regen docs / examples for 2.2.2

I wonder if introducing "mp" as a shortcut to monkeypatch
is a good idea, actually :)

.hgtags

@@ -31,3 +31,18 @@ c59d3fa8681a5b5966b8375b16fccd64a3a8dbeb 1.3.3
 79ef6377705184c55633d456832eea318fedcf61 1.3.4
 79ef6377705184c55633d456832eea318fedcf61 1.3.4
 90fffd35373e9f125af233f78b19416f0938d841 1.3.4
+e9e127acd6f0497324ef7f40cfb997cad4c4cd17 2.0.0
+e4497c2aed358c1988cf7be83ca9394c3c707fa2 2.0.1
+84e5c54b72448194a0f6f815da7e048ac8019d50 2.0.2
+2ef82d82daacb72733a3a532a95c5a37164e5819 2.0.3
+2ef82d82daacb72733a3a532a95c5a37164e5819 2.0.3
+c777dcad166548b7499564cb49ae5c8b4b07f935 2.0.3
+c777dcad166548b7499564cb49ae5c8b4b07f935 2.0.3
+49f11dbff725acdcc5fe3657cbcdf9ae04e25bbc 2.0.3
+49f11dbff725acdcc5fe3657cbcdf9ae04e25bbc 2.0.3
+363e5a5a59c803e6bc176a6f9cc4bf1a1ca2dab0 2.0.3
+e5e1746a197f0398356a43fbe2eebac9690f795d 2.1.0
+5864412c6f3c903384243bd315639d101d7ebc67 2.1.2
+12a05d59249f80276e25fd8b96e8e545b1332b7a 2.1.3
+1522710369337d96bf9568569d5f0ca9b38a74e0 2.2.0
+3da8cec6c5326ed27c144c9b6d7a64a648370005 2.2.1

AUTHORS

@@ -1,16 +1,17 @@
 Holger Krekel, holger at merlinux eu
-Benjamin Peterson, benjamin at python org
-Ronny Pfannschmidt, Ronny.Pfannschmidt at gmx de
-Guido Wesdorp, johnny at johnnydebris net 
-Samuele Pedroni, pedronis at openend se 
-Carl Friedrich Bolz, cfbolz at gmx de
-Armin Rigo, arigo at tunes org 
-Maciek Fijalkowski, fijal at genesilico pl
-Brian Dorsey, briandorsey at gmail com 
 merlinux GmbH, Germany, office at merlinux eu
 
 Contributors include:: 
 
+Ronny Pfannschmidt
+Benjamin Peterson
+Floris Bruynooghe
+Samuele Pedroni
+Carl Friedrich Bolz
+Armin Rigo
+Maciek Fijalkowski
+Guido Wesdorp
+Brian Dorsey
 Ross Lawley
 Ralf Schmitt
 Chris Lamb 
@@ -21,3 +22,4 @@ Jan Balster
 Grig Gheorghiu
 Bob Ippolito
 Christian Tismer
+Daniel Nuri

CHANGELOG

@@ -1,4 +1,249 @@
-Changes between 1.3.4 and 2.0.0dev0
+Changes between 2.2.1 and 2.2.2
+----------------------------------------
+
+- fix issue101: wrong args to unittest.TestCase test function now
+  produce better output
+- fix issue102: report more useful errors and hints for when a 
+  test directory was renamed and some pyc/__pycache__ remain
+- fix issue106: allow parametrize to be applied multiple times
+  e.g. from module, class and at function level.
+- fix issue107: actually perform session scope finalization
+- don't check in parametrize if indirect parameters are funcarg names
+- add chdir method to monkeypatch funcarg
+- fix crash resulting from calling monkeypatch undo a second time
+- fix issue115: make --collectonly robust against early failure
+  (missing files/directories)
+- "-qq --collectonly" now shows only files and the number of tests in them
+- "-q --collectonly" now shows test ids
+- allow adding of attributes to test reports such that it also works
+  with distributed testing (no upgrade of pytest-xdist needed)
+
+Changes between 2.2.0 and 2.2.1
+----------------------------------------
+
+- fix issue99 (in pytest and py) internallerrors with resultlog now
+  produce better output - fixed by normalizing pytest_internalerror 
+  input arguments.
+- fix issue97 / traceback issues (in pytest and py) improve traceback output
+  in conjunction with jinja2 and cython which hack tracebacks
+- fix issue93 (in pytest and pytest-xdist) avoid "delayed teardowns":
+  the final test in a test node will now run its teardown directly
+  instead of waiting for the end of the session. Thanks Dave Hunt for
+  the good reporting and feedback.  The pytest_runtest_protocol as well
+  as the pytest_runtest_teardown hooks now have "nextitem" available 
+  which will be None indicating the end of the test run.
+- fix collection crash due to unknown-source collected items, thanks
+  to Ralf Schmitt (fixed by depending on a more recent pylib)
+
+Changes between 2.1.3 and 2.2.0
+----------------------------------------
+
+- fix issue90: introduce eager tearing down of test items so that
+  teardown function are called earlier.
+- add an all-powerful metafunc.parametrize function which allows to 
+  parametrize test function arguments in multiple steps and therefore
+  from indepdenent plugins and palces. 
+- add a @pytest.mark.parametrize helper which allows to easily
+  call a test function with different argument values
+- Add examples to the "parametrize" example page, including a quick port 
+  of Test scenarios and the new parametrize function and decorator.
+- introduce registration for "pytest.mark.*" helpers via ini-files
+  or through plugin hooks.  Also introduce a "--strict" option which 
+  will treat unregistered markers as errors
+  allowing to avoid typos and maintain a well described set of markers
+  for your test suite.  See exaples at http://pytest.org/latest/mark.html
+  and its links.
+- issue50: introduce "-m marker" option to select tests based on markers
+  (this is a stricter and more predictable version of '-k' in that "-m"
+  only matches complete markers and has more obvious rules for and/or
+  semantics.
+- new feature to help optimizing the speed of your tests: 
+  --durations=N option for displaying N slowest test calls 
+  and setup/teardown methods.
+- fix issue87: --pastebin now works with python3
+- fix issue89: --pdb with unexpected exceptions in doctest work more sensibly
+- fix and cleanup pytest's own test suite to not leak FDs 
+- fix issue83: link to generated funcarg list
+- fix issue74: pyarg module names are now checked against imp.find_module false positives
+- fix compatibility with twisted/trial-11.1.0 use cases
+- simplify Node.listchain
+- simplify junitxml output code by relying on py.xml
+- add support for skip properties on unittest classes and functions
+
+Changes between 2.1.2 and 2.1.3
+----------------------------------------
+
+- fix issue79: assertion rewriting failed on some comparisons in boolops
+- correctly handle zero length arguments (a la pytest '')
+- fix issue67 / junitxml now contains correct test durations, thanks ronny
+- fix issue75 / skipping test failure on jython
+- fix issue77 / Allow assertrepr_compare hook to apply to a subset of tests
+
+Changes between 2.1.1 and 2.1.2
+----------------------------------------
+
+- fix assertion rewriting on files with windows newlines on some Python versions
+- refine test discovery by package/module name (--pyargs), thanks Florian Mayer
+- fix issue69 / assertion rewriting fixed on some boolean operations
+- fix issue68 / packages now work with assertion rewriting
+- fix issue66: use different assertion rewriting caches when the -O option is passed
+- don't try assertion rewriting on Jython, use reinterp
+
+Changes between 2.1.0 and 2.1.1
+----------------------------------------------
+
+- fix issue64 / pytest.set_trace now works within pytest_generate_tests hooks
+- fix issue60 / fix error conditions involving the creation of __pycache__
+- fix issue63 / assertion rewriting on inserts involving strings containing '%'
+- fix assertion rewriting on calls with a ** arg
+- don't cache rewritten modules if bytecode generation is disabled
+- fix assertion rewriting in read-only directories
+- fix issue59: provide system-out/err tags for junitxml output
+- fix issue61: assertion rewriting on boolean operations with 3 or more operands
+- you can now build a man page with "cd doc ; make man"
+
+Changes between 2.0.3 and 2.1.0.DEV
+----------------------------------------------
+
+- fix issue53 call nosestyle setup functions with correct ordering
+- fix issue58 and issue59: new assertion code fixes
+- merge Benjamin's assertionrewrite branch: now assertions
+  for test modules on python 2.6 and above are done by rewriting
+  the AST and saving the pyc file before the test module is imported.
+  see doc/assert.txt for more info.
+- fix issue43: improve doctests with better traceback reporting on
+  unexpected exceptions
+- fix issue47: timing output in junitxml for test cases is now correct
+- fix issue48: typo in MarkInfo repr leading to exception
+- fix issue49: avoid confusing error when initizaliation partially fails
+- fix issue44: env/username expansion for junitxml file path
+- show releaselevel information in test runs for pypy
+- reworked doc pages for better navigation and PDF generation
+- report KeyboardInterrupt even if interrupted during session startup
+- fix issue 35 - provide PDF doc version and download link from index page
+
+Changes between 2.0.2 and 2.0.3
+----------------------------------------------
+
+- fix issue38: nicer tracebacks on calls to hooks, particularly early
+  configure/sessionstart ones
+
+- fix missing skip reason/meta information in junitxml files, reported
+  via http://lists.idyll.org/pipermail/testing-in-python/2011-March/003928.html
+
+- fix issue34: avoid collection failure with "test" prefixed classes
+  deriving from object.
+
+- don't require zlib (and other libs) for genscript plugin without
+  --genscript actually being used.
+
+- speed up skips (by not doing a full traceback represenation
+  internally)
+
+- fix issue37: avoid invalid characters in junitxml's output
+
+Changes between 2.0.1 and 2.0.2
+----------------------------------------------
+
+- tackle issue32 - speed up test runs of very quick test functions
+  by reducing the relative overhead
+
+- fix issue30 - extended xfail/skipif handling and improved reporting.
+  If you have a syntax error in your skip/xfail
+  expressions you now get nice error reports.
+
+  Also you can now access module globals from xfail/skipif
+  expressions so that this for example works now::
+ 
+    import pytest
+    import mymodule
+    @pytest.mark.skipif("mymodule.__version__[0] == "1")
+    def test_function():
+        pass
+
+  This will not run the test function if the module's version string 
+  does not start with a "1".  Note that specifying a string instead
+  of a boolean expressions allows py.test to report meaningful information 
+  when summarizing a test run as to what conditions lead to skipping 
+  (or xfail-ing) tests.
+
+- fix issue28 - setup_method and pytest_generate_tests work together
+  The setup_method fixture method now gets called also for
+  test function invocations generated from the pytest_generate_tests
+  hook.
+
+- fix issue27 - collectonly and keyword-selection (-k) now work together
+  Also, if you do "py.test --collectonly -q" you now get a flat list
+  of test ids that you can use to paste to the py.test commandline
+  in order to execute a particular test.
+
+- fix issue25 avoid reported problems with --pdb and python3.2/encodings output
+
+- fix issue23 - tmpdir argument now works on Python3.2 and WindowsXP
+  Starting with Python3.2 os.symlink may be supported. By requiring
+  a newer py lib version the py.path.local() implementation acknowledges
+  this.
+
+- fixed typos in the docs (thanks Victor Garcia, Brianna Laugher) and particular
+  thanks to Laura Creighton who also revieved parts of the documentation.
+
+- fix slighly wrong output of verbose progress reporting for classes 
+  (thanks Amaury)
+
+- more precise (avoiding of) deprecation warnings for node.Class|Function accesses
+
+- avoid std unittest assertion helper code in tracebacks (thanks Ronny)
+
+Changes between 2.0.0 and 2.0.1
+----------------------------------------------
+
+- refine and unify initial capturing so that it works nicely
+  even if the logging module is used on an early-loaded conftest.py
+  file or plugin.
+- allow to omit "()" in test ids to allow for uniform test ids
+  as produced by Alfredo's nice pytest.vim plugin.
+- fix issue12 - show plugin versions with "--version" and
+  "--traceconfig" and also document how to add extra information 
+  to reporting test header
+- fix issue17 (import-* reporting issue on python3) by
+  requiring py>1.4.0 (1.4.1 is going to include it)
+- fix issue10 (numpy arrays truth checking) by refining
+  assertion interpretation in py lib
+- fix issue15: make nose compatibility tests compatible
+  with python3 (now that nose-1.0 supports python3)
+- remove somewhat surprising "same-conftest" detection because
+  it ignores conftest.py when they appear in several subdirs.
+- improve assertions ("not in"), thanks Floris Bruynooghe
+- improve behaviour/warnings when running on top of "python -OO"
+  (assertions and docstrings are turned off, leading to potential
+  false positives)
+- introduce a pytest_cmdline_processargs(args) hook
+  to allow dynamic computation of command line arguments.
+  This fixes a regression because py.test prior to 2.0
+  allowed to set command line options from conftest.py
+  files which so far pytest-2.0 only allowed from ini-files now.
+- fix issue7: assert failures in doctest modules.
+  unexpected failures in doctests will not generally
+  show nicer, i.e. within the doctest failing context.
+- fix issue9: setup/teardown functions for an xfail-marked
+  test will report as xfail if they fail but report as normally
+  passing (not xpassing) if they succeed.  This only is true
+  for "direct" setup/teardown invocations because teardown_class/
+  teardown_module cannot closely relate to a single test.
+- fix issue14: no logging errors at process exit
+- refinements to "collecting" output on non-ttys
+- refine internal plugin registration and --traceconfig output
+- introduce a mechanism to prevent/unregister plugins from the 
+  command line, see http://pytest.org/plugins.html#cmdunregister
+- activate resultlog plugin by default
+- fix regression wrt yielded tests which due to the
+  collection-before-running semantics were not 
+  setup as with pytest 1.3.4.  Note, however, that
+  the recommended and much cleaner way to do test 
+  parametraization remains the "pytest_generate_tests"
+  mechanism, see the docs.
+
+Changes between 1.3.4 and 2.0.0
 ----------------------------------------------
 
 - pytest-2.0 is now its own package and depends on pylib-2.0

ISSUES.txt

@@ -1,3 +1,24 @@
+refine parametrize API in 2.2 series
+-------------------------------------------------------------
+tags: critical feature 2.2
+
+extend metafunc.parametrize to better support indirection
+by specifying a setupfunc(request, val) which will _substitute_
+the funcarg factory.   Here is an example:
+
+    def setupdb(request, val):
+        # setup "resource" based on test request and the values passed 
+        # in to parametrize.  setupfunc is called for each such value.
+        # you may use request.addfinalizer() or request.cached_setup ...
+        return db
+
+    @pytest.mark.parametrize("db", ["pg", "mysql"], setupfunc=setupdb)
+    def test_heavy_functional_test(db):
+        ...
+
+There would be no need to write funcarg factories for this example, only
+to explain the attributes and functionality of "request".
+
 checks / deprecations for next release
 ---------------------------------------------------------------
 tags: bug 2.4 core xdist
@@ -7,17 +28,27 @@ tags: bug 2.4 core xdist
   the protocol now - setup/teardown is called at module level.
   consider making calling of setup/teardown configurable
 
-profiling / hook call optimization
--------------------------------------
-tags: enhancement 2.1
+optimizations 
+---------------------------------------------------------------
+tags: 2.4 core
 
-bench/bench.py reveals that for very quick running
-unit tests the hook architecture is a bit slow.
-Profile and improve hook calls.
+- look at ihook optimization such that all lookups for
+  hooks relating to the same fspath are cached.
+
+fix start/finish partial finailization problem
+---------------------------------------------------------------
+tags: bug core
+
+if a configure/runtest_setup/sessionstart/... hook invocation partially
+fails the sessionfinishes is not called.  Each hook implementation
+should better be repsonsible for registering a cleanup/finalizer
+appropriately to avoid this issue.  Moreover/Alternatively, we could 
+record which implementations of a hook succeeded and only call their
+teardown.
 
 do early-teardown of test modules
 -----------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 currently teardowns are called when the next tests is setup
 except for the function/method level where interally
@@ -29,7 +60,7 @@ prints of teardown-code appear in the setup of the next test.
 
 consider and document __init__ file usage in test directories
 ---------------------------------------------------------------
-tags: bug 2.1 core
+tags: bug 2.3 core
 
 Currently, a test module is imported with its fully qualified
 package path, determined by checking __init__ files upwards.
@@ -44,7 +75,7 @@ certain scenarios makes sense.
 
 relax requirement to have tests/testing contain an __init__
 ----------------------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 bb: http://bitbucket.org/hpk42/py-trunk/issue/64
 
 A local test run of a "tests" directory may work
@@ -55,25 +86,24 @@ i.e. port the nose-logic of unloading a test module.
 
 customize test function collection
 -------------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 - introduce py.test.mark.nocollect for not considering a function for
   test collection at all.  maybe also introduce a py.test.mark.test to
   explicitely mark a function to become a tested one.  Lookup JUnit ways
   of tagging tests.
 
-- allow an easy way to customize "test_", "Test" prefixes for file paths
-  and test function/class names. the current customizable Item requires
-  too much code/concepts to influence this collection matching.
-  maybe introduce pytest_pycollect_filters = {
-    'file': 'test*.py',
-    'function': 'test*',
-    'class': 'Test*',
-  }
+introduce pytest.mark.importorskip
+-------------------------------------------------------
+tags: feature 2.3
+
+in addition to the imperative pytest.importorskip also introduce
+a pytest.mark.importorskip so that the test count is more correct.
+
 
 introduce py.test.mark.platform
 -------------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 Introduce nice-to-spell platform-skipping, examples:
 
@@ -88,9 +118,19 @@ etc. Idea is to allow Python expressions which can operate
 on common spellings for operating systems and python
 interpreter versions.
 
+pytest.mark.xfail signature change
+-------------------------------------------------------
+tags: feature 2.3
+
+change to pytest.mark.xfail(reason, (optional)condition)
+to better implement the word meaning.  It also signals
+better that we always have some kind of an implementation
+reason that can be formualated.
+Compatibility? how to introduce a new name/keep compat?
+
 introduce py.test.mark registration
 -----------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 introduce a hook that allows to register a named mark decorator
 with documentation and add "py.test --marks" to get
@@ -99,7 +139,7 @@ definitions.
 
 allow to non-intrusively apply skipfs/xfail/marks
 ---------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 use case: mark a module or directory structures
 to be skipped on certain platforms (i.e. no import
@@ -110,14 +150,14 @@ from conftests or plugins.
 
 explicit referencing of conftest.py files
 -----------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 allow to name conftest.py files (in sub directories) that should
 be imported early, as to include command line options.
 
 improve central py.test ini file
 ----------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 introduce more declarative configuration options:
 - (to-be-collected test directories)
@@ -128,33 +168,16 @@ introduce more declarative configuration options:
 
 new documentation
 ----------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 - logo py.test
 - examples for unittest or functional testing
 - resource management for functional testing
 - patterns: page object
-- parametrized testing
-- better / more integrated plugin docs
-
-generalize parametrized testing to generate combinations
--------------------------------------------------------------
-tags: feature 2.1
-
-think about extending metafunc.addcall or add a new method to allow to
-generate tests with combinations of all generated versions - what to do
-about "id" and "param" in such combinations though?
-
-introduce py.test.mark.multi
------------------------------------------
-tags: feature 1.3
-
-introduce py.test.mark.multi to specify a number
-of values for a given function argument.
 
 have imported module mismatch honour relative paths
 --------------------------------------------------------
-tags: bug 2.1
+tags: bug 2.3
 
 With 1.1.1 py.test fails at least on windows if an import
 is relative and compared against an absolute conftest.py
@@ -162,7 +185,7 @@ path. Normalize.
 
 call termination with small timeout
 -------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 test: testing/pytest/dist/test_dsession.py - test_terminate_on_hanging_node
 
 Call gateway group termination with a small timeout if available.
@@ -170,7 +193,7 @@ Should make dist-testing less likely to leave lost processes.
 
 consider globals: py.test.ensuretemp and config
 --------------------------------------------------------------
-tags: experimental-wish 2.1
+tags: experimental-wish 2.3
 
 consider deprecating py.test.ensuretemp and py.test.config
 to further reduce py.test globality.  Also consider
@@ -179,7 +202,7 @@ a plugin rather than being there from the start.
 
 consider allowing funcargs for setup methods
 --------------------------------------------------------------
-tags: experimental-wish 2.1
+tags: experimental-wish 2.3
 
 Users have expressed the wish to have funcargs available to setup
 functions.  Experiment with allowing funcargs there - it might
@@ -189,13 +212,20 @@ factories with a request object that not have a cls/function
 attributes.  However, how to handle parametrized test functions
 and funcargs?
 
-setup_function -> request can be like it is now
-setup_class -> request has no request.function
-setup_module -> request has no request.cls
+maybe introduce a setup method like:
+
+    setup_invocation(self, request)
+
+which has full access to the test invocation through "request"
+through which you can get funcargvalues, use cached_setup etc.
+Therefore, the access to funcargs would be indirect but it
+could be consistently implemented.  setup_invocation() would
+be a "glue" function for bringing together the xUnit and funcargs
+world.
 
 consider pytest_addsyspath hook
 -----------------------------------------
-tags: 2.1
+tags: 2.3
 
 py.test could call a new pytest_addsyspath() in order to systematically
 allow manipulation of sys.path and to inhibit it via --no-addsyspath
@@ -207,7 +237,7 @@ and pytest_configure.
 
 show plugin information in test header
 ----------------------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 Now that external plugins are becoming more numerous
 it would be useful to have external plugins along with
@@ -215,7 +245,7 @@ their versions displayed as a header line.
 
 deprecate global py.test.config usage
 ----------------------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 py.test.ensuretemp and py.test.config are probably the last
 objects containing global state.  Often using them is not
@@ -225,9 +255,32 @@ as others.
 
 remove deprecated bits in collect.py
 -------------------------------------------------------------------
-tags: feature 2.1
+tags: feature 2.3
 
 In an effort to further simplify code, review and remove deprecated bits
 in collect.py.  Probably good:
 - inline consider_file/dir methods, no need to have them
   subclass-overridable because of hooks
+
+implement fslayout decorator
+---------------------------------
+tags: feature 2.3
+
+Improve the way how tests can work with pre-made examples,
+keeping the layout close to the test function:
+
+@pytest.mark.fslayout("""
+    conftest.py:
+        #  empty
+    tests/
+        test_%(NAME)s:  # becomes test_run1.py
+            def test_function(self):
+                pass
+""")
+def test_run(pytester, fslayout):
+    p = fslayout.find("test_*.py")
+    result = pytester.runpytest(p)
+    assert result.ret == 0
+    assert result.passed == 1
+
+

_pytest/__init__.py

@@ -1 +1,2 @@
 #
+__version__ = '2.2.2'

_pytest/assertion/__init__.py

@@ -0,0 +1,119 @@
+"""
+support for presenting detailed information in failing assertions.
+"""
+import py
+import sys
+import pytest
+from _pytest.monkeypatch import monkeypatch
+from _pytest.assertion import util
+
+def pytest_addoption(parser):
+    group = parser.getgroup("debugconfig")
+    group.addoption('--assert', action="store", dest="assertmode",
+                    choices=("rewrite", "reinterp", "plain",),
+                    default="rewrite", metavar="MODE",
+                    help="""control assertion debugging tools.
+'plain' performs no assertion debugging.
+'reinterp' reinterprets assert statements after they failed to provide assertion expression information.
+'rewrite' (the default) rewrites assert statements in test modules on import
+to provide assert expression information. """)
+    group.addoption('--no-assert', action="store_true", default=False,
+        dest="noassert", help="DEPRECATED equivalent to --assert=plain")
+    group.addoption('--nomagic', action="store_true", default=False,
+        dest="nomagic", help="DEPRECATED equivalent to --assert=plain")
+
+class AssertionState:
+    """State for the assertion plugin."""
+
+    def __init__(self, config, mode):
+        self.mode = mode
+        self.trace = config.trace.root.get("assertion")
+
+def pytest_configure(config):
+    mode = config.getvalue("assertmode")
+    if config.getvalue("noassert") or config.getvalue("nomagic"):
+        mode = "plain"
+    if mode == "rewrite":
+        try:
+            import ast
+        except ImportError:
+            mode = "reinterp"
+        else:
+            if sys.platform.startswith('java'):
+                mode = "reinterp"
+    if mode != "plain":
+        _load_modules(mode)
+        m = monkeypatch()
+        config._cleanup.append(m.undo)
+        m.setattr(py.builtin.builtins, 'AssertionError',
+                  reinterpret.AssertionError)
+    hook = None
+    if mode == "rewrite":
+        hook = rewrite.AssertionRewritingHook()
+        sys.meta_path.append(hook)
+    warn_about_missing_assertion(mode)
+    config._assertstate = AssertionState(config, mode)
+    config._assertstate.hook = hook
+    config._assertstate.trace("configured with mode set to %r" % (mode,))
+
+def pytest_unconfigure(config):
+    hook = config._assertstate.hook
+    if hook is not None:
+        sys.meta_path.remove(hook)
+
+def pytest_collection(session):
+    # this hook is only called when test modules are collected
+    # so for example not in the master process of pytest-xdist
+    # (which does not collect test modules)
+    hook = session.config._assertstate.hook
+    if hook is not None:
+        hook.set_session(session)
+
+def pytest_runtest_setup(item):
+    def callbinrepr(op, left, right):
+        hook_result = item.ihook.pytest_assertrepr_compare(
+            config=item.config, op=op, left=left, right=right)
+        for new_expl in hook_result:
+            if new_expl:
+                res = '\n~'.join(new_expl)
+                if item.config.getvalue("assertmode") == "rewrite":
+                    # The result will be fed back a python % formatting
+                    # operation, which will fail if there are extraneous
+                    # '%'s in the string. Escape them here.
+                    res = res.replace("%", "%%")
+                return res
+    util._reprcompare = callbinrepr
+
+def pytest_runtest_teardown(item):
+    util._reprcompare = None
+
+def pytest_sessionfinish(session):
+    hook = session.config._assertstate.hook
+    if hook is not None:
+        hook.session = None
+
+def _load_modules(mode):
+    """Lazily import assertion related code."""
+    global rewrite, reinterpret
+    from _pytest.assertion import reinterpret
+    if mode == "rewrite":
+        from _pytest.assertion import rewrite
+
+def warn_about_missing_assertion(mode):
+    try:
+        assert False
+    except AssertionError:
+        pass
+    else:
+        if mode == "rewrite":
+            specifically = ("assertions which are not in test modules "
+                            "will be ignored")
+        else:
+            specifically = "failing tests may report as passing"
+
+        sys.stderr.write("WARNING: " + specifically +
+                        " because assert statements are not executed "
+                        "by the underlying Python interpreter "
+                        "(are you using python -O?)\n")
+
+pytest_assertrepr_compare = util.assertrepr_compare

_pytest/assertion/newinterpret.py

@@ -0,0 +1,333 @@
+"""
+Find intermediate evalutation results in assert statements through builtin AST.
+This should replace oldinterpret.py eventually.
+"""
+
+import sys
+import ast
+
+import py
+from _pytest.assertion import util
+from _pytest.assertion.reinterpret import BuiltinAssertionError
+
+
+if sys.platform.startswith("java") and sys.version_info < (2, 5, 2):
+    # See http://bugs.jython.org/issue1497
+    _exprs = ("BoolOp", "BinOp", "UnaryOp", "Lambda", "IfExp", "Dict",
+              "ListComp", "GeneratorExp", "Yield", "Compare", "Call",
+              "Repr", "Num", "Str", "Attribute", "Subscript", "Name",
+              "List", "Tuple")
+    _stmts = ("FunctionDef", "ClassDef", "Return", "Delete", "Assign",
+              "AugAssign", "Print", "For", "While", "If", "With", "Raise",
+              "TryExcept", "TryFinally", "Assert", "Import", "ImportFrom",
+              "Exec", "Global", "Expr", "Pass", "Break", "Continue")
+    _expr_nodes = set(getattr(ast, name) for name in _exprs)
+    _stmt_nodes = set(getattr(ast, name) for name in _stmts)
+    def _is_ast_expr(node):
+        return node.__class__ in _expr_nodes
+    def _is_ast_stmt(node):
+        return node.__class__ in _stmt_nodes
+else:
+    def _is_ast_expr(node):
+        return isinstance(node, ast.expr)
+    def _is_ast_stmt(node):
+        return isinstance(node, ast.stmt)
+
+
+class Failure(Exception):
+    """Error found while interpreting AST."""
+
+    def __init__(self, explanation=""):
+        self.cause = sys.exc_info()
+        self.explanation = explanation
+
+
+def interpret(source, frame, should_fail=False):
+    mod = ast.parse(source)
+    visitor = DebugInterpreter(frame)
+    try:
+        visitor.visit(mod)
+    except Failure:
+        failure = sys.exc_info()[1]
+        return getfailure(failure)
+    if should_fail:
+        return ("(assertion failed, but when it was re-run for "
+                "printing intermediate values, it did not fail.  Suggestions: "
+                "compute assert expression before the assert or use --assert=plain)")
+
+def run(offending_line, frame=None):
+    if frame is None:
+        frame = py.code.Frame(sys._getframe(1))
+    return interpret(offending_line, frame)
+
+def getfailure(e):
+    explanation = util.format_explanation(e.explanation)
+    value = e.cause[1]
+    if str(value):
+        lines = explanation.split('\n')
+        lines[0] += "  << %s" % (value,)
+        explanation = '\n'.join(lines)
+    text = "%s: %s" % (e.cause[0].__name__, explanation)
+    if text.startswith('AssertionError: assert '):
+        text = text[16:]
+    return text
+
+operator_map = {
+    ast.BitOr : "|",
+    ast.BitXor : "^",
+    ast.BitAnd : "&",
+    ast.LShift : "<<",
+    ast.RShift : ">>",
+    ast.Add : "+",
+    ast.Sub : "-",
+    ast.Mult : "*",
+    ast.Div : "/",
+    ast.FloorDiv : "//",
+    ast.Mod : "%",
+    ast.Eq : "==",
+    ast.NotEq : "!=",
+    ast.Lt : "<",
+    ast.LtE : "<=",
+    ast.Gt : ">",
+    ast.GtE : ">=",
+    ast.Pow : "**",
+    ast.Is : "is",
+    ast.IsNot : "is not",
+    ast.In : "in",
+    ast.NotIn : "not in"
+}
+
+unary_map = {
+    ast.Not : "not %s",
+    ast.Invert : "~%s",
+    ast.USub : "-%s",
+    ast.UAdd : "+%s"
+}
+
+
+class DebugInterpreter(ast.NodeVisitor):
+    """Interpret AST nodes to gleam useful debugging information. """
+
+    def __init__(self, frame):
+        self.frame = frame
+
+    def generic_visit(self, node):
+        # Fallback when we don't have a special implementation.
+        if _is_ast_expr(node):
+            mod = ast.Expression(node)
+            co = self._compile(mod)
+            try:
+                result = self.frame.eval(co)
+            except Exception:
+                raise Failure()
+            explanation = self.frame.repr(result)
+            return explanation, result
+        elif _is_ast_stmt(node):
+            mod = ast.Module([node])
+            co = self._compile(mod, "exec")
+            try:
+                self.frame.exec_(co)
+            except Exception:
+                raise Failure()
+            return None, None
+        else:
+            raise AssertionError("can't handle %s" %(node,))
+
+    def _compile(self, source, mode="eval"):
+        return compile(source, "<assertion interpretation>", mode)
+
+    def visit_Expr(self, expr):
+        return self.visit(expr.value)
+
+    def visit_Module(self, mod):
+        for stmt in mod.body:
+            self.visit(stmt)
+
+    def visit_Name(self, name):
+        explanation, result = self.generic_visit(name)
+        # See if the name is local.
+        source = "%r in locals() is not globals()" % (name.id,)
+        co = self._compile(source)
+        try:
+            local = self.frame.eval(co)
+        except Exception:
+            # have to assume it isn't
+            local = None
+        if local is None or not self.frame.is_true(local):
+            return name.id, result
+        return explanation, result
+
+    def visit_Compare(self, comp):
+        left = comp.left
+        left_explanation, left_result = self.visit(left)
+        for op, next_op in zip(comp.ops, comp.comparators):
+            next_explanation, next_result = self.visit(next_op)
+            op_symbol = operator_map[op.__class__]
+            explanation = "%s %s %s" % (left_explanation, op_symbol,
+                                        next_explanation)
+            source = "__exprinfo_left %s __exprinfo_right" % (op_symbol,)
+            co = self._compile(source)
+            try:
+                result = self.frame.eval(co, __exprinfo_left=left_result,
+                                         __exprinfo_right=next_result)
+            except Exception:
+                raise Failure(explanation)
+            try:
+                if not self.frame.is_true(result):
+                    break
+            except KeyboardInterrupt:
+                raise
+            except:
+                break
+            left_explanation, left_result = next_explanation, next_result
+
+        if util._reprcompare is not None:
+            res = util._reprcompare(op_symbol, left_result, next_result)
+            if res:
+                explanation = res
+        return explanation, result
+
+    def visit_BoolOp(self, boolop):
+        is_or = isinstance(boolop.op, ast.Or)
+        explanations = []
+        for operand in boolop.values:
+            explanation, result = self.visit(operand)
+            explanations.append(explanation)
+            if result == is_or:
+                break
+        name = is_or and " or " or " and "
+        explanation = "(" + name.join(explanations) + ")"
+        return explanation, result
+
+    def visit_UnaryOp(self, unary):
+        pattern = unary_map[unary.op.__class__]
+        operand_explanation, operand_result = self.visit(unary.operand)
+        explanation = pattern % (operand_explanation,)
+        co = self._compile(pattern % ("__exprinfo_expr",))
+        try:
+            result = self.frame.eval(co, __exprinfo_expr=operand_result)
+        except Exception:
+            raise Failure(explanation)
+        return explanation, result
+
+    def visit_BinOp(self, binop):
+        left_explanation, left_result = self.visit(binop.left)
+        right_explanation, right_result = self.visit(binop.right)
+        symbol = operator_map[binop.op.__class__]
+        explanation = "(%s %s %s)" % (left_explanation, symbol,
+                                      right_explanation)
+        source = "__exprinfo_left %s __exprinfo_right" % (symbol,)
+        co = self._compile(source)
+        try:
+            result = self.frame.eval(co, __exprinfo_left=left_result,
+                                     __exprinfo_right=right_result)
+        except Exception:
+            raise Failure(explanation)
+        return explanation, result
+
+    def visit_Call(self, call):
+        func_explanation, func = self.visit(call.func)
+        arg_explanations = []
+        ns = {"__exprinfo_func" : func}
+        arguments = []
+        for arg in call.args:
+            arg_explanation, arg_result = self.visit(arg)
+            arg_name = "__exprinfo_%s" % (len(ns),)
+            ns[arg_name] = arg_result
+            arguments.append(arg_name)
+            arg_explanations.append(arg_explanation)
+        for keyword in call.keywords:
+            arg_explanation, arg_result = self.visit(keyword.value)
+            arg_name = "__exprinfo_%s" % (len(ns),)
+            ns[arg_name] = arg_result
+            keyword_source = "%s=%%s" % (keyword.arg)
+            arguments.append(keyword_source % (arg_name,))
+            arg_explanations.append(keyword_source % (arg_explanation,))
+        if call.starargs:
+            arg_explanation, arg_result = self.visit(call.starargs)
+            arg_name = "__exprinfo_star"
+            ns[arg_name] = arg_result
+            arguments.append("*%s" % (arg_name,))
+            arg_explanations.append("*%s" % (arg_explanation,))
+        if call.kwargs:
+            arg_explanation, arg_result = self.visit(call.kwargs)
+            arg_name = "__exprinfo_kwds"
+            ns[arg_name] = arg_result
+            arguments.append("**%s" % (arg_name,))
+            arg_explanations.append("**%s" % (arg_explanation,))
+        args_explained = ", ".join(arg_explanations)
+        explanation = "%s(%s)" % (func_explanation, args_explained)
+        args = ", ".join(arguments)
+        source = "__exprinfo_func(%s)" % (args,)
+        co = self._compile(source)
+        try:
+            result = self.frame.eval(co, **ns)
+        except Exception:
+            raise Failure(explanation)
+        pattern = "%s\n{%s = %s\n}"
+        rep = self.frame.repr(result)
+        explanation = pattern % (rep, rep, explanation)
+        return explanation, result
+
+    def _is_builtin_name(self, name):
+        pattern = "%r not in globals() and %r not in locals()"
+        source = pattern % (name.id, name.id)
+        co = self._compile(source)
+        try:
+            return self.frame.eval(co)
+        except Exception:
+            return False
+
+    def visit_Attribute(self, attr):
+        if not isinstance(attr.ctx, ast.Load):
+            return self.generic_visit(attr)
+        source_explanation, source_result = self.visit(attr.value)
+        explanation = "%s.%s" % (source_explanation, attr.attr)
+        source = "__exprinfo_expr.%s" % (attr.attr,)
+        co = self._compile(source)
+        try:
+            result = self.frame.eval(co, __exprinfo_expr=source_result)
+        except Exception:
+            raise Failure(explanation)
+        explanation = "%s\n{%s = %s.%s\n}" % (self.frame.repr(result),
+                                              self.frame.repr(result),
+                                              source_explanation, attr.attr)
+        # Check if the attr is from an instance.
+        source = "%r in getattr(__exprinfo_expr, '__dict__', {})"
+        source = source % (attr.attr,)
+        co = self._compile(source)
+        try:
+            from_instance = self.frame.eval(co, __exprinfo_expr=source_result)
+        except Exception:
+            from_instance = None
+        if from_instance is None or self.frame.is_true(from_instance):
+            rep = self.frame.repr(result)
+            pattern = "%s\n{%s = %s\n}"
+            explanation = pattern % (rep, rep, explanation)
+        return explanation, result
+
+    def visit_Assert(self, assrt):
+        test_explanation, test_result = self.visit(assrt.test)
+        explanation = "assert %s" % (test_explanation,)
+        if not self.frame.is_true(test_result):
+            try:
+                raise BuiltinAssertionError
+            except Exception:
+                raise Failure(explanation)
+        return explanation, test_result
+
+    def visit_Assign(self, assign):
+        value_explanation, value_result = self.visit(assign.value)
+        explanation = "... = %s" % (value_explanation,)
+        name = ast.Name("__exprinfo_expr", ast.Load(),
+                        lineno=assign.value.lineno,
+                        col_offset=assign.value.col_offset)
+        new_assign = ast.Assign(assign.targets, name, lineno=assign.lineno,
+                                col_offset=assign.col_offset)
+        mod = ast.Module([new_assign])
+        co = self._compile(mod, "exec")
+        try:
+            self.frame.exec_(co, __exprinfo_expr=value_result)
+        except Exception:
+            raise Failure(explanation)
+        return explanation, value_result

_pytest/assertion/oldinterpret.py

@@ -0,0 +1,552 @@
+import py
+import sys, inspect
+from compiler import parse, ast, pycodegen
+from _pytest.assertion.util import format_explanation
+from _pytest.assertion.reinterpret import BuiltinAssertionError
+
+passthroughex = py.builtin._sysex
+
+class Failure:
+    def __init__(self, node):
+        self.exc, self.value, self.tb = sys.exc_info()
+        self.node = node
+
+class View(object):
+    """View base class.
+
+    If C is a subclass of View, then C(x) creates a proxy object around
+    the object x.  The actual class of the proxy is not C in general,
+    but a *subclass* of C determined by the rules below.  To avoid confusion
+    we call view class the class of the proxy (a subclass of C, so of View)
+    and object class the class of x.
+
+    Attributes and methods not found in the proxy are automatically read on x.
+    Other operations like setting attributes are performed on the proxy, as
+    determined by its view class.  The object x is available from the proxy
+    as its __obj__ attribute.
+
+    The view class selection is determined by the __view__ tuples and the
+    optional __viewkey__ method.  By default, the selected view class is the
+    most specific subclass of C whose __view__ mentions the class of x.
+    If no such subclass is found, the search proceeds with the parent
+    object classes.  For example, C(True) will first look for a subclass
+    of C with __view__ = (..., bool, ...) and only if it doesn't find any
+    look for one with __view__ = (..., int, ...), and then ..., object,...
+    If everything fails the class C itself is considered to be the default.
+
+    Alternatively, the view class selection can be driven by another aspect
+    of the object x, instead of the class of x, by overriding __viewkey__.
+    See last example at the end of this module.
+    """
+
+    _viewcache = {}
+    __view__ = ()
+
+    def __new__(rootclass, obj, *args, **kwds):
+        self = object.__new__(rootclass)
+        self.__obj__ = obj
+        self.__rootclass__ = rootclass
+        key = self.__viewkey__()
+        try:
+            self.__class__ = self._viewcache[key]
+        except KeyError:
+            self.__class__ = self._selectsubclass(key)
+        return self
+
+    def __getattr__(self, attr):
+        # attributes not found in the normal hierarchy rooted on View
+        # are looked up in the object's real class
+        return getattr(self.__obj__, attr)
+
+    def __viewkey__(self):
+        return self.__obj__.__class__
+
+    def __matchkey__(self, key, subclasses):
+        if inspect.isclass(key):
+            keys = inspect.getmro(key)
+        else:
+            keys = [key]
+        for key in keys:
+            result = [C for C in subclasses if key in C.__view__]
+            if result:
+                return result
+        return []
+
+    def _selectsubclass(self, key):
+        subclasses = list(enumsubclasses(self.__rootclass__))
+        for C in subclasses:
+            if not isinstance(C.__view__, tuple):
+                C.__view__ = (C.__view__,)
+        choices = self.__matchkey__(key, subclasses)
+        if not choices:
+            return self.__rootclass__
+        elif len(choices) == 1:
+            return choices[0]
+        else:
+            # combine the multiple choices
+            return type('?', tuple(choices), {})
+
+    def __repr__(self):
+        return '%s(%r)' % (self.__rootclass__.__name__, self.__obj__)
+
+
+def enumsubclasses(cls):
+    for subcls in cls.__subclasses__():
+        for subsubclass in enumsubclasses(subcls):
+            yield subsubclass
+    yield cls
+
+
+class Interpretable(View):
+    """A parse tree node with a few extra methods."""
+    explanation = None
+
+    def is_builtin(self, frame):
+        return False
+
+    def eval(self, frame):
+        # fall-back for unknown expression nodes
+        try:
+            expr = ast.Expression(self.__obj__)
+            expr.filename = '<eval>'
+            self.__obj__.filename = '<eval>'
+            co = pycodegen.ExpressionCodeGenerator(expr).getCode()
+            result = frame.eval(co)
+        except passthroughex:
+            raise
+        except:
+            raise Failure(self)
+        self.result = result
+        self.explanation = self.explanation or frame.repr(self.result)
+
+    def run(self, frame):
+        # fall-back for unknown statement nodes
+        try:
+            expr = ast.Module(None, ast.Stmt([self.__obj__]))
+            expr.filename = '<run>'
+            co = pycodegen.ModuleCodeGenerator(expr).getCode()
+            frame.exec_(co)
+        except passthroughex:
+            raise
+        except:
+            raise Failure(self)
+
+    def nice_explanation(self):
+        return format_explanation(self.explanation)
+
+
+class Name(Interpretable):
+    __view__ = ast.Name
+
+    def is_local(self, frame):
+        source = '%r in locals() is not globals()' % self.name
+        try:
+            return frame.is_true(frame.eval(source))
+        except passthroughex:
+            raise
+        except:
+            return False
+
+    def is_global(self, frame):
+        source = '%r in globals()' % self.name
+        try:
+            return frame.is_true(frame.eval(source))
+        except passthroughex:
+            raise
+        except:
+            return False
+
+    def is_builtin(self, frame):
+        source = '%r not in locals() and %r not in globals()' % (
+            self.name, self.name)
+        try:
+            return frame.is_true(frame.eval(source))
+        except passthroughex:
+            raise
+        except:
+            return False
+
+    def eval(self, frame):
+        super(Name, self).eval(frame)
+        if not self.is_local(frame):
+            self.explanation = self.name
+
+class Compare(Interpretable):
+    __view__ = ast.Compare
+
+    def eval(self, frame):
+        expr = Interpretable(self.expr)
+        expr.eval(frame)
+        for operation, expr2 in self.ops:
+            if hasattr(self, 'result'):
+                # shortcutting in chained expressions
+                if not frame.is_true(self.result):
+                    break
+            expr2 = Interpretable(expr2)
+            expr2.eval(frame)
+            self.explanation = "%s %s %s" % (
+                expr.explanation, operation, expr2.explanation)
+            source = "__exprinfo_left %s __exprinfo_right" % operation
+            try:
+                self.result = frame.eval(source,
+                                         __exprinfo_left=expr.result,
+                                         __exprinfo_right=expr2.result)
+            except passthroughex:
+                raise
+            except:
+                raise Failure(self)
+            expr = expr2
+
+class And(Interpretable):
+    __view__ = ast.And
+
+    def eval(self, frame):
+        explanations = []
+        for expr in self.nodes:
+            expr = Interpretable(expr)
+            expr.eval(frame)
+            explanations.append(expr.explanation)
+            self.result = expr.result
+            if not frame.is_true(expr.result):
+                break
+        self.explanation = '(' + ' and '.join(explanations) + ')'
+
+class Or(Interpretable):
+    __view__ = ast.Or
+
+    def eval(self, frame):
+        explanations = []
+        for expr in self.nodes:
+            expr = Interpretable(expr)
+            expr.eval(frame)
+            explanations.append(expr.explanation)
+            self.result = expr.result
+            if frame.is_true(expr.result):
+                break
+        self.explanation = '(' + ' or '.join(explanations) + ')'
+
+
+# == Unary operations ==
+keepalive = []
+for astclass, astpattern in {
+    ast.Not    : 'not __exprinfo_expr',
+    ast.Invert : '(~__exprinfo_expr)',
+    }.items():
+
+    class UnaryArith(Interpretable):
+        __view__ = astclass
+
+        def eval(self, frame, astpattern=astpattern):
+            expr = Interpretable(self.expr)
+            expr.eval(frame)
+            self.explanation = astpattern.replace('__exprinfo_expr',
+                                                  expr.explanation)
+            try:
+                self.result = frame.eval(astpattern,
+                                         __exprinfo_expr=expr.result)
+            except passthroughex:
+                raise
+            except:
+                raise Failure(self)
+
+    keepalive.append(UnaryArith)
+
+# == Binary operations ==
+for astclass, astpattern in {
+    ast.Add    : '(__exprinfo_left + __exprinfo_right)',
+    ast.Sub    : '(__exprinfo_left - __exprinfo_right)',
+    ast.Mul    : '(__exprinfo_left * __exprinfo_right)',
+    ast.Div    : '(__exprinfo_left / __exprinfo_right)',
+    ast.Mod    : '(__exprinfo_left % __exprinfo_right)',
+    ast.Power  : '(__exprinfo_left ** __exprinfo_right)',
+    }.items():
+
+    class BinaryArith(Interpretable):
+        __view__ = astclass
+
+        def eval(self, frame, astpattern=astpattern):
+            left = Interpretable(self.left)
+            left.eval(frame)
+            right = Interpretable(self.right)
+            right.eval(frame)
+            self.explanation = (astpattern
+                                .replace('__exprinfo_left',  left .explanation)
+                                .replace('__exprinfo_right', right.explanation))
+            try:
+                self.result = frame.eval(astpattern,
+                                         __exprinfo_left=left.result,
+                                         __exprinfo_right=right.result)
+            except passthroughex:
+                raise
+            except:
+                raise Failure(self)
+
+    keepalive.append(BinaryArith)
+
+
+class CallFunc(Interpretable):
+    __view__ = ast.CallFunc
+
+    def is_bool(self, frame):
+        source = 'isinstance(__exprinfo_value, bool)'
+        try:
+            return frame.is_true(frame.eval(source,
+                                            __exprinfo_value=self.result))
+        except passthroughex:
+            raise
+        except:
+            return False
+
+    def eval(self, frame):
+        node = Interpretable(self.node)
+        node.eval(frame)
+        explanations = []
+        vars = {'__exprinfo_fn': node.result}
+        source = '__exprinfo_fn('
+        for a in self.args:
+            if isinstance(a, ast.Keyword):
+                keyword = a.name
+                a = a.expr
+            else:
+                keyword = None
+            a = Interpretable(a)
+            a.eval(frame)
+            argname = '__exprinfo_%d' % len(vars)
+            vars[argname] = a.result
+            if keyword is None:
+                source += argname + ','
+                explanations.append(a.explanation)
+            else:
+                source += '%s=%s,' % (keyword, argname)
+                explanations.append('%s=%s' % (keyword, a.explanation))
+        if self.star_args:
+            star_args = Interpretable(self.star_args)
+            star_args.eval(frame)
+            argname = '__exprinfo_star'
+            vars[argname] = star_args.result
+            source += '*' + argname + ','
+            explanations.append('*' + star_args.explanation)
+        if self.dstar_args:
+            dstar_args = Interpretable(self.dstar_args)
+            dstar_args.eval(frame)
+            argname = '__exprinfo_kwds'
+            vars[argname] = dstar_args.result
+            source += '**' + argname + ','
+            explanations.append('**' + dstar_args.explanation)
+        self.explanation = "%s(%s)" % (
+            node.explanation, ', '.join(explanations))
+        if source.endswith(','):
+            source = source[:-1]
+        source += ')'
+        try:
+            self.result = frame.eval(source, **vars)
+        except passthroughex:
+            raise
+        except:
+            raise Failure(self)
+        if not node.is_builtin(frame) or not self.is_bool(frame):
+            r = frame.repr(self.result)
+            self.explanation = '%s\n{%s = %s\n}' % (r, r, self.explanation)
+
+class Getattr(Interpretable):
+    __view__ = ast.Getattr
+
+    def eval(self, frame):
+        expr = Interpretable(self.expr)
+        expr.eval(frame)
+        source = '__exprinfo_expr.%s' % self.attrname
+        try:
+            self.result = frame.eval(source, __exprinfo_expr=expr.result)
+        except passthroughex:
+            raise
+        except:
+            raise Failure(self)
+        self.explanation = '%s.%s' % (expr.explanation, self.attrname)
+        # if the attribute comes from the instance, its value is interesting
+        source = ('hasattr(__exprinfo_expr, "__dict__") and '
+                  '%r in __exprinfo_expr.__dict__' % self.attrname)
+        try:
+            from_instance = frame.is_true(
+                frame.eval(source, __exprinfo_expr=expr.result))
+        except passthroughex:
+            raise
+        except:
+            from_instance = True
+        if from_instance:
+            r = frame.repr(self.result)
+            self.explanation = '%s\n{%s = %s\n}' % (r, r, self.explanation)
+
+# == Re-interpretation of full statements ==
+
+class Assert(Interpretable):
+    __view__ = ast.Assert
+
+    def run(self, frame):
+        test = Interpretable(self.test)
+        test.eval(frame)
+        # print the result as  'assert <explanation>'
+        self.result = test.result
+        self.explanation = 'assert ' + test.explanation
+        if not frame.is_true(test.result):
+            try:
+                raise BuiltinAssertionError
+            except passthroughex:
+                raise
+            except:
+                raise Failure(self)
+
+class Assign(Interpretable):
+    __view__ = ast.Assign
+
+    def run(self, frame):
+        expr = Interpretable(self.expr)
+        expr.eval(frame)
+        self.result = expr.result
+        self.explanation = '... = ' + expr.explanation
+        # fall-back-run the rest of the assignment
+        ass = ast.Assign(self.nodes, ast.Name('__exprinfo_expr'))
+        mod = ast.Module(None, ast.Stmt([ass]))
+        mod.filename = '<run>'
+        co = pycodegen.ModuleCodeGenerator(mod).getCode()
+        try:
+            frame.exec_(co, __exprinfo_expr=expr.result)
+        except passthroughex:
+            raise
+        except:
+            raise Failure(self)
+
+class Discard(Interpretable):
+    __view__ = ast.Discard
+
+    def run(self, frame):
+        expr = Interpretable(self.expr)
+        expr.eval(frame)
+        self.result = expr.result
+        self.explanation = expr.explanation
+
+class Stmt(Interpretable):
+    __view__ = ast.Stmt
+
+    def run(self, frame):
+        for stmt in self.nodes:
+            stmt = Interpretable(stmt)
+            stmt.run(frame)
+
+
+def report_failure(e):
+    explanation = e.node.nice_explanation()
+    if explanation:
+        explanation = ", in: " + explanation
+    else:
+        explanation = ""
+    sys.stdout.write("%s: %s%s\n" % (e.exc.__name__, e.value, explanation))
+
+def check(s, frame=None):
+    if frame is None:
+        frame = sys._getframe(1)
+        frame = py.code.Frame(frame)
+    expr = parse(s, 'eval')
+    assert isinstance(expr, ast.Expression)
+    node = Interpretable(expr.node)
+    try:
+        node.eval(frame)
+    except passthroughex:
+        raise
+    except Failure:
+        e = sys.exc_info()[1]
+        report_failure(e)
+    else:
+        if not frame.is_true(node.result):
+            sys.stderr.write("assertion failed: %s\n" % node.nice_explanation())
+
+
+###########################################################
+# API / Entry points
+# #########################################################
+
+def interpret(source, frame, should_fail=False):
+    module = Interpretable(parse(source, 'exec').node)
+    #print "got module", module
+    if isinstance(frame, py.std.types.FrameType):
+        frame = py.code.Frame(frame)
+    try:
+        module.run(frame)
+    except Failure:
+        e = sys.exc_info()[1]
+        return getfailure(e)
+    except passthroughex:
+        raise
+    except:
+        import traceback
+        traceback.print_exc()
+    if should_fail:
+        return ("(assertion failed, but when it was re-run for "
+                "printing intermediate values, it did not fail.  Suggestions: "
+                "compute assert expression before the assert or use --assert=plain)")
+    else:
+        return None
+
+def getmsg(excinfo):
+    if isinstance(excinfo, tuple):
+        excinfo = py.code.ExceptionInfo(excinfo)
+    #frame, line = gettbline(tb)
+    #frame = py.code.Frame(frame)
+    #return interpret(line, frame)
+
+    tb = excinfo.traceback[-1]
+    source = str(tb.statement).strip()
+    x = interpret(source, tb.frame, should_fail=True)
+    if not isinstance(x, str):
+        raise TypeError("interpret returned non-string %r" % (x,))
+    return x
+
+def getfailure(e):
+    explanation = e.node.nice_explanation()
+    if str(e.value):
+        lines = explanation.split('\n')
+        lines[0] += "  << %s" % (e.value,)
+        explanation = '\n'.join(lines)
+    text = "%s: %s" % (e.exc.__name__, explanation)
+    if text.startswith('AssertionError: assert '):
+        text = text[16:]
+    return text
+
+def run(s, frame=None):
+    if frame is None:
+        frame = sys._getframe(1)
+        frame = py.code.Frame(frame)
+    module = Interpretable(parse(s, 'exec').node)
+    try:
+        module.run(frame)
+    except Failure:
+        e = sys.exc_info()[1]
+        report_failure(e)
+
+
+if __name__ == '__main__':
+    # example:
+    def f():
+        return 5
+    def g():
+        return 3
+    def h(x):
+        return 'never'
+    check("f() * g() == 5")
+    check("not f()")
+    check("not (f() and g() or 0)")
+    check("f() == g()")
+    i = 4
+    check("i == f()")
+    check("len(f()) == 0")
+    check("isinstance(2+3+4, float)")
+
+    run("x = i")
+    check("x == 5")
+
+    run("assert not f(), 'oops'")
+    run("a, b, c = 1, 2")
+    run("a, b, c = f()")
+
+    check("max([f(),g()]) == 4")
+    check("'hello'[g()] == 'h'")
+    run("'guk%d' % h(f())")

_pytest/assertion/reinterpret.py

@@ -0,0 +1,48 @@
+import sys
+import py
+
+BuiltinAssertionError = py.builtin.builtins.AssertionError
+
+class AssertionError(BuiltinAssertionError):
+    def __init__(self, *args):
+        BuiltinAssertionError.__init__(self, *args)
+        if args:
+            try:
+                self.msg = str(args[0])
+            except py.builtin._sysex:
+                raise
+            except:
+                self.msg = "<[broken __repr__] %s at %0xd>" %(
+                    args[0].__class__, id(args[0]))
+        else:
+            f = py.code.Frame(sys._getframe(1))
+            try:
+                source = f.code.fullsource
+                if source is not None:
+                    try:
+                        source = source.getstatement(f.lineno, assertion=True)
+                    except IndexError:
+                        source = None
+                    else:
+                        source = str(source.deindent()).strip()
+            except py.error.ENOENT:
+                source = None
+                # this can also occur during reinterpretation, when the
+                # co_filename is set to "<run>".
+            if source:
+                self.msg = reinterpret(source, f, should_fail=True)
+            else:
+                self.msg = "<could not determine information>"
+            if not self.args:
+                self.args = (self.msg,)
+
+if sys.version_info > (3, 0):
+    AssertionError.__module__ = "builtins"
+    reinterpret_old = "old reinterpretation not available for py3"
+else:
+    from _pytest.assertion.oldinterpret import interpret as reinterpret_old
+if sys.version_info >= (2, 6) or (sys.platform.startswith("java")):
+    from _pytest.assertion.newinterpret import interpret as reinterpret
+else:
+    reinterpret = reinterpret_old
+

_pytest/assertion/rewrite.py

@@ -0,0 +1,598 @@
+"""Rewrite assertion AST to produce nice error messages"""
+
+import ast
+import collections
+import errno
+import itertools
+import imp
+import marshal
+import os
+import struct
+import sys
+import types
+
+import py
+from _pytest.assertion import util
+
+
+# Windows gives ENOENT in places *nix gives ENOTDIR.
+if sys.platform.startswith("win"):
+    PATH_COMPONENT_NOT_DIR = errno.ENOENT
+else:
+    PATH_COMPONENT_NOT_DIR = errno.ENOTDIR
+
+# py.test caches rewritten pycs in __pycache__.
+if hasattr(imp, "get_tag"):
+    PYTEST_TAG = imp.get_tag() + "-PYTEST"
+else:
+    if hasattr(sys, "pypy_version_info"):
+        impl = "pypy"
+    elif sys.platform == "java":
+        impl = "jython"
+    else:
+        impl = "cpython"
+    ver = sys.version_info
+    PYTEST_TAG = "%s-%s%s-PYTEST" % (impl, ver[0], ver[1])
+    del ver, impl
+
+PYC_EXT = ".py" + "c" if __debug__ else "o"
+PYC_TAIL = "." + PYTEST_TAG + PYC_EXT
+
+REWRITE_NEWLINES = sys.version_info[:2] != (2, 7) and sys.version_info < (3, 2)
+
+class AssertionRewritingHook(object):
+    """Import hook which rewrites asserts."""
+
+    def __init__(self):
+        self.session = None
+        self.modules = {}
+
+    def set_session(self, session):
+        self.fnpats = session.config.getini("python_files")
+        self.session = session
+
+    def find_module(self, name, path=None):
+        if self.session is None:
+            return None
+        sess = self.session
+        state = sess.config._assertstate
+        state.trace("find_module called for: %s" % name)
+        names = name.rsplit(".", 1)
+        lastname = names[-1]
+        pth = None
+        if path is not None and len(path) == 1:
+            pth = path[0]
+        if pth is None:
+            try:
+                fd, fn, desc = imp.find_module(lastname, path)
+            except ImportError:
+                return None
+            if fd is not None:
+                fd.close()
+            tp = desc[2]
+            if tp == imp.PY_COMPILED:
+                if hasattr(imp, "source_from_cache"):
+                    fn = imp.source_from_cache(fn)
+                else:
+                    fn = fn[:-1]
+            elif tp != imp.PY_SOURCE:
+                # Don't know what this is.
+                return None
+        else:
+            fn = os.path.join(pth, name.rpartition(".")[2] + ".py")
+        fn_pypath = py.path.local(fn)
+        # Is this a test file?
+        if not sess.isinitpath(fn):
+            # We have to be very careful here because imports in this code can
+            # trigger a cycle.
+            self.session = None
+            try:
+                for pat in self.fnpats:
+                    if fn_pypath.fnmatch(pat):
+                        state.trace("matched test file %r" % (fn,))
+                        break
+                else:
+                    return None
+            finally:
+                self.session = sess
+        else:
+            state.trace("matched test file (was specified on cmdline): %r" % (fn,))
+        # The requested module looks like a test file, so rewrite it. This is
+        # the most magical part of the process: load the source, rewrite the
+        # asserts, and load the rewritten source. We also cache the rewritten
+        # module code in a special pyc. We must be aware of the possibility of
+        # concurrent py.test processes rewriting and loading pycs. To avoid
+        # tricky race conditions, we maintain the following invariant: The
+        # cached pyc is always a complete, valid pyc. Operations on it must be
+        # atomic. POSIX's atomic rename comes in handy.
+        write = not sys.dont_write_bytecode
+        cache_dir = os.path.join(fn_pypath.dirname, "__pycache__")
+        if write:
+            try:
+                os.mkdir(cache_dir)
+            except OSError:
+                e = sys.exc_info()[1].errno
+                if e == errno.EEXIST:
+                    # Either the __pycache__ directory already exists (the
+                    # common case) or it's blocked by a non-dir node. In the
+                    # latter case, we'll ignore it in _write_pyc.
+                    pass
+                elif e == PATH_COMPONENT_NOT_DIR:
+                    # One of the path components was not a directory, likely
+                    # because we're in a zip file.
+                    write = False
+                elif e == errno.EACCES:
+                    state.trace("read only directory: %r" % (fn_pypath.dirname,))
+                    write = False
+                else:
+                    raise
+        cache_name = fn_pypath.basename[:-3] + PYC_TAIL
+        pyc = os.path.join(cache_dir, cache_name)
+        # Notice that even if we're in a read-only directory, I'm going to check
+        # for a cached pyc. This may not be optimal...
+        co = _read_pyc(fn_pypath, pyc)
+        if co is None:
+            state.trace("rewriting %r" % (fn,))
+            co = _rewrite_test(state, fn_pypath)
+            if co is None:
+                # Probably a SyntaxError in the test.
+                return None
+            if write:
+                _make_rewritten_pyc(state, fn_pypath, pyc, co)
+        else:
+            state.trace("found cached rewritten pyc for %r" % (fn,))
+        self.modules[name] = co, pyc
+        return self
+
+    def load_module(self, name):
+        co, pyc = self.modules.pop(name)
+        # I wish I could just call imp.load_compiled here, but __file__ has to
+        # be set properly. In Python 3.2+, this all would be handled correctly
+        # by load_compiled.
+        mod = sys.modules[name] = imp.new_module(name)
+        try:
+            mod.__file__ = co.co_filename
+            # Normally, this attribute is 3.2+.
+            mod.__cached__ = pyc
+            py.builtin.exec_(co, mod.__dict__)
+        except:
+            del sys.modules[name]
+            raise
+        return sys.modules[name]
+
+def _write_pyc(co, source_path, pyc):
+    # Technically, we don't have to have the same pyc format as (C)Python, since
+    # these "pycs" should never be seen by builtin import. However, there's
+    # little reason deviate, and I hope sometime to be able to use
+    # imp.load_compiled to load them. (See the comment in load_module above.)
+    mtime = int(source_path.mtime())
+    try:
+        fp = open(pyc, "wb")
+    except IOError:
+        err = sys.exc_info()[1].errno
+        if err == PATH_COMPONENT_NOT_DIR:
+            # This happens when we get a EEXIST in find_module creating the
+            # __pycache__ directory and __pycache__ is by some non-dir node.
+            return False
+        raise
+    try:
+        fp.write(imp.get_magic())
+        fp.write(struct.pack("<l", mtime))
+        marshal.dump(co, fp)
+    finally:
+        fp.close()
+    return True
+
+RN = "\r\n".encode("utf-8")
+N = "\n".encode("utf-8")
+
+def _rewrite_test(state, fn):
+    """Try to read and rewrite *fn* and return the code object."""
+    try:
+        source = fn.read("rb")
+    except EnvironmentError:
+        return None
+    # On Python versions which are not 2.7 and less than or equal to 3.1, the
+    # parser expects *nix newlines.
+    if REWRITE_NEWLINES:
+        source = source.replace(RN, N) + N
+    try:
+        tree = ast.parse(source)
+    except SyntaxError:
+        # Let this pop up again in the real import.
+        state.trace("failed to parse: %r" % (fn,))
+        return None
+    rewrite_asserts(tree)
+    try:
+        co = compile(tree, fn.strpath, "exec")
+    except SyntaxError:
+        # It's possible that this error is from some bug in the
+        # assertion rewriting, but I don't know of a fast way to tell.
+        state.trace("failed to compile: %r" % (fn,))
+        return None
+    return co
+
+def _make_rewritten_pyc(state, fn, pyc, co):
+    """Try to dump rewritten code to *pyc*."""
+    if sys.platform.startswith("win"):
+        # Windows grants exclusive access to open files and doesn't have atomic
+        # rename, so just write into the final file.
+        _write_pyc(co, fn, pyc)
+    else:
+        # When not on windows, assume rename is atomic. Dump the code object
+        # into a file specific to this process and atomically replace it.
+        proc_pyc = pyc + "." + str(os.getpid())
+        if _write_pyc(co, fn, proc_pyc):
+            os.rename(proc_pyc, pyc)
+
+def _read_pyc(source, pyc):
+    """Possibly read a py.test pyc containing rewritten code.
+
+    Return rewritten code if successful or None if not.
+    """
+    try:
+        fp = open(pyc, "rb")
+    except IOError:
+        return None
+    try:
+        try:
+            mtime = int(source.mtime())
+            data = fp.read(8)
+        except EnvironmentError:
+            return None
+        # Check for invalid or out of date pyc file.
+        if (len(data) != 8 or
+            data[:4] != imp.get_magic() or
+            struct.unpack("<l", data[4:])[0] != mtime):
+            return None
+        co = marshal.load(fp)
+        if not isinstance(co, types.CodeType):
+            # That's interesting....
+            return None
+        return co
+    finally:
+        fp.close()
+
+
+def rewrite_asserts(mod):
+    """Rewrite the assert statements in mod."""
+    AssertionRewriter().run(mod)
+
+
+_saferepr = py.io.saferepr
+from _pytest.assertion.util import format_explanation as _format_explanation
+
+def _format_boolop(explanations, is_or):
+    return "(" + (is_or and " or " or " and ").join(explanations) + ")"
+
+def _call_reprcompare(ops, results, expls, each_obj):
+    for i, res, expl in zip(range(len(ops)), results, expls):
+        try:
+            done = not res
+        except Exception:
+            done = True
+        if done:
+            break
+    if util._reprcompare is not None:
+        custom = util._reprcompare(ops[i], each_obj[i], each_obj[i + 1])
+        if custom is not None:
+            return custom
+    return expl
+
+
+unary_map = {
+    ast.Not : "not %s",
+    ast.Invert : "~%s",
+    ast.USub : "-%s",
+    ast.UAdd : "+%s"
+}
+
+binop_map = {
+    ast.BitOr : "|",
+    ast.BitXor : "^",
+    ast.BitAnd : "&",
+    ast.LShift : "<<",
+    ast.RShift : ">>",
+    ast.Add : "+",
+    ast.Sub : "-",
+    ast.Mult : "*",
+    ast.Div : "/",
+    ast.FloorDiv : "//",
+    ast.Mod : "%",
+    ast.Eq : "==",
+    ast.NotEq : "!=",
+    ast.Lt : "<",
+    ast.LtE : "<=",
+    ast.Gt : ">",
+    ast.GtE : ">=",
+    ast.Pow : "**",
+    ast.Is : "is",
+    ast.IsNot : "is not",
+    ast.In : "in",
+    ast.NotIn : "not in"
+}
+
+
+def set_location(node, lineno, col_offset):
+    """Set node location information recursively."""
+    def _fix(node, lineno, col_offset):
+        if "lineno" in node._attributes:
+            node.lineno = lineno
+        if "col_offset" in node._attributes:
+            node.col_offset = col_offset
+        for child in ast.iter_child_nodes(node):
+            _fix(child, lineno, col_offset)
+    _fix(node, lineno, col_offset)
+    return node
+
+
+class AssertionRewriter(ast.NodeVisitor):
+
+    def run(self, mod):
+        """Find all assert statements in *mod* and rewrite them."""
+        if not mod.body:
+            # Nothing to do.
+            return
+        # Insert some special imports at the top of the module but after any
+        # docstrings and __future__ imports.
+        aliases = [ast.alias(py.builtin.builtins.__name__, "@py_builtins"),
+                   ast.alias("_pytest.assertion.rewrite", "@pytest_ar")]
+        expect_docstring = True
+        pos = 0
+        lineno = 0
+        for item in mod.body:
+            if (expect_docstring and isinstance(item, ast.Expr) and
+                isinstance(item.value, ast.Str)):
+                doc = item.value.s
+                if "PYTEST_DONT_REWRITE" in doc:
+                    # The module has disabled assertion rewriting.
+                    return
+                lineno += len(doc) - 1
+                expect_docstring = False
+            elif (not isinstance(item, ast.ImportFrom) or item.level > 0 or
+                  item.module != "__future__"):
+                lineno = item.lineno
+                break
+            pos += 1
+        imports = [ast.Import([alias], lineno=lineno, col_offset=0)
+                   for alias in aliases]
+        mod.body[pos:pos] = imports
+        # Collect asserts.
+        nodes = [mod]
+        while nodes:
+            node = nodes.pop()
+            for name, field in ast.iter_fields(node):
+                if isinstance(field, list):
+                    new = []
+                    for i, child in enumerate(field):
+                        if isinstance(child, ast.Assert):
+                            # Transform assert.
+                            new.extend(self.visit(child))
+                        else:
+                            new.append(child)
+                            if isinstance(child, ast.AST):
+                                nodes.append(child)
+                    setattr(node, name, new)
+                elif (isinstance(field, ast.AST) and
+                      # Don't recurse into expressions as they can't contain
+                      # asserts.
+                      not isinstance(field, ast.expr)):
+                    nodes.append(field)
+
+    def variable(self):
+        """Get a new variable."""
+        # Use a character invalid in python identifiers to avoid clashing.
+        name = "@py_assert" + str(next(self.variable_counter))
+        self.variables.append(name)
+        return name
+
+    def assign(self, expr):
+        """Give *expr* a name."""
+        name = self.variable()
+        self.statements.append(ast.Assign([ast.Name(name, ast.Store())], expr))
+        return ast.Name(name, ast.Load())
+
+    def display(self, expr):
+        """Call py.io.saferepr on the expression."""
+        return self.helper("saferepr", expr)
+
+    def helper(self, name, *args):
+        """Call a helper in this module."""
+        py_name = ast.Name("@pytest_ar", ast.Load())
+        attr = ast.Attribute(py_name, "_" + name, ast.Load())
+        return ast.Call(attr, list(args), [], None, None)
+
+    def builtin(self, name):
+        """Return the builtin called *name*."""
+        builtin_name = ast.Name("@py_builtins", ast.Load())
+        return ast.Attribute(builtin_name, name, ast.Load())
+
+    def explanation_param(self, expr):
+        specifier = "py" + str(next(self.variable_counter))
+        self.explanation_specifiers[specifier] = expr
+        return "%(" + specifier + ")s"
+
+    def push_format_context(self):
+        self.explanation_specifiers = {}
+        self.stack.append(self.explanation_specifiers)
+
+    def pop_format_context(self, expl_expr):
+        current = self.stack.pop()
+        if self.stack:
+            self.explanation_specifiers = self.stack[-1]
+        keys = [ast.Str(key) for key in current.keys()]
+        format_dict = ast.Dict(keys, list(current.values()))
+        form = ast.BinOp(expl_expr, ast.Mod(), format_dict)
+        name = "@py_format" + str(next(self.variable_counter))
+        self.on_failure.append(ast.Assign([ast.Name(name, ast.Store())], form))
+        return ast.Name(name, ast.Load())
+
+    def generic_visit(self, node):
+        """Handle expressions we don't have custom code for."""
+        assert isinstance(node, ast.expr)
+        res = self.assign(node)
+        return res, self.explanation_param(self.display(res))
+
+    def visit_Assert(self, assert_):
+        if assert_.msg:
+            # There's already a message. Don't mess with it.
+            return [assert_]
+        self.statements = []
+        self.cond_chain = ()
+        self.variables = []
+        self.variable_counter = itertools.count()
+        self.stack = []
+        self.on_failure = []
+        self.push_format_context()
+        # Rewrite assert into a bunch of statements.
+        top_condition, explanation = self.visit(assert_.test)
+        # Create failure message.
+        body = self.on_failure
+        negation = ast.UnaryOp(ast.Not(), top_condition)
+        self.statements.append(ast.If(negation, body, []))
+        explanation = "assert " + explanation
+        template = ast.Str(explanation)
+        msg = self.pop_format_context(template)
+        fmt = self.helper("format_explanation", msg)
+        err_name = ast.Name("AssertionError", ast.Load())
+        exc = ast.Call(err_name, [fmt], [], None, None)
+        if sys.version_info[0] >= 3:
+            raise_ = ast.Raise(exc, None)
+        else:
+            raise_ = ast.Raise(exc, None, None)
+        body.append(raise_)
+        # Clear temporary variables by setting them to None.
+        if self.variables:
+            variables = [ast.Name(name, ast.Store()) for name in self.variables]
+            clear = ast.Assign(variables, ast.Name("None", ast.Load()))
+            self.statements.append(clear)
+        # Fix line numbers.
+        for stmt in self.statements:
+            set_location(stmt, assert_.lineno, assert_.col_offset)
+        return self.statements
+
+    def visit_Name(self, name):
+        # Check if the name is local or not.
+        locs = ast.Call(self.builtin("locals"), [], [], None, None)
+        globs = ast.Call(self.builtin("globals"), [], [], None, None)
+        ops = [ast.In(), ast.IsNot()]
+        test = ast.Compare(ast.Str(name.id), ops, [locs, globs])
+        expr = ast.IfExp(test, self.display(name), ast.Str(name.id))
+        return name, self.explanation_param(expr)
+
+    def visit_BoolOp(self, boolop):
+        res_var = self.variable()
+        expl_list = self.assign(ast.List([], ast.Load()))
+        app = ast.Attribute(expl_list, "append", ast.Load())
+        is_or = int(isinstance(boolop.op, ast.Or))
+        body = save = self.statements
+        fail_save = self.on_failure
+        levels = len(boolop.values) - 1
+        self.push_format_context()
+        # Process each operand, short-circuting if needed.
+        for i, v in enumerate(boolop.values):
+            if i:
+                fail_inner = []
+                self.on_failure.append(ast.If(cond, fail_inner, []))
+                self.on_failure = fail_inner
+            self.push_format_context()
+            res, expl = self.visit(v)
+            body.append(ast.Assign([ast.Name(res_var, ast.Store())], res))
+            expl_format = self.pop_format_context(ast.Str(expl))
+            call = ast.Call(app, [expl_format], [], None, None)
+            self.on_failure.append(ast.Expr(call))
+            if i < levels:
+                cond = res
+                if is_or:
+                    cond = ast.UnaryOp(ast.Not(), cond)
+                inner = []
+                self.statements.append(ast.If(cond, inner, []))
+                self.statements = body = inner
+        self.statements = save
+        self.on_failure = fail_save
+        expl_template = self.helper("format_boolop", expl_list, ast.Num(is_or))
+        expl = self.pop_format_context(expl_template)
+        return ast.Name(res_var, ast.Load()), self.explanation_param(expl)
+
+    def visit_UnaryOp(self, unary):
+        pattern = unary_map[unary.op.__class__]
+        operand_res, operand_expl = self.visit(unary.operand)
+        res = self.assign(ast.UnaryOp(unary.op, operand_res))
+        return res, pattern % (operand_expl,)
+
+    def visit_BinOp(self, binop):
+        symbol = binop_map[binop.op.__class__]
+        left_expr, left_expl = self.visit(binop.left)
+        right_expr, right_expl = self.visit(binop.right)
+        explanation = "(%s %s %s)" % (left_expl, symbol, right_expl)
+        res = self.assign(ast.BinOp(left_expr, binop.op, right_expr))
+        return res, explanation
+
+    def visit_Call(self, call):
+        new_func, func_expl = self.visit(call.func)
+        arg_expls = []
+        new_args = []
+        new_kwargs = []
+        new_star = new_kwarg = None
+        for arg in call.args:
+            res, expl = self.visit(arg)
+            new_args.append(res)
+            arg_expls.append(expl)
+        for keyword in call.keywords:
+            res, expl = self.visit(keyword.value)
+            new_kwargs.append(ast.keyword(keyword.arg, res))
+            arg_expls.append(keyword.arg + "=" + expl)
+        if call.starargs:
+            new_star, expl = self.visit(call.starargs)
+            arg_expls.append("*" + expl)
+        if call.kwargs:
+            new_kwarg, expl = self.visit(call.kwargs)
+            arg_expls.append("**" + expl)
+        expl = "%s(%s)" % (func_expl, ', '.join(arg_expls))
+        new_call = ast.Call(new_func, new_args, new_kwargs, new_star, new_kwarg)
+        res = self.assign(new_call)
+        res_expl = self.explanation_param(self.display(res))
+        outer_expl = "%s\n{%s = %s\n}" % (res_expl, res_expl, expl)
+        return res, outer_expl
+
+    def visit_Attribute(self, attr):
+        if not isinstance(attr.ctx, ast.Load):
+            return self.generic_visit(attr)
+        value, value_expl = self.visit(attr.value)
+        res = self.assign(ast.Attribute(value, attr.attr, ast.Load()))
+        res_expl = self.explanation_param(self.display(res))
+        pat = "%s\n{%s = %s.%s\n}"
+        expl = pat % (res_expl, res_expl, value_expl, attr.attr)
+        return res, expl
+
+    def visit_Compare(self, comp):
+        self.push_format_context()
+        left_res, left_expl = self.visit(comp.left)
+        res_variables = [self.variable() for i in range(len(comp.ops))]
+        load_names = [ast.Name(v, ast.Load()) for v in res_variables]
+        store_names = [ast.Name(v, ast.Store()) for v in res_variables]
+        it = zip(range(len(comp.ops)), comp.ops, comp.comparators)
+        expls = []
+        syms = []
+        results = [left_res]
+        for i, op, next_operand in it:
+            next_res, next_expl = self.visit(next_operand)
+            results.append(next_res)
+            sym = binop_map[op.__class__]
+            syms.append(ast.Str(sym))
+            expl = "%s %s %s" % (left_expl, sym, next_expl)
+            expls.append(ast.Str(expl))
+            res_expr = ast.Compare(left_res, [op], [next_res])
+            self.statements.append(ast.Assign([store_names[i]], res_expr))
+            left_res, left_expl = next_res, next_expl
+        # Use py.code._reprcompare if that's available.
+        expl_call = self.helper("call_reprcompare",
+                                ast.Tuple(syms, ast.Load()),
+                                ast.Tuple(load_names, ast.Load()),
+                                ast.Tuple(expls, ast.Load()),
+                                ast.Tuple(results, ast.Load()))
+        if len(comp.ops) > 1:
+            res = ast.BoolOp(ast.And(), load_names)
+        else:
+            res = load_names[0]
+        return res, self.explanation_param(self.pop_format_context(expl_call))

_pytest/assertion/util.py

@@ -1,45 +1,78 @@
-"""
-support for presented detailed information in failing assertions.
-"""
+"""Utilities for assertion debugging"""
+
 import py
-import sys
-from _pytest.monkeypatch import monkeypatch
 
-def pytest_addoption(parser):
-    group = parser.getgroup("debugconfig")
-    group._addoption('--no-assert', action="store_true", default=False,
-        dest="noassert",
-        help="disable python assert expression reinterpretation."),
 
-def pytest_configure(config):
-    # The _pytesthook attribute on the AssertionError is used by
-    # py._code._assertionnew to detect this plugin was loaded and in
-    # turn call the hooks defined here as part of the
-    # DebugInterpreter.
-    config._monkeypatch = m = monkeypatch()
-    if not config.getvalue("noassert") and not config.getvalue("nomagic"):
-        warn_about_missing_assertion()
-        def callbinrepr(op, left, right):
-            hook_result = config.hook.pytest_assertrepr_compare(
-                config=config, op=op, left=left, right=right)
-            for new_expl in hook_result:
-                if new_expl:
-                    return '\n~'.join(new_expl)
-        m.setattr(py.builtin.builtins,
-                  'AssertionError', py.code._AssertionError)
-        m.setattr(py.code, '_reprcompare', callbinrepr)
+# The _reprcompare attribute on the util module is used by the new assertion
+# interpretation code and assertion rewriter to detect this plugin was
+# loaded and in turn call the hooks defined here as part of the
+# DebugInterpreter.
+_reprcompare = None
 
-def pytest_unconfigure(config):
-    config._monkeypatch.undo()
+def format_explanation(explanation):
+    """This formats an explanation
 
-def warn_about_missing_assertion():
-    try:
-        assert False
-    except AssertionError:
-        pass
-    else:
-        py.std.warnings.warn("Assertions are turned off!"
-                             " (are you using python -O?)")
+    Normally all embedded newlines are escaped, however there are
+    three exceptions: \n{, \n} and \n~.  The first two are intended
+    cover nested explanations, see function and attribute explanations
+    for examples (.visit_Call(), visit_Attribute()).  The last one is
+    for when one explanation needs to span multiple lines, e.g. when
+    displaying diffs.
+    """
+    # simplify 'assert False where False = ...'
+    where = 0
+    while True:
+        start = where = explanation.find("False\n{False = ", where)
+        if where == -1:
+            break
+        level = 0
+        for i, c in enumerate(explanation[start:]):
+            if c == "{":
+                level += 1
+            elif c == "}":
+                level -= 1
+                if not level:
+                    break
+        else:
+            raise AssertionError("unbalanced braces: %r" % (explanation,))
+        end = start + i
+        where = end
+        if explanation[end - 1] == '\n':
+            explanation = (explanation[:start] + explanation[start+15:end-1] +
+                           explanation[end+1:])
+            where -= 17
+    raw_lines = (explanation or '').split('\n')
+    # escape newlines not followed by {, } and ~
+    lines = [raw_lines[0]]
+    for l in raw_lines[1:]:
+        if l.startswith('{') or l.startswith('}') or l.startswith('~'):
+            lines.append(l)
+        else:
+            lines[-1] += '\\n' + l
+
+    result = lines[:1]
+    stack = [0]
+    stackcnt = [0]
+    for line in lines[1:]:
+        if line.startswith('{'):
+            if stackcnt[-1]:
+                s = 'and   '
+            else:
+                s = 'where '
+            stack.append(len(result))
+            stackcnt[-1] += 1
+            stackcnt.append(0)
+            result.append(' +' + '  '*(len(stack)-1) + s + line[1:])
+        elif line.startswith('}'):
+            assert line.startswith('}')
+            stack.pop()
+            stackcnt.pop()
+            result[stack[-1]] += line[1:]
+        else:
+            assert line.startswith('~')
+            result.append('  '*len(stack) + line[1:])
+    assert len(stack) == 1
+    return '\n'.join(result)
 
 
 # Provide basestring in python3
@@ -49,10 +82,11 @@ except NameError:
     basestring = str
 
 
-def pytest_assertrepr_compare(op, left, right):
+def assertrepr_compare(op, left, right):
     """return specialised explanations for some operators/operands"""
-    left_repr = py.io.saferepr(left, maxsize=30)
-    right_repr = py.io.saferepr(right, maxsize=30)
+    width = 80 - 15 - len(op) - 2 # 15 chars indentation, 1 space around op
+    left_repr = py.io.saferepr(left, maxsize=int(width/2))
+    right_repr = py.io.saferepr(right, maxsize=width-len(left_repr))
     summary = '%s %s %s' % (left_repr, op, right_repr)
 
     issequence = lambda x: isinstance(x, (list, tuple))
@@ -72,6 +106,9 @@ def pytest_assertrepr_compare(op, left, right):
             elif isdict(left) and isdict(right):
                 explanation = _diff_text(py.std.pprint.pformat(left),
                                          py.std.pprint.pformat(right))
+        elif op == 'not in':
+            if istext(left) and istext(right):
+                explanation = _notin_text(left, right)
     except py.builtin._sysex:
         raise
     except:
@@ -155,3 +192,22 @@ def _compare_eq_set(left, right):
         for item in diff_right:
             explanation.append(py.io.saferepr(item))
     return explanation
+
+
+def _notin_text(term, text):
+    index = text.find(term)
+    head = text[:index]
+    tail = text[index+len(term):]
+    correct_text = head + tail
+    diff = _diff_text(correct_text, text)
+    newdiff = ['%s is contained here:' % py.io.saferepr(term, maxsize=42)]
+    for line in diff:
+        if line.startswith('Skipping'):
+            continue
+        if line.startswith('- '):
+            continue
+        if line.startswith('+ '):
+            newdiff.append('  ' + line[2:])
+        else:
+            newdiff.append(line)
+    return newdiff

_pytest/capture.py

@@ -11,22 +11,22 @@ def pytest_addoption(parser):
     group._addoption('-s', action="store_const", const="no", dest="capture",
         help="shortcut for --capture=no.")
 
+@pytest.mark.tryfirst
+def pytest_cmdline_parse(pluginmanager, args):
+    # we want to perform capturing already for plugin/conftest loading
+    if '-s' in args or "--capture=no" in args:
+        method = "no"
+    elif hasattr(os, 'dup') and '--capture=sys' not in args:
+        method = "fd"
+    else:
+        method = "sys"
+    capman = CaptureManager(method)
+    pluginmanager.register(capman, "capturemanager")
+
 def addouterr(rep, outerr):
-    repr = getattr(rep, 'longrepr', None)
-    if not hasattr(repr, 'addsection'):
-        return
     for secname, content in zip(["out", "err"], outerr):
         if content:
-            repr.addsection("Captured std%s" % secname, content.rstrip())
-
-def pytest_configure(config):
-    config.pluginmanager.register(CaptureManager(), 'capturemanager')
-
-def pytest_unconfigure(config):
-    capman = config.pluginmanager.getplugin('capturemanager')
-    while capman._method2capture:
-        name, cap = capman._method2capture.popitem()
-        cap.reset()
+            rep.sections.append(("Captured std%s" % secname, content))
 
 class NoCapture:
     def startall(self):
@@ -39,8 +39,9 @@ class NoCapture:
         return "", ""
 
 class CaptureManager:
-    def __init__(self):
+    def __init__(self, defaultmethod=None):
         self._method2capture = {}
+        self._defaultmethod = defaultmethod
 
     def _maketempfile(self):
         f = py.std.tempfile.TemporaryFile()
@@ -77,16 +78,22 @@ class CaptureManager:
             method = "sys"
         return method
 
+    def reset_capturings(self):
+        for name, cap in self._method2capture.items():
+            cap.reset()
+
     def resumecapture_item(self, item):
         method = self._getmethod(item.config, item.fspath)
         if not hasattr(item, 'outerr'):
             item.outerr = ('', '') # we accumulate outerr on the item
         return self.resumecapture(method)
 
-    def resumecapture(self, method):
+    def resumecapture(self, method=None):
         if hasattr(self, '_capturing'):
             raise ValueError("cannot resume, already capturing with %r" %
                 (self._capturing,))
+        if method is None:
+            method = self._defaultmethod
         cap = self._method2capture.get(method)
         self._capturing = method
         if cap is None:
@@ -156,17 +163,6 @@ class CaptureManager:
     def pytest_runtest_teardown(self, item):
         self.resumecapture_item(item)
 
-    def pytest__teardown_final(self, __multicall__, session):
-        method = self._getmethod(session.config, None)
-        self.resumecapture(method)
-        try:
-            rep = __multicall__.execute()
-        finally:
-            outerr = self.suspendcapture()
-        if rep:
-            addouterr(rep, outerr)
-        return rep
-
     def pytest_keyboard_interrupt(self, excinfo):
         if hasattr(self, '_capturing'):
             self.suspendcapture()
@@ -184,18 +180,16 @@ class CaptureManager:
         return rep
 
 def pytest_funcarg__capsys(request):
-    """captures writes to sys.stdout/sys.stderr and makes
-    them available successively via a ``capsys.readouterr()`` method
-    which returns a ``(out, err)`` tuple of captured snapshot strings.
+    """enables capturing of writes to sys.stdout/sys.stderr and makes
+    captured output available via ``capsys.readouterr()`` method calls
+    which return a ``(out, err)`` tuple.
     """
     return CaptureFuncarg(py.io.StdCapture)
 
 def pytest_funcarg__capfd(request):
-    """captures writes to file descriptors 1 and 2 and makes
-    snapshotted ``(out, err)`` string tuples available
-    via the ``capsys.readouterr()`` method.  If the underlying
-    platform does not have ``os.dup`` (e.g. Jython) tests using
-    this funcarg will automatically skip.
+    """enables capturing of writes to file descriptors 1 and 2 and makes
+    captured output available via ``capsys.readouterr()`` method calls
+    which return a ``(out, err)`` tuple.
     """
     if not hasattr(os, 'dup'):
         py.test.skip("capfd funcarg needs os.dup")

_pytest/config.py

@@ -8,10 +8,16 @@ import pytest
 def pytest_cmdline_parse(pluginmanager, args):
     config = Config(pluginmanager)
     config.parse(args)
-    if config.option.debug:
-        config.trace.root.setwriter(sys.stderr.write)
     return config
 
+def pytest_unconfigure(config):
+    while 1:
+        try:
+            fin = config._cleanup.pop()
+        except IndexError:
+            break
+        fin()
+
 class Parser:
     """ Parser for command line arguments. """
 
@@ -34,7 +40,7 @@ class Parser:
 
     def getgroup(self, name, description="", after=None):
         """ get (or create) a named option Group.
-       
+
         :name: unique name of the option group.
         :description: long description for --help output.
         :after: name of other group, used for ordering --help output.
@@ -77,6 +83,7 @@ class Parser:
         self._inidict[name] = (help, type, default)
         self._ininames.append(name)
 
+
 class OptionGroup:
     def __init__(self, name, description="", parser=None):
         self.name = name
@@ -125,7 +132,6 @@ class Conftest(object):
         self._onimport = onimport
         self._conftestpath2mod = {}
         self._confcutdir = confcutdir
-        self._md5cache = {}
 
     def setinitial(self, args):
         """ try to find a first anchor path for looking up global values
@@ -179,14 +185,7 @@ class Conftest(object):
                 else:
                     conftestpath = path.join("conftest.py")
                     if conftestpath.check(file=1):
-                        key = conftestpath.computehash()
-                        # XXX logging about conftest loading
-                        if key not in self._md5cache:
-                            clist.append(self.importconftest(conftestpath))
-                            self._md5cache[key] = conftestpath
-                        else:
-                            # use some kind of logging
-                            print ("WARN: not loading %s" % conftestpath)
+                        clist.append(self.importconftest(conftestpath))
                 clist[:0] = self.getconftestmodules(dp)
             self._path2confmods[path] = clist
         # be defensive: avoid changes from caller side to
@@ -259,6 +258,20 @@ class Config(object):
         self._conftest = Conftest(onimport=self._onimportconftest)
         self.hook = self.pluginmanager.hook
         self._inicache = {}
+        self._cleanup = []
+
+    @classmethod
+    def fromdictargs(cls, option_dict, args):
+        """ constructor useable for subprocesses. """
+        config = cls()
+        # XXX slightly crude way to initialize capturing
+        import _pytest.capture
+        _pytest.capture.pytest_cmdline_parse(config.pluginmanager, args)
+        config._preparse(args, addopts=False)
+        config.option.__dict__.update(option_dict)
+        for x in config.option.plugins:
+            config.pluginmanager.consider_pluginarg(x)
+        return config
 
     def _onimportconftest(self, conftestmodule):
         self.trace("loaded conftestmodule %r" %(conftestmodule,))
@@ -278,13 +291,15 @@ class Config(object):
 
     def _setinitialconftest(self, args):
         # capture output during conftest init (#issue93)
-        name = hasattr(os, 'dup') and 'StdCaptureFD' or 'StdCapture'
-        cap = getattr(py.io, name)()
+        # XXX introduce load_conftest hook to avoid needing to know
+        # about capturing plugin here
+        capman = self.pluginmanager.getplugin("capturemanager")
+        capman.resumecapture()
         try:
             try:
                 self._conftest.setinitial(args)
             finally:
-                out, err = cap.reset()
+                out, err = capman.suspendcapture() # logging might have got it
         except:
             sys.stdout.write(out)
             sys.stderr.write(err)
@@ -300,11 +315,13 @@ class Config(object):
         if addopts:
             args[:] = self.getini("addopts") + args
         self._checkversion()
+        self.pluginmanager.consider_preparse(args)
         self.pluginmanager.consider_setuptools_entrypoints()
         self.pluginmanager.consider_env()
-        self.pluginmanager.consider_preparse(args)
         self._setinitialconftest(args)
         self.pluginmanager.do_addoption(self._parser)
+        if addopts:
+            self.hook.pytest_cmdline_preparse(config=self, args=args)
 
     def _checkversion(self):
         minver = self.inicfg.get('minversion', None)
@@ -322,6 +339,7 @@ class Config(object):
         # Note that this can only be called once per testing process.
         assert not hasattr(self, 'args'), (
                 "can only parse cmdline args at most once per Config object")
+        self._origargs = args
         self._preparse(args)
         self._parser.hints.extend(self.pluginmanager._hints)
         args = self._parser.parse_setoption(args, self.option)
@@ -329,6 +347,14 @@ class Config(object):
             args.append(py.std.os.getcwd())
         self.args = args
 
+    def addinivalue_line(self, name, line):
+        """ add a line to an ini-file option. The option must have been
+        declared but might not yet be set in which case the line becomes the
+        the first line in its value. """
+        x = self.getini(name)
+        assert isinstance(x, list)
+        x.append(line) # modifies the cached list inline
+
     def getini(self, name):
         """ return configuration value from an ini file. If the
         specified name hasn't been registered through a prior ``parse.addini``
@@ -397,7 +423,9 @@ class Config(object):
             return self._getconftest(name, path, check=False)
 
     def getvalueorskip(self, name, path=None):
-        """ (deprecated) return getvalue(name) or call py.test.skip if no value exists. """
+        """ (deprecated) return getvalue(name) or call
+        py.test.skip if no value exists. """
+        __tracebackhide__ = True
         try:
             val = self.getvalue(name, path)
             if val is None:
@@ -408,7 +436,7 @@ class Config(object):
 
 
 def getcfg(args, inibasenames):
-    args = [x for x in args if str(x)[0] != "-"]
+    args = [x for x in args if not str(x).startswith("-")]
     if not args:
         args = [py.path.local()]
     for arg in args:
@@ -421,7 +449,7 @@ def getcfg(args, inibasenames):
                     if 'pytest' in iniconfig.sections:
                         return iniconfig['pytest']
     return {}
-   
+
 def findupwards(current, basename):
     current = py.path.local(current)
     while 1:

_pytest/core.py

@@ -11,18 +11,15 @@ assert py.__version__.split(".")[:2] >= ['1', '4'], ("installation problem: "
     "%s is too old, remove or upgrade 'py'" % (py.__version__))
 
 default_plugins = (
- "config mark session terminal runner python pdb unittest capture skipping "
+ "config mark main terminal runner python pdb unittest capture skipping "
  "tmpdir monkeypatch recwarn pastebin helpconfig nose assertion genscript "
- "junitxml doctest").split()
-
-IMPORTPREFIX = "pytest_"
+ "junitxml resultlog doctest").split()
 
 class TagTracer:
-    def __init__(self, prefix="[pytest] "):
+    def __init__(self):
         self._tag2proc = {}
         self.writer = None
         self.indent = 0
-        self.prefix = prefix
 
     def get(self, name):
         return TagTracerSub(self, (name,))
@@ -32,7 +29,7 @@ class TagTracer:
             if args:
                 indent = "  " * self.indent
                 content = " ".join(map(str, args))
-                self.writer("%s%s%s\n" %(self.prefix, indent, content))
+                self.writer("%s%s [%s]\n" %(indent, content, ":".join(tags)))
         try:
             self._tag2proc[tags](tags, args)
         except KeyError:
@@ -62,9 +59,11 @@ class TagTracerSub:
 class PluginManager(object):
     def __init__(self, load=False):
         self._name2plugin = {}
+        self._listattrcache = {}
         self._plugins = []
         self._hints = []
         self.trace = TagTracer().get("pluginmanage")
+        self._plugin_distinfo = []
         if os.environ.get('PYTEST_DEBUG'):
             err = sys.stderr
             encoding = getattr(err, 'encoding', 'utf8')
@@ -79,20 +78,12 @@ class PluginManager(object):
             for spec in default_plugins:
                 self.import_plugin(spec)
 
-    def _getpluginname(self, plugin, name):
-        if name is None:
-            if hasattr(plugin, '__name__'):
-                name = plugin.__name__.split(".")[-1]
-            else:
-                name = id(plugin)
-        return name
-
     def register(self, plugin, name=None, prepend=False):
         assert not self.isregistered(plugin), plugin
-        assert not self.isregistered(plugin), plugin
-        name = self._getpluginname(plugin, name)
+        name = name or getattr(plugin, '__name__', str(id(plugin)))
         if name in self._name2plugin:
             return False
+        #self.trace("registering", name, plugin)
         self._name2plugin[name] = plugin
         self.call_plugin(plugin, "pytest_addhooks", {'pluginmanager': self})
         self.hook.pytest_plugin_registered(manager=self, plugin=plugin)
@@ -112,7 +103,7 @@ class PluginManager(object):
                 del self._name2plugin[name]
 
     def isregistered(self, plugin, name=None):
-        if self._getpluginname(plugin, name) in self._name2plugin:
+        if self.getplugin(name) is not None:
             return True
         for val in self._name2plugin.values():
             if plugin == val:
@@ -129,18 +120,15 @@ class PluginManager(object):
             py.test.skip("plugin %r is missing" % name)
 
     def hasplugin(self, name):
-        try:
-            self.getplugin(name)
-            return True
-        except KeyError:
-            return False
+        return bool(self.getplugin(name))
 
     def getplugin(self, name):
+        if name is None:
+            return None
         try:
             return self._name2plugin[name]
         except KeyError:
-            impname = canonical_importname(name)
-            return self._name2plugin[impname]
+            return self._name2plugin.get("_pytest." + name, None)
 
     # API for bootstrapping
     #
@@ -160,19 +148,32 @@ class PluginManager(object):
         except ImportError:
             return # XXX issue a warning
         for ep in iter_entry_points('pytest11'):
-            name = canonical_importname(ep.name)
-            if name in self._name2plugin:
+            name = ep.name
+            if name.startswith("pytest_"):
+                name = name[7:]
+            if ep.name in self._name2plugin or name in self._name2plugin:
                 continue
             try:
                 plugin = ep.load()
             except DistributionNotFound:
                 continue
+            self._plugin_distinfo.append((ep.dist, plugin))
             self.register(plugin, name=name)
 
     def consider_preparse(self, args):
         for opt1,opt2 in zip(args, args[1:]):
             if opt1 == "-p":
-                self.import_plugin(opt2)
+                self.consider_pluginarg(opt2)
+
+    def consider_pluginarg(self, arg):
+        if arg.startswith("no:"):
+            name = arg[3:]
+            if self.getplugin(name) is not None:
+                self.unregister(None, name=name)
+            self._name2plugin[name] = -1
+        else:
+            if self.getplugin(arg) is None:
+                self.import_plugin(arg)
 
     def consider_conftest(self, conftestmodule):
         if self.register(conftestmodule, name=conftestmodule.__file__):
@@ -186,15 +187,19 @@ class PluginManager(object):
             for spec in attr:
                 self.import_plugin(spec)
 
-    def import_plugin(self, spec):
-        assert isinstance(spec, str)
-        modname = canonical_importname(spec)
-        if modname in self._name2plugin:
+    def import_plugin(self, modname):
+        assert isinstance(modname, str)
+        if self.getplugin(modname) is not None:
             return
         try:
+            #self.trace("importing", modname)
             mod = importplugin(modname)
         except KeyboardInterrupt:
             raise
+        except ImportError:
+            if modname.startswith("pytest_"):
+                return self.import_plugin(modname[7:])
+            raise
         except:
             e = py.std.sys.exc_info()[1]
             if not hasattr(py.test, 'skip'):
@@ -206,6 +211,14 @@ class PluginManager(object):
             self.register(mod, modname)
             self.consider_module(mod)
 
+    def pytest_configure(self, config):
+        config.addinivalue_line("markers",
+            "tryfirst: mark a hook implementation function such that the "
+            "plugin machinery will try to call it first/as early as possible.")
+        config.addinivalue_line("markers",
+            "trylast: mark a hook implementation function such that the "
+            "plugin machinery will try to call it last/as late as possible.")
+
     def pytest_plugin_registered(self, plugin):
         import pytest
         dic = self.call_plugin(plugin, "pytest_namespace", {}) or {}
@@ -259,8 +272,15 @@ class PluginManager(object):
         config.hook.pytest_unconfigure(config=config)
         config.pluginmanager.unregister(self)
 
-    def notify_exception(self, excinfo):
-        excrepr = excinfo.getrepr(funcargs=True, showlocals=True)
+    def notify_exception(self, excinfo, option=None):
+        if option and option.fulltrace:
+            style = "long"
+        else:
+            style = "native"
+        excrepr = excinfo.getrepr(funcargs=True,
+            showlocals=getattr(option, 'showlocals', False),
+            style=style,
+        )
         res = self.hook.pytest_internalerror(excrepr=excrepr)
         if not py.builtin.any(res):
             for line in str(excrepr).split("\n"):
@@ -270,6 +290,11 @@ class PluginManager(object):
     def listattr(self, attrname, plugins=None):
         if plugins is None:
             plugins = self._plugins
+        key = (attrname,) + tuple(plugins)
+        try:
+            return list(self._listattrcache[key])
+        except KeyError:
+            pass
         l = []
         last = []
         for plugin in plugins:
@@ -284,40 +309,25 @@ class PluginManager(object):
             except AttributeError:
                 continue
         l.extend(last)
+        self._listattrcache[key] = list(l)
         return l
 
     def call_plugin(self, plugin, methname, kwargs):
         return MultiCall(methods=self.listattr(methname, plugins=[plugin]),
                 kwargs=kwargs, firstresult=True).execute()
 
-def canonical_importname(name):
-    if '.' in name:
-        return name
-    name = name.lower()
-    if not name.startswith(IMPORTPREFIX):
-        name = IMPORTPREFIX + name
-    return name
 
 def importplugin(importspec):
-    #print "importing", importspec
+    name = importspec
     try:
-        return __import__(importspec, None, None, '__doc__')
+        mod = "_pytest." + name
+        return __import__(mod, None, None, '__doc__')
     except ImportError:
-        e = py.std.sys.exc_info()[1]
-        if str(e).find(importspec) == -1:
-            raise
-        name = importspec
-        try:
-            if name.startswith("pytest_"):
-                name = importspec[7:]
-            return __import__("_pytest.%s" %(name), None, None, '__doc__')
-        except ImportError:
-            e = py.std.sys.exc_info()[1]
-            if str(e).find(name) == -1:
-                raise
-            # show the original exception, not the failing internal one
-            return __import__(importspec, None, None, '__doc__')
-
+        #e = py.std.sys.exc_info()[1]
+        #if str(e).find(name) == -1:
+        #    raise
+        pass #
+    return __import__(importspec, None, None, '__doc__')
 
 class MultiCall:
     """ execute a call into multiple python functions/methods. """
@@ -354,14 +364,20 @@ class MultiCall:
         return kwargs
 
 def varnames(func):
+    try:
+        return func._varnames
+    except AttributeError:
+        pass
     if not inspect.isfunction(func) and not inspect.ismethod(func):
         func = getattr(func, '__call__', func)
     ismethod = inspect.ismethod(func)
     rawcode = py.code.getrawcode(func)
     try:
-        return rawcode.co_varnames[ismethod:rawcode.co_argcount]
+        x = rawcode.co_varnames[ismethod:rawcode.co_argcount]
     except AttributeError:
-        return ()
+        x = ()
+    py.builtin._getfuncdict(func)['_varnames'] = x
+    return x
 
 class HookRelay:
     def __init__(self, hookspecs, pm, prefix="pytest_"):
@@ -378,9 +394,6 @@ class HookRelay:
         added = False
         for name, method in vars(hookspecs).items():
             if name.startswith(prefix):
-                if not method.__doc__:
-                    raise ValueError("docstring required for hook %r, in %r"
-                        % (method, hookspecs))
                 firstresult = getattr(method, 'firstresult', False)
                 hc = HookCaller(self, name, firstresult=firstresult)
                 setattr(self, name, hc)
@@ -426,10 +439,7 @@ _preinit = []
 def _preloadplugins():
     _preinit.append(PluginManager(load=True))
 
-def main(args=None, plugins=None):
-    """ returned exit code integer, after an in-process testing run
-    with the given command line arguments, preloading an optional list
-    of passed in plugin objects. """
+def _prepareconfig(args=None, plugins=None):
     if args is None:
         args = sys.argv[1:]
     elif isinstance(args, py.path.local):
@@ -443,13 +453,19 @@ def main(args=None, plugins=None):
     else: # subsequent calls to main will create a fresh instance
         _pluginmanager = PluginManager(load=True)
     hook = _pluginmanager.hook
+    if plugins:
+        for plugin in plugins:
+            _pluginmanager.register(plugin)
+    return hook.pytest_cmdline_parse(
+            pluginmanager=_pluginmanager, args=args)
+
+def main(args=None, plugins=None):
+    """ returned exit code integer, after an in-process testing run
+    with the given command line arguments, preloading an optional list
+    of passed in plugin objects. """
     try:
-        if plugins:
-            for plugin in plugins:
-                _pluginmanager.register(plugin)
-        config = hook.pytest_cmdline_parse(
-                pluginmanager=_pluginmanager, args=args)
-        exitstatus = hook.pytest_cmdline_main(config=config)
+        config = _prepareconfig(args, plugins)
+        exitstatus = config.hook.pytest_cmdline_main(config=config)
     except UsageError:
         e = sys.exc_info()[1]
         sys.stderr.write("ERROR: %s\n" %(e.args[0],))

_pytest/doctest.py

@@ -34,7 +34,9 @@ class ReprFailDoctest(TerminalRepr):
 
 class DoctestItem(pytest.Item):
     def repr_failure(self, excinfo):
-        if excinfo.errisinstance(py.std.doctest.DocTestFailure):
+        doctest = py.std.doctest
+        if excinfo.errisinstance((doctest.DocTestFailure,
+                                  doctest.UnexpectedException)):
             doctestfailure = excinfo.value
             example = doctestfailure.example
             test = doctestfailure.test
@@ -50,12 +52,15 @@ class DoctestItem(pytest.Item):
             for line in filelines[i:lineno]:
                 lines.append("%03d %s" % (i+1, line))
                 i += 1
-            lines += checker.output_difference(example,
-                    doctestfailure.got, REPORT_UDIFF).split("\n")
+            if excinfo.errisinstance(doctest.DocTestFailure):
+                lines += checker.output_difference(example,
+                        doctestfailure.got, REPORT_UDIFF).split("\n")
+            else:
+                inner_excinfo = py.code.ExceptionInfo(excinfo.value.exc_info)
+                lines += ["UNEXPECTED EXCEPTION: %s" %
+                            repr(inner_excinfo.value)]
+                lines += py.std.traceback.format_exception(*excinfo.value.exc_info)
             return ReprFailDoctest(reprlocation, lines)
-        elif excinfo.errisinstance(py.std.doctest.UnexpectedException):
-            excinfo = py.code.ExceptionInfo(excinfo.value.exc_info)
-            return super(DoctestItem, self).repr_failure(excinfo)
         else:
             return super(DoctestItem, self).repr_failure(excinfo)
 

_pytest/genscript.py

@@ -1,8 +1,5 @@
 """ generate a single-file self-contained version of py.test """
 import py
-import pickle
-import zlib
-import base64
 
 def find_toplevel(name):
     for syspath in py.std.sys.path:
@@ -31,9 +28,9 @@ def pkg_to_mapping(name):
     return name2src
 
 def compress_mapping(mapping):
-    data = pickle.dumps(mapping, 2)
-    data = zlib.compress(data, 9)
-    data = base64.encodestring(data)
+    data = py.std.pickle.dumps(mapping, 2)
+    data = py.std.zlib.compress(data, 9)
+    data = py.std.base64.encodestring(data)
     data = data.decode('ascii')
     return data
 
@@ -44,7 +41,6 @@ def compress_packages(names):
         mapping.update(pkg_to_mapping(name))
     return compress_mapping(mapping)
 
-
 def generate_script(entry, packages):
     data = compress_packages(packages)
     tmpl = py.path.local(__file__).dirpath().join('standalonetemplate.py')

_pytest/helpconfig.py

@@ -1,7 +1,8 @@
 """ version info, help messages, tracing configuration.  """
 import py
 import pytest
-import inspect, sys
+import os, inspect, sys
+from _pytest.core import varnames
 
 def pytest_addoption(parser):
     group = parser.getgroup('debugconfig')
@@ -15,12 +16,31 @@ def pytest_addoption(parser):
     group.addoption('--traceconfig',
                action="store_true", dest="traceconfig", default=False,
                help="trace considerations of conftest.py files."),
-    group._addoption('--nomagic',
-               action="store_true", dest="nomagic", default=False,
-               help="don't reinterpret asserts, no traceback cutting. ")
     group.addoption('--debug',
                action="store_true", dest="debug", default=False,
-               help="generate and show internal debugging information.")
+               help="store internal tracing debug information in 'pytestdebug.log'.")
+
+
+def pytest_cmdline_parse(__multicall__):
+    config = __multicall__.execute()
+    if config.option.debug:
+        path = os.path.abspath("pytestdebug.log")
+        f = open(path, 'w')
+        config._debugfile = f
+        f.write("versions pytest-%s, py-%s, python-%s\ncwd=%s\nargs=%s\n\n" %(
+            pytest.__version__, py.__version__, ".".join(map(str, sys.version_info)),
+            os.getcwd(), config._origargs))
+        config.trace.root.setwriter(f.write)
+        sys.stderr.write("writing pytestdebug information to %s\n" % path)
+    return config
+
+@pytest.mark.trylast
+def pytest_unconfigure(config):
+    if hasattr(config, '_debugfile'):
+        config._debugfile.close()
+        sys.stderr.write("wrote pytestdebug information to %s\n" %
+            config._debugfile.name)
+        config.trace.root.setwriter(None)
 
 
 def pytest_cmdline_main(config):
@@ -28,10 +48,15 @@ def pytest_cmdline_main(config):
         p = py.path.local(pytest.__file__)
         sys.stderr.write("This is py.test version %s, imported from %s\n" %
             (pytest.__version__, p))
+        plugininfo = getpluginversioninfo(config)
+        if plugininfo:
+            for line in plugininfo:
+                sys.stderr.write(line + "\n")
         return 0
     elif config.option.help:
         config.pluginmanager.do_configure(config)
         showhelp(config)
+        config.pluginmanager.do_unconfigure(config)
         return 0
 
 def showhelp(config):
@@ -69,18 +94,37 @@ conftest_options = [
     ('pytest_plugins', 'list of plugin names to load'),
 ]
 
+def getpluginversioninfo(config):
+    lines = []
+    plugininfo = config.pluginmanager._plugin_distinfo
+    if plugininfo:
+        lines.append("setuptools registered plugins:")
+        for dist, plugin in plugininfo:
+            loc = getattr(plugin, '__file__', repr(plugin))
+            content = "%s-%s at %s" % (dist.project_name, dist.version, loc)
+            lines.append("  " + content)
+    return lines
+
 def pytest_report_header(config):
     lines = []
     if config.option.debug or config.option.traceconfig:
         lines.append("using: pytest-%s pylib-%s" %
             (pytest.__version__,py.__version__))
-            
+
+        verinfo = getpluginversioninfo(config)
+        if verinfo:
+            lines.extend(verinfo)
+
     if config.option.traceconfig:
         lines.append("active plugins:")
         plugins = []
         items = config.pluginmanager._name2plugin.items()
         for name, plugin in items:
-            lines.append("    %-20s: %s" %(name, repr(plugin)))
+            if hasattr(plugin, '__file__'):
+                r = plugin.__file__
+            else:
+                r = repr(plugin)
+            lines.append("    %-20s: %s" %(name, r))
     return lines
 
 
@@ -112,12 +156,11 @@ def pytest_plugin_registered(manager, plugin):
                 fail = True
         else:
             #print "checking", method
-            method_args = getargs(method)
-            #print "method_args", method_args
+            method_args = list(varnames(method))
             if '__multicall__' in method_args:
                 method_args.remove('__multicall__')
             hook = hooks[name]
-            hookargs = getargs(hook)
+            hookargs = varnames(hook)
             for arg in method_args:
                 if arg not in hookargs:
                     Print("argument %r not available"  %(arg, ))
@@ -139,11 +182,6 @@ def isgenerichook(name):
     return name == "pytest_plugins" or \
            name.startswith("pytest_funcarg__")
 
-def getargs(func):
-    args = inspect.getargs(py.code.getrawcode(func))[0]
-    startindex = inspect.ismethod(func) and 1 or 0
-    return args[startindex:]
-
 def collectattr(obj):
     methods = {}
     for apiname in dir(obj):

_pytest/hookspec.py

@@ -19,6 +19,9 @@ def pytest_cmdline_parse(pluginmanager, args):
     """return initialized config object, parsing the specified args. """
 pytest_cmdline_parse.firstresult = True
 
+def pytest_cmdline_preparse(config, args):
+    """modify command line arguments before option parsing. """
+
 def pytest_addoption(parser):
     """add optparse-style options and ini-style config values via calls
     to ``parser.addoption`` and ``parser.addini(...)``.
@@ -118,16 +121,23 @@ def pytest_generate_tests(metafunc):
 def pytest_itemstart(item, node=None):
     """ (deprecated, use pytest_runtest_logstart). """
 
-def pytest_runtest_protocol(item):
-    """ implements the standard runtest_setup/call/teardown protocol including
-    capturing exceptions and calling reporting hooks on the results accordingly.
+def pytest_runtest_protocol(item, nextitem):
+    """ implements the runtest_setup/call/teardown protocol for
+    the given test item, including capturing exceptions and calling
+    reporting hooks.
+
+    :arg item: test item for which the runtest protocol is performed.
+
+    :arg nexitem: the scheduled-to-be-next test item (or None if this
+                  is the end my friend).  This argument is passed on to
+                  :py:func:`pytest_runtest_teardown`.
 
     :return boolean: True if no further hook implementations should be invoked.
     """
 pytest_runtest_protocol.firstresult = True
 
 def pytest_runtest_logstart(nodeid, location):
-    """ signal the start of a test run. """
+    """ signal the start of running a single test item. """
 
 def pytest_runtest_setup(item):
     """ called before ``pytest_runtest_call(item)``. """
@@ -135,8 +145,14 @@ def pytest_runtest_setup(item):
 def pytest_runtest_call(item):
     """ called to execute the test ``item``. """
 
-def pytest_runtest_teardown(item):
-    """ called after ``pytest_runtest_call``. """
+def pytest_runtest_teardown(item, nextitem):
+    """ called after ``pytest_runtest_call``.
+
+    :arg nexitem: the scheduled-to-be-next test item (None if no further
+                  test item is scheduled).  This argument can be used to
+                  perform exact teardowns, i.e. calling just enough finalizers
+                  so that nextitem only needs to call setup-functions.
+    """
 
 def pytest_runtest_makereport(item, call):
     """ return a :py:class:`_pytest.runner.TestReport` object
@@ -146,15 +162,8 @@ def pytest_runtest_makereport(item, call):
 pytest_runtest_makereport.firstresult = True
 
 def pytest_runtest_logreport(report):
-    """ process item test report. """
-
-# special handling for final teardown - somewhat internal for now
-def pytest__teardown_final(session):
-    """ called before test session finishes. """
-pytest__teardown_final.firstresult = True
-
-def pytest__teardown_final_logerror(report, session):
-    """ called if runtest_teardown_final failed. """
+    """ process a test setup/call/teardown report relating to
+    the respective phase of executing a test. """
 
 # -------------------------------------------------------------------------
 # test session related hooks
@@ -202,7 +211,6 @@ def pytest_doctest_prepare_content(content):
     """ return processed content for a given doctest"""
 pytest_doctest_prepare_content.firstresult = True
 
-
 # -------------------------------------------------------------------------
 # error handling and internal debugging hooks
 # -------------------------------------------------------------------------

_pytest/junitxml.py

@@ -5,8 +5,54 @@ Based on initial code from Ross Lawley.
 
 import py
 import os
+import re
+import sys
 import time
 
+
+# Python 2.X and 3.X compatibility
+try:
+    unichr(65)
+except NameError:
+    unichr = chr
+try:
+    unicode('A')
+except NameError:
+    unicode = str
+try:
+    long(1)
+except NameError:
+    long = int
+
+
+class Junit(py.xml.Namespace):
+    pass
+
+
+# We need to get the subset of the invalid unicode ranges according to
+# XML 1.0 which are valid in this python build.  Hence we calculate
+# this dynamically instead of hardcoding it.  The spec range of valid
+# chars is: Char ::= #x9 | #xA | #xD | [#x20-#xD7FF] | [#xE000-#xFFFD]
+#                    | [#x10000-#x10FFFF]
+_illegal_unichrs = [(0x00, 0x08), (0x0B, 0x0C), (0x0E, 0x19),
+                   (0xD800, 0xDFFF), (0xFDD0, 0xFFFF)]
+_illegal_ranges = [unicode("%s-%s") % (unichr(low), unichr(high))
+                  for (low, high) in _illegal_unichrs
+                  if low < sys.maxunicode]
+illegal_xml_re = re.compile(unicode('[%s]') %
+                            unicode('').join(_illegal_ranges))
+del _illegal_unichrs
+del _illegal_ranges
+
+def bin_xml_escape(arg):
+    def repl(matchobj):
+        i = ord(matchobj.group())
+        if i <= 0xFF:
+            return unicode('#x%02X') % i
+        else:
+            return unicode('#x%04X') % i
+    return illegal_xml_re.sub(repl, py.xml.escape(arg))
+
 def pytest_addoption(parser):
     group = parser.getgroup("terminal reporting")
     group.addoption('--junitxml', action="store", dest="xmlpath",
@@ -28,109 +74,103 @@ def pytest_unconfigure(config):
         del config._xml
         config.pluginmanager.unregister(xml)
 
+
 class LogXML(object):
     def __init__(self, logfile, prefix):
-        self.logfile = logfile
+        logfile = os.path.expanduser(os.path.expandvars(logfile))
+        self.logfile = os.path.normpath(logfile)
         self.prefix = prefix
-        self.test_logs = []
+        self.tests = []
         self.passed = self.skipped = 0
         self.failed = self.errors = 0
-        self._durations = {}
 
     def _opentestcase(self, report):
         names = report.nodeid.split("::")
         names[0] = names[0].replace("/", '.')
-        names = tuple(names)
-        d = {'time': self._durations.pop(names, "0")}
         names = [x.replace(".py", "") for x in names if x != "()"]
         classnames = names[:-1]
         if self.prefix:
             classnames.insert(0, self.prefix)
-        d['classname'] = ".".join(classnames)
-        d['name'] = py.xml.escape(names[-1])
-        attrs = ['%s="%s"' % item for item in sorted(d.items())]
-        self.test_logs.append("\n<testcase %s>" % " ".join(attrs))
+        self.tests.append(Junit.testcase(
+            classname=".".join(classnames),
+            name=names[-1],
+            time=getattr(report, 'duration', 0)
+        ))
 
-    def _closetestcase(self):
-        self.test_logs.append("</testcase>")
-
-    def appendlog(self, fmt, *args):
-        args = tuple([py.xml.escape(arg) for arg in args])
-        self.test_logs.append(fmt % args)
+    def append(self, obj):
+        self.tests[-1].append(obj)
 
     def append_pass(self, report):
         self.passed += 1
-        self._opentestcase(report)
-        self._closetestcase()
 
     def append_failure(self, report):
-        self._opentestcase(report)
         #msg = str(report.longrepr.reprtraceback.extraline)
         if "xfail" in report.keywords:
-            self.appendlog(
-                '<skipped message="xfail-marked test passes unexpectedly"/>')
+            self.append(
+                Junit.skipped(message="xfail-marked test passes unexpectedly"))
             self.skipped += 1
         else:
-            self.appendlog('<failure message="test failure">%s</failure>',
-                report.longrepr)
+            sec = dict(report.sections)
+            fail = Junit.failure(message="test failure")
+            fail.append(str(report.longrepr))
+            self.append(fail)
+            for name in ('out', 'err'):
+                content = sec.get("Captured std%s" % name)
+                if content:
+                    tag = getattr(Junit, 'system-'+name)
+                    self.append(tag(bin_xml_escape(content)))
             self.failed += 1
-        self._closetestcase()
 
     def append_collect_failure(self, report):
-        self._opentestcase(report)
         #msg = str(report.longrepr.reprtraceback.extraline)
-        self.appendlog('<failure message="collection failure">%s</failure>',
-            report.longrepr)
-        self._closetestcase()
+        self.append(Junit.failure(str(report.longrepr),
+                                  message="collection failure"))
         self.errors += 1
 
     def append_collect_skipped(self, report):
-        self._opentestcase(report)
         #msg = str(report.longrepr.reprtraceback.extraline)
-        self.appendlog('<skipped message="collection skipped">%s</skipped>',
-            report.longrepr)
-        self._closetestcase()
+        self.append(Junit.skipped(str(report.longrepr),
+                                  message="collection skipped"))
         self.skipped += 1
 
     def append_error(self, report):
-        self._opentestcase(report)
-        self.appendlog('<error message="test setup failure">%s</error>',
-            report.longrepr)
-        self._closetestcase()
+        self.append(Junit.error(str(report.longrepr),
+                                message="test setup failure"))
         self.errors += 1
 
     def append_skipped(self, report):
-        self._opentestcase(report)
         if "xfail" in report.keywords:
-            self.appendlog(
-                '<skipped message="expected test failure">%s</skipped>',
-                report.keywords['xfail'])
+            self.append(Junit.skipped(str(report.keywords['xfail']),
+                                      message="expected test failure"))
         else:
-            self.appendlog("<skipped/>")
-        self._closetestcase()
+            filename, lineno, skipreason = report.longrepr
+            if skipreason.startswith("Skipped: "):
+                skipreason = skipreason[9:]
+            self.append(
+                Junit.skipped("%s:%s: %s" % report.longrepr,
+                              type="pytest.skip",
+                              message=skipreason
+                ))
         self.skipped += 1
 
     def pytest_runtest_logreport(self, report):
         if report.passed:
-            self.append_pass(report)
+            if report.when == "call": # ignore setup/teardown
+                self._opentestcase(report)
+                self.append_pass(report)
         elif report.failed:
+            self._opentestcase(report)
             if report.when != "call":
                 self.append_error(report)
             else:
                 self.append_failure(report)
         elif report.skipped:
+            self._opentestcase(report)
             self.append_skipped(report)
 
-    def pytest_runtest_call(self, item, __multicall__):
-        names = tuple(item.listnames())
-        start = time.time()
-        try:
-            return __multicall__.execute()
-        finally:
-            self._durations[names] = time.time() - start
-
     def pytest_collectreport(self, report):
         if not report.passed:
+            self._opentestcase(report)
             if report.failed:
                 self.append_collect_failure(report)
             else:
@@ -139,10 +179,11 @@ class LogXML(object):
     def pytest_internalerror(self, excrepr):
         self.errors += 1
         data = py.xml.escape(excrepr)
-        self.test_logs.append(
-            '\n<testcase classname="pytest" name="internal">'
-            '    <error message="internal error">'
-            '%s</error></testcase>' % data)
+        self.tests.append(
+            Junit.testcase(
+                    Junit.error(data, message="internal error"),
+                    classname="pytest",
+                    name="internal"))
 
     def pytest_sessionstart(self, session):
         self.suite_start_time = time.time()
@@ -156,17 +197,17 @@ class LogXML(object):
         suite_stop_time = time.time()
         suite_time_delta = suite_stop_time - self.suite_start_time
         numtests = self.passed + self.failed
+
         logfile.write('<?xml version="1.0" encoding="utf-8"?>')
-        logfile.write('<testsuite ')
-        logfile.write('name="" ')
-        logfile.write('errors="%i" ' % self.errors)
-        logfile.write('failures="%i" ' % self.failed)
-        logfile.write('skips="%i" ' % self.skipped)
-        logfile.write('tests="%i" ' % numtests)
-        logfile.write('time="%.3f"' % suite_time_delta)
-        logfile.write(' >')
-        logfile.writelines(self.test_logs)
-        logfile.write('</testsuite>')
+        logfile.write(Junit.testsuite(
+            self.tests,
+            name="",
+            errors=self.errors,
+            failures=self.failed,
+            skips=self.skipped,
+            tests=numtests,
+            time="%.3f" % suite_time_delta,
+        ).unicode(indent=0))
         logfile.close()
 
     def pytest_terminal_summary(self, terminalreporter):

_pytest/main.py

@@ -2,7 +2,7 @@
 
 import py
 import pytest, _pytest
-import os, sys
+import os, sys, imp
 tracebackcutdir = py.path.local(_pytest.__file__).dirpath()
 
 # exitcodes for the command line
@@ -11,6 +11,8 @@ EXIT_TESTSFAILED = 1
 EXIT_INTERRUPTED = 2
 EXIT_INTERNALERROR = 3
 
+name_re = py.std.re.compile("^[a-zA-Z_]\w*$")
+
 def pytest_addoption(parser):
     parser.addini("norecursedirs", "directory patterns to avoid for recursion",
         type="args", default=('.*', 'CVS', '_darcs', '{arch}'))
@@ -27,6 +29,9 @@ def pytest_addoption(parser):
                action="store", type="int", dest="maxfail", default=0,
                help="exit after first num failures or errors.")
 
+    group._addoption('--strict', action="store_true",
+               help="run pytest in strict mode, warnings become errors.")
+
     group = parser.getgroup("collect", "collection")
     group.addoption('--collectonly',
         action="store_true", dest="collectonly",
@@ -40,29 +45,31 @@ def pytest_addoption(parser):
         help="only load conftest.py's relative to specified dir.")
 
     group = parser.getgroup("debugconfig",
-        "test process debugging and configuration")
+        "test session debugging and configuration")
     group.addoption('--basetemp', dest="basetemp", default=None, metavar="dir",
                help="base temporary directory for this test run.")
 
 
 def pytest_namespace():
-    return dict(collect=dict(Item=Item, Collector=Collector, File=File))
-        
+    collect = dict(Item=Item, Collector=Collector, File=File, Session=Session)
+    return dict(collect=collect)
+
 def pytest_configure(config):
     py.test.config = config # compatibiltiy
     if config.option.exitfirst:
         config.option.maxfail = 1
 
-def pytest_cmdline_main(config):
-    """ default command line protocol for initialization, session,
-    running tests and reporting. """
+def wrap_session(config, doit):
+    """Skeleton command line program"""
     session = Session(config)
     session.exitstatus = EXIT_OK
+    initstate = 0
     try:
         config.pluginmanager.do_configure(config)
+        initstate = 1
         config.hook.pytest_sessionstart(session=session)
-        config.hook.pytest_collection(session=session)
-        config.hook.pytest_runtestloop(session=session)
+        initstate = 2
+        doit(config, session)
     except pytest.UsageError:
         raise
     except KeyboardInterrupt:
@@ -71,30 +78,40 @@ def pytest_cmdline_main(config):
         session.exitstatus = EXIT_INTERRUPTED
     except:
         excinfo = py.code.ExceptionInfo()
-        config.pluginmanager.notify_exception(excinfo)
+        config.pluginmanager.notify_exception(excinfo, config.option)
         session.exitstatus = EXIT_INTERNALERROR
         if excinfo.errisinstance(SystemExit):
             sys.stderr.write("mainloop: caught Spurious SystemExit!\n")
+    if initstate >= 2:
+        config.hook.pytest_sessionfinish(session=session,
+            exitstatus=session.exitstatus or (session._testsfailed and 1))
     if not session.exitstatus and session._testsfailed:
         session.exitstatus = EXIT_TESTSFAILED
-    config.hook.pytest_sessionfinish(session=session,
-        exitstatus=session.exitstatus)
-    config.pluginmanager.do_unconfigure(config)
+    if initstate >= 1:
+        config.pluginmanager.do_unconfigure(config)
     return session.exitstatus
 
+def pytest_cmdline_main(config):
+    return wrap_session(config, _main)
+
+def _main(config, session):
+    """ default command line protocol for initialization, session,
+    running tests and reporting. """
+    config.hook.pytest_collection(session=session)
+    config.hook.pytest_runtestloop(session=session)
+
 def pytest_collection(session):
-    session.perform_collect()
-    hook = session.config.hook
-    hook.pytest_collection_modifyitems(session=session,
-        config=session.config, items=session.items)
-    hook.pytest_collection_finish(session=session)
-    return True
+    return session.perform_collect()
 
 def pytest_runtestloop(session):
     if session.config.option.collectonly:
         return True
-    for item in session.session.items:
-        item.config.hook.pytest_runtest_protocol(item=item)
+    for i, item in enumerate(session.items):
+        try:
+            nextitem = session.items[i+1]
+        except IndexError:
+            nextitem = None
+        item.config.hook.pytest_runtest_protocol(item=item, nextitem=nextitem)
         if session.shouldstop:
             raise session.Interrupted(session.shouldstop)
     return True
@@ -121,13 +138,10 @@ class HookProxy:
 
 def compatproperty(name):
     def fget(self):
-        #print "retrieving %r property from %s" %(name, self.fspath)
-        py.log._apiwarn("2.0", "use pytest.%s for "
-            "test collection and item classes" % name)
         return getattr(pytest, name)
     return property(fget, None, None,
         "deprecated attribute %r, use pytest.%s" % (name,name))
-    
+
 class Node(object):
     """ base class for all Nodes in the collection tree.
     Collector subclasses have children, Items are terminal nodes."""
@@ -138,13 +152,13 @@ class Node(object):
 
         #: the parent collector node.
         self.parent = parent
-        
+
         #: the test config object
         self.config = config or parent.config
 
         #: the collection this node is part of
         self.session = session or parent.session
-        
+
         #: filesystem path where this node was collected from
         self.fspath = getattr(parent, 'fspath', None)
         self.ihook = self.session.gethookproxy(self.fspath)
@@ -152,10 +166,19 @@ class Node(object):
 
     Module = compatproperty("Module")
     Class = compatproperty("Class")
+    Instance = compatproperty("Instance")
     Function = compatproperty("Function")
     File = compatproperty("File")
     Item = compatproperty("Item")
 
+    def _getcustomclass(self, name):
+        cls = getattr(self, name)
+        if cls != getattr(pytest, name):
+            py.log._apiwarn("2.0", "use of node.%s is deprecated, "
+                "use pytest_pycollect_makeitem(...) to create custom "
+                "collection nodes" % name)
+        return cls
+
     def __repr__(self):
         return "<%s %r>" %(self.__class__.__name__, getattr(self, 'name', None))
 
@@ -210,13 +233,13 @@ class Node(object):
     def listchain(self):
         """ return list of all parent collectors up to self,
             starting from root of collection tree. """
-        l = [self]
-        while 1:
-            x = l[0]
-            if x.parent is not None: # and x.parent.parent is not None:
-                l.insert(0, x.parent)
-            else:
-                return l
+        chain = []
+        item = self
+        while item is not None:
+            chain.append(item)
+            item = item.parent
+        chain.reverse()
+        return chain
 
     def listnames(self):
         return [x.name for x in self.listchain()]
@@ -311,6 +334,8 @@ class Item(Node):
     """ a basic test invocation item. Note that for a single function
     there might be multiple test invocation items.
     """
+    nextitem = None
+
     def reportinfo(self):
         return self.fspath, None, ""
 
@@ -320,7 +345,13 @@ class Item(Node):
             return self._location
         except AttributeError:
             location = self.reportinfo()
-            fspath = self.session.fspath.bestrelpath(location[0])
+            # bestrelpath is a quite slow function
+            cache = self.config.__dict__.setdefault("_bestrelpathcache", {})
+            try:
+                fspath = cache[location[0]]
+            except KeyError:
+                fspath = self.session.fspath.bestrelpath(location[0])
+                cache[location[0]] = fspath
             location = (fspath, location[1], str(location[2]))
             self._location = location
             return location
@@ -336,7 +367,7 @@ class Session(FSCollector):
     def __init__(self, config):
         super(Session, self).__init__(py.path.local(), parent=None,
             config=config, session=self)
-        self.config.pluginmanager.register(self, name="session", prepend=True)
+        assert self.config.pluginmanager.register(self, name="session", prepend=True)
         self._testsfailed = 0
         self.shouldstop = False
         self.trace = config.trace.root.get("collection")
@@ -362,6 +393,16 @@ class Session(FSCollector):
         return HookProxy(fspath, self.config)
 
     def perform_collect(self, args=None, genitems=True):
+        hook = self.config.hook
+        try:
+            items = self._perform_collect(args, genitems)
+            hook.pytest_collection_modifyitems(session=self,
+                config=self.config, items=items)
+        finally:
+            hook.pytest_collection_finish(session=self)
+        return items
+
+    def _perform_collect(self, args, genitems):
         if args is None:
             args = self.config.args
         self.trace("perform_collect", self, args)
@@ -369,6 +410,7 @@ class Session(FSCollector):
         self._notfound = []
         self._initialpaths = set()
         self._initialparts = []
+        self.items = items = []
         for arg in args:
             parts = self._parsearg(arg)
             self._initialparts.append(parts)
@@ -384,7 +426,6 @@ class Session(FSCollector):
         if not genitems:
             return rep.result
         else:
-            self.items = items = []
             if rep.passed:
                 for node in rep.result:
                     self.items.extend(self.genitems(node))
@@ -439,16 +480,29 @@ class Session(FSCollector):
         return True
 
     def _tryconvertpyarg(self, x):
-        try:
-            mod = __import__(x, None, None, ['__doc__'])
-        except (ValueError, ImportError):
-            return x
-        p = py.path.local(mod.__file__)
-        if p.purebasename == "__init__":
-            p = p.dirpath()
-        else:
-            p = p.new(basename=p.purebasename+".py")
-        return p
+        mod = None
+        path = [os.path.abspath('.')] + sys.path
+        for name in x.split('.'):
+            # ignore anything that's not a proper name here
+            # else something like --pyargs will mess up '.'
+            # since imp.find_module will actually sometimes work for it
+            # but it's supposed to be considered a filesystem path
+            # not a package
+            if name_re.match(name) is None:
+                return x
+            try:
+                fd, mod, type_ = imp.find_module(name, path)
+            except ImportError:
+                return x
+            else:
+                if fd is not None:
+                    fd.close()
+
+            if type_[2] != imp.PKG_DIRECTORY:
+                path = [os.path.dirname(mod)]
+            else:
+                path = [mod]
+        return mod
 
     def _parsearg(self, arg):
         """ return (fspath, names) tuple after checking the file exists. """
@@ -466,7 +520,7 @@ class Session(FSCollector):
             raise pytest.UsageError(msg + arg)
         parts[0] = path
         return parts
-   
+
     def matchnodes(self, matching, names):
         self.trace("matchnodes", matching, names)
         self.trace.root.indent += 1
@@ -494,9 +548,15 @@ class Session(FSCollector):
             node.ihook.pytest_collectstart(collector=node)
             rep = node.ihook.pytest_make_collect_report(collector=node)
             if rep.passed:
+                has_matched = False
                 for x in rep.result:
                     if x.name == name:
                         resultnodes.extend(self.matchnodes([x], nextnames))
+                        has_matched = True
+                # XXX accept IDs that don't have "()" for class instances
+                if not has_matched and len(rep.result) == 1 and x.name == "()":
+                    nextnames.insert(0, name)
+                    resultnodes.extend(self.matchnodes([x], nextnames))
             node.ihook.pytest_collectreport(report=rep)
         return resultnodes
 

_pytest/mark.py

@@ -14,12 +14,37 @@ def pytest_addoption(parser):
              "Terminate expression with ':' to make the first match match "
              "all subsequent tests (usually file-order). ")
 
+    group._addoption("-m",
+        action="store", dest="markexpr", default="", metavar="MARKEXPR",
+        help="only run tests matching given mark expression.  "
+             "example: -m 'mark1 and not mark2'."
+             )
+
+    group.addoption("--markers", action="store_true", help=
+        "show markers (builtin, plugin and per-project ones).")
+
+    parser.addini("markers", "markers for test functions", 'linelist')
+
+def pytest_cmdline_main(config):
+    if config.option.markers:
+        config.pluginmanager.do_configure(config)
+        tw = py.io.TerminalWriter()
+        for line in config.getini("markers"):
+            name, rest = line.split(":", 1)
+            tw.write("@pytest.mark.%s:" %  name, bold=True)
+            tw.line(rest)
+            tw.line()
+        config.pluginmanager.do_unconfigure(config)
+        return 0
+pytest_cmdline_main.tryfirst = True
+
 def pytest_collection_modifyitems(items, config):
     keywordexpr = config.option.keyword
-    if not keywordexpr:
+    matchexpr = config.option.markexpr
+    if not keywordexpr and not matchexpr:
         return
     selectuntil = False
-    if keywordexpr[-1] == ":":
+    if keywordexpr[-1:] == ":":
         selectuntil = True
         keywordexpr = keywordexpr[:-1]
 
@@ -29,21 +54,38 @@ def pytest_collection_modifyitems(items, config):
         if keywordexpr and skipbykeyword(colitem, keywordexpr):
             deselected.append(colitem)
         else:
-            remaining.append(colitem)
             if selectuntil:
                 keywordexpr = None
+            if matchexpr:
+                if not matchmark(colitem, matchexpr):
+                    deselected.append(colitem)
+                    continue
+            remaining.append(colitem)
 
     if deselected:
         config.hook.pytest_deselected(items=deselected)
         items[:] = remaining
 
+class BoolDict:
+    def __init__(self, mydict):
+        self._mydict = mydict
+    def __getitem__(self, name):
+        return name in self._mydict
+
+def matchmark(colitem, matchexpr):
+    return eval(matchexpr, {}, BoolDict(colitem.obj.__dict__))
+
+def pytest_configure(config):
+    if config.option.strict:
+        pytest.mark._config = config
+
 def skipbykeyword(colitem, keywordexpr):
     """ return True if they given keyword expression means to
         skip this collector/item.
     """
     if not keywordexpr:
         return
-    
+
     itemkeywords = getkeywords(colitem)
     for key in filter(None, keywordexpr.split()):
         eor = key[:1] == '-'
@@ -77,20 +119,36 @@ class MarkGenerator:
          @py.test.mark.slowtest
          def test_function():
             pass
-  
+
     will set a 'slowtest' :class:`MarkInfo` object
     on the ``test_function`` object. """
 
     def __getattr__(self, name):
         if name[0] == "_":
             raise AttributeError(name)
+        if hasattr(self, '_config'):
+            self._check(name)
         return MarkDecorator(name)
 
+    def _check(self, name):
+        try:
+            if name in self._markers:
+                return
+        except AttributeError:
+            pass
+        self._markers = l = set()
+        for line in self._config.getini("markers"):
+            beginning = line.split(":", 1)
+            x = beginning[0].split("(", 1)[0]
+            l.add(x)
+        if name not in self._markers:
+            raise AttributeError("%r not a registered marker" % (name,))
+
 class MarkDecorator:
     """ A decorator for test functions and test classes.  When applied
     it will create :class:`MarkInfo` objects which may be
-    :ref:`retrieved by hooks as item keywords`  MarkDecorator instances
-    are usually created by writing::
+    :ref:`retrieved by hooks as item keywords <excontrolskip>`.
+    MarkDecorator instances are often created like this::
 
         mark1 = py.test.mark.NAME              # simple MarkDecorator
         mark2 = py.test.mark.NAME(name1=value) # parametrized MarkDecorator
@@ -133,8 +191,7 @@ class MarkDecorator:
                         holder = MarkInfo(self.markname, self.args, self.kwargs)
                         setattr(func, self.markname, holder)
                     else:
-                        holder.kwargs.update(self.kwargs)
-                        holder.args += self.args
+                        holder.add(self.args, self.kwargs)
                 return func
         kw = self.kwargs.copy()
         kw.update(kwargs)
@@ -150,27 +207,20 @@ class MarkInfo:
         self.args = args
         #: keyword argument dictionary, empty if nothing specified
         self.kwargs = kwargs
+        self._arglist = [(args, kwargs.copy())]
 
     def __repr__(self):
         return "<MarkInfo %r args=%r kwargs=%r>" % (
-                self._name, self.args, self.kwargs)
+                self.name, self.args, self.kwargs)
+
+    def add(self, args, kwargs):
+        """ add a MarkInfo with the given args and kwargs. """
+        self._arglist.append((args, kwargs))
+        self.args += args
+        self.kwargs.update(kwargs)
+
+    def __iter__(self):
+        """ yield MarkInfo objects each relating to a marking-call. """
+        for args, kwargs in self._arglist:
+            yield MarkInfo(self.name, args, kwargs)
 
-def pytest_itemcollected(item):
-    if not isinstance(item, pytest.Function):
-        return
-    try:
-        func = item.obj.__func__
-    except AttributeError:
-        func = getattr(item.obj, 'im_func', item.obj)
-    pyclasses = (pytest.Class, pytest.Module)
-    for node in item.listchain():
-        if isinstance(node, pyclasses):
-            marker = getattr(node.obj, 'pytestmark', None)
-            if marker is not None:
-                if isinstance(marker, list):
-                    for mark in marker:
-                        mark(func)
-                else:
-                    marker(func)
-        node = node.parent
-    item.keywords.update(py.builtin._getfuncdict(func))

_pytest/monkeypatch.py

@@ -13,9 +13,10 @@ def pytest_funcarg__monkeypatch(request):
         monkeypatch.setenv(name, value, prepend=False)
         monkeypatch.delenv(name, value, raising=True)
         monkeypatch.syspath_prepend(path)
+        monkeypatch.chdir(path)
 
-    All modifications will be undone when the requesting
-    test function finished its execution.  The ``raising``
+    All modifications will be undone after the requesting
+    test function has finished. The ``raising``
     parameter determines if a KeyError or AttributeError
     will be raised if the set/deletion operation has no target.
     """
@@ -30,6 +31,7 @@ class monkeypatch:
     def __init__(self):
         self._setattr = []
         self._setitem = []
+        self._cwd = None
 
     def setattr(self, obj, name, value, raising=True):
         """ set attribute ``name`` on ``obj`` to ``value``, by default
@@ -83,6 +85,17 @@ class monkeypatch:
             self._savesyspath = sys.path[:]
         sys.path.insert(0, str(path))
 
+    def chdir(self, path):
+        """ change the current working directory to the specified path
+        path can be a string or a py.path.local object
+        """
+        if self._cwd is None:
+            self._cwd = os.getcwd()
+        if hasattr(path, "chdir"):
+            path.chdir()
+        else:
+            os.chdir(path)
+
     def undo(self):
         """ undo previous changes.  This call consumes the
         undo stack.  Calling it a second time has no effect unless
@@ -95,9 +108,17 @@ class monkeypatch:
         self._setattr[:] = []
         for dictionary, name, value in self._setitem:
             if value is notset:
-                del dictionary[name]
+                try:
+                    del dictionary[name]
+                except KeyError:
+                    pass # was already deleted, so we have the desired state
             else:
                 dictionary[name] = value
         self._setitem[:] = []
         if hasattr(self, '_savesyspath'):
             sys.path[:] = self._savesyspath
+            del self._savesyspath
+
+        if self._cwd is not None:
+            os.chdir(self._cwd)
+            self._cwd = None

_pytest/nose.py

@@ -13,6 +13,7 @@ def pytest_runtest_makereport(__multicall__, item, call):
             call.excinfo = call2.excinfo
 
 
+@pytest.mark.trylast
 def pytest_runtest_setup(item):
     if isinstance(item, (pytest.Function)):
         if isinstance(item.parent, pytest.Generator):

_pytest/pastebin.py

@@ -38,7 +38,11 @@ def pytest_unconfigure(config):
         del tr._tw.__dict__['write']
 
 def getproxy():
-    return py.std.xmlrpclib.ServerProxy(url.xmlrpc).pastes
+    if sys.version_info < (3, 0):
+        from xmlrpclib import ServerProxy
+    else:
+        from xmlrpc.client import ServerProxy
+    return ServerProxy(url.xmlrpc).pastes
 
 def pytest_terminal_summary(terminalreporter):
     if terminalreporter.config.option.pastebin != "failed":

_pytest/pdb.py

@@ -19,11 +19,13 @@ def pytest_configure(config):
 class pytestPDB:
     """ Pseudo PDB that defers to the real pdb. """
     item = None
+    collector = None
 
     def set_trace(self):
         """ invoke PDB set_trace debugging, dropping any IO capturing. """
         frame = sys._getframe().f_back
-        item = getattr(self, 'item', None)
+        item = self.item or self.collector
+
         if item is not None:
             capman = item.config.pluginmanager.getplugin("capturemanager")
             out, err = capman.suspendcapture()
@@ -38,6 +40,14 @@ def pdbitem(item):
     pytestPDB.item = item
 pytest_runtest_setup = pytest_runtest_call = pytest_runtest_teardown = pdbitem
 
+@pytest.mark.tryfirst
+def pytest_make_collect_report(__multicall__, collector):
+    try:
+        pytestPDB.collector = collector
+        return __multicall__.execute()
+    finally:
+        pytestPDB.collector = None
+
 def pytest_runtest_makereport():
     pytestPDB.item = None
     
@@ -52,12 +62,21 @@ class PdbInvoke:
         if "xfail" in rep.keywords:
             return rep
         # we assume that the above execute() suspended capturing
-        tw = py.io.TerminalWriter()
+        # XXX we re-use the TerminalReporter's terminalwriter
+        # because this seems to avoid some encoding related troubles
+        # for not completely clear reasons.
+        tw = item.config.pluginmanager.getplugin("terminalreporter")._tw
         tw.line()
         tw.sep(">", "traceback")
         rep.toterminal(tw)
         tw.sep(">", "entering PDB")
-        post_mortem(call.excinfo._excinfo[2])
+        # A doctest.UnexpectedException is not useful for post_mortem.
+        # Use the underlying exception instead:
+        if isinstance(call.excinfo.value, py.std.doctest.UnexpectedException):
+            tb = call.excinfo.value.exc_info[2]
+        else:
+            tb = call.excinfo._excinfo[2]
+        post_mortem(tb)
         rep._pdbshown = True
         return rep
 

_pytest/pytester.py

@@ -6,7 +6,7 @@ import re
 import inspect
 import time
 from fnmatch import fnmatch
-from _pytest.session import Session
+from _pytest.main import Session, EXIT_OK
 from py.builtin import print_
 from _pytest.core import HookRelay
 
@@ -25,6 +25,7 @@ def pytest_configure(config):
         _pytest_fullpath
     except NameError:
         _pytest_fullpath = os.path.abspath(pytest.__file__.rstrip("oc"))
+        _pytest_fullpath = _pytest_fullpath.replace("$py.class", ".py")
 
 def pytest_funcarg___pytest(request):
     return PytestArg(request)
@@ -236,13 +237,14 @@ class TmpTestdir:
     def _makefile(self, ext, args, kwargs):
         items = list(kwargs.items())
         if args:
-            source = "\n".join(map(str, args)) + "\n"
+            source = py.builtin._totext("\n").join(
+                map(py.builtin._totext, args)) + py.builtin._totext("\n")
             basename = self.request.function.__name__
             items.insert(0, (basename, source))
         ret = None
         for name, value in items:
             p = self.tmpdir.join(name).new(ext=ext)
-            source = str(py.code.Source(value)).lstrip()
+            source = py.builtin._totext(py.code.Source(value)).lstrip()
             p.write(source.encode("utf-8"), "wb")
             if ret is None:
                 ret = p
@@ -291,13 +293,19 @@ class TmpTestdir:
         assert '::' not in str(arg)
         p = py.path.local(arg)
         x = session.fspath.bestrelpath(p)
-        return session.perform_collect([x], genitems=False)[0]
+        config.hook.pytest_sessionstart(session=session)
+        res = session.perform_collect([x], genitems=False)[0]
+        config.hook.pytest_sessionfinish(session=session, exitstatus=EXIT_OK)
+        return res
 
     def getpathnode(self, path):
-        config = self.parseconfig(path)
+        config = self.parseconfigure(path)
         session = Session(config)
         x = session.fspath.bestrelpath(path)
-        return session.perform_collect([x], genitems=False)[0]
+        config.hook.pytest_sessionstart(session=session)
+        res = session.perform_collect([x], genitems=False)[0]
+        config.hook.pytest_sessionfinish(session=session, exitstatus=EXIT_OK)
+        return res
 
     def genitems(self, colitems):
         session = colitems[0].session
@@ -306,14 +314,6 @@ class TmpTestdir:
             result.extend(session.genitems(colitem))
         return result
 
-    def inline_genitems(self, *args):
-        #config = self.parseconfig(*args)
-        config = self.parseconfigure(*args)
-        rec = self.getreportrecorder(config)
-        session = Session(config)
-        session.perform_collect()
-        return session.items, rec
-
     def runitem(self, source):
         # used from runner functional tests
         item = self.getitem(source)
@@ -334,62 +334,57 @@ class TmpTestdir:
         l = list(args) + [p]
         reprec = self.inline_run(*l)
         reports = reprec.getreports("pytest_runtest_logreport")
-        assert len(reports) == 1, reports
-        return reports[0]
+        assert len(reports) == 3, reports # setup/call/teardown
+        return reports[1]
+
+    def inline_genitems(self, *args):
+        return self.inprocess_run(list(args) + ['--collectonly'])
 
     def inline_run(self, *args):
-        args = ("-s", ) + args # otherwise FD leakage
-        config = self.parseconfig(*args)
-        reprec = self.getreportrecorder(config)
-        #config.pluginmanager.do_configure(config)
-        config.hook.pytest_cmdline_main(config=config)
-        #config.pluginmanager.do_unconfigure(config)
-        return reprec
+        items, rec = self.inprocess_run(args)
+        return rec
 
-    def config_preparse(self):
-        config = self.Config()
-        for plugin in self.plugins:
-            if isinstance(plugin, str):
-                config.pluginmanager.import_plugin(plugin)
-            else:
-                if isinstance(plugin, dict):
-                    plugin = PseudoPlugin(plugin)
-                if not config.pluginmanager.isregistered(plugin):
-                    config.pluginmanager.register(plugin)
-        return config
+    def inprocess_run(self, args, plugins=None):
+        rec = []
+        items = []
+        class Collect:
+            def pytest_configure(x, config):
+                rec.append(self.getreportrecorder(config))
+            def pytest_itemcollected(self, item):
+                items.append(item)
+        if not plugins:
+            plugins = []
+        plugins.append(Collect())
+        ret = self.pytestmain(list(args), plugins=[Collect()])
+        reprec = rec[0]
+        reprec.ret = ret
+        assert len(rec) == 1
+        return items, reprec
 
     def parseconfig(self, *args):
-        if not args:
-            args = (self.tmpdir,)
-        config = self.config_preparse()
-        args = list(args)
+        args = [str(x) for x in args]
         for x in args:
             if str(x).startswith('--basetemp'):
                 break
         else:
             args.append("--basetemp=%s" % self.tmpdir.dirpath('basetemp'))
-        config.parse(args)
+        import _pytest.core
+        config = _pytest.core._prepareconfig(args, self.plugins)
+        # the in-process pytest invocation needs to avoid leaking FDs
+        # so we register a "reset_capturings" callmon the capturing manager
+        # and make sure it gets called
+        config._cleanup.append(
+            config.pluginmanager.getplugin("capturemanager").reset_capturings)
+        import _pytest.config
+        self.request.addfinalizer(
+            lambda: _pytest.config.pytest_unconfigure(config))
         return config
 
-    def reparseconfig(self, args=None):
-        """ this is used from tests that want to re-invoke parse(). """
-        if not args:
-            args = [self.tmpdir]
-        oldconfig = getattr(py.test, 'config', None)
-        try:
-            c = py.test.config = self.Config()
-            c.basetemp = py.path.local.make_numbered_dir(prefix="reparse",
-                keep=0, rootdir=self.tmpdir, lock_timeout=None)
-            c.parse(args)
-            return c
-        finally:
-            py.test.config = oldconfig
-
     def parseconfigure(self, *args):
         config = self.parseconfig(*args)
         config.pluginmanager.do_configure(config)
         self.request.addfinalizer(lambda:
-            config.pluginmanager.do_unconfigure(config))
+        config.pluginmanager.do_unconfigure(config))
         return config
 
     def getitem(self,  source, funcname="test_func"):
@@ -409,7 +404,6 @@ class TmpTestdir:
             self.makepyfile(__init__ = "#")
         self.config = config = self.parseconfigure(path, *configargs)
         node = self.getnode(config, path)
-        #config.pluginmanager.do_unconfigure(config)
         return node
 
     def collect_by_name(self, modcol, name):
@@ -426,9 +420,16 @@ class TmpTestdir:
         return py.std.subprocess.Popen(cmdargs, stdout=stdout, stderr=stderr, **kw)
 
     def pytestmain(self, *args, **kwargs):
-        ret = pytest.main(*args, **kwargs)
-        if ret == 2:
-            raise KeyboardInterrupt()
+        class ResetCapturing:
+            @pytest.mark.trylast
+            def pytest_unconfigure(self, config):
+                capman = config.pluginmanager.getplugin("capturemanager")
+                capman.reset_capturings()
+        plugins = kwargs.setdefault("plugins", [])
+        rc = ResetCapturing()
+        plugins.append(rc)
+        return pytest.main(*args, **kwargs)
+
     def run(self, *cmdargs):
         return self._run(*cmdargs)
 
@@ -515,6 +516,10 @@ class TmpTestdir:
 
     def spawn(self, cmd, expect_timeout=10.0):
         pexpect = py.test.importorskip("pexpect", "2.4")
+        if hasattr(sys, 'pypy_version_info') and '64' in py.std.platform.machine():
+            pytest.skip("pypy-64 bit not supported")
+        if sys.platform == "darwin":
+            pytest.xfail("pexpect does not work reliably on darwin?!")
         logfile = self.tmpdir.join("spawn.out")
         child = pexpect.spawn(cmd, logfile=logfile.open("w"))
         child.timeout = expect_timeout
@@ -527,10 +532,6 @@ def getdecoded(out):
             return "INTERNAL not-utf8-decodeable, truncated string:\n%s" % (
                     py.io.saferepr(out),)
 
-class PseudoPlugin:
-    def __init__(self, vars):
-        self.__dict__.update(vars)
-
 class ReportRecorder(object):
     def __init__(self, hook):
         self.hook = hook
@@ -552,10 +553,17 @@ class ReportRecorder(object):
     def getreports(self, names="pytest_runtest_logreport pytest_collectreport"):
         return [x.report for x in self.getcalls(names)]
 
-    def matchreport(self, inamepart="", names="pytest_runtest_logreport pytest_collectreport", when=None):
+    def matchreport(self, inamepart="",
+        names="pytest_runtest_logreport pytest_collectreport", when=None):
         """ return a testreport whose dotted import path matches """
         l = []
         for rep in self.getreports(names=names):
+            try:
+                if not when and rep.when != "call" and rep.passed:
+                    # setup/teardown passing reports - let's ignore those
+                    continue
+            except AttributeError:
+                pass
             if when and getattr(rep, 'when', None) != when:
                 continue
             if not inamepart or inamepart in rep.nodeid.split("::"):

_pytest/python.py

@@ -4,6 +4,7 @@ import inspect
 import sys
 import pytest
 from py._code.code import TerminalRepr
+from _pytest.monkeypatch import monkeypatch
 
 import _pytest
 cutdir = py.path.local(_pytest.__file__).dirpath()
@@ -26,6 +27,24 @@ def pytest_cmdline_main(config):
         showfuncargs(config)
         return 0
 
+
+def pytest_generate_tests(metafunc):
+    try:
+        param = metafunc.function.parametrize
+    except AttributeError:
+        return
+    for p in param:
+        metafunc.parametrize(*p.args, **p.kwargs)
+
+def pytest_configure(config):
+    config.addinivalue_line("markers",
+        "parametrize(argnames, argvalues): call a test function multiple "
+        "times passing in multiple different argument value sets. Example: "
+        "@parametrize('arg1', [1,2]) would lead to two calls of the decorated "
+        "test function, one with arg1=1 and another with arg1=2."
+    )
+
+
 @pytest.mark.trylast
 def pytest_namespace():
     raises.Exception = pytest.fail.Exception
@@ -70,10 +89,13 @@ def pytest_pycollect_makeitem(__multicall__, collector, name, obj):
     res = __multicall__.execute()
     if res is not None:
         return res
-    if collector._istestclasscandidate(name, obj):
+    if inspect.isclass(obj):
         #if hasattr(collector.obj, 'unittest'):
         #    return # we assume it's a mixin class for a TestCase derived one
-        return Class(name, parent=collector)
+        if collector.classnamefilter(name):
+            if not hasinit(obj):
+                Class = collector._getcustomclass("Class")
+                return Class(name, parent=collector)
     elif collector.funcnamefilter(name) and hasattr(obj, '__call__'):
         if is_generator(obj):
             return Generator(name, parent=collector)
@@ -135,6 +157,7 @@ class PyobjMixin(object):
             obj = obj.place_as
 
         self._fslineno = py.code.getfslineno(obj)
+        assert isinstance(self._fslineno[1], int), obj
         return self._fslineno
 
     def reportinfo(self):
@@ -152,6 +175,7 @@ class PyobjMixin(object):
         else:
             fspath, lineno = self._getfslineno()
             modpath = self.getmodpath()
+        assert isinstance(lineno, int)
         return fspath, lineno, modpath
 
 class PyCollectorMixin(PyobjMixin, pytest.Collector):
@@ -160,7 +184,7 @@ class PyCollectorMixin(PyobjMixin, pytest.Collector):
         for prefix in self.config.getini("python_functions"):
             if name.startswith(prefix):
                 return True
-        
+
     def classnamefilter(self, name):
         for prefix in self.config.getini("python_classes"):
             if name.startswith(prefix):
@@ -193,18 +217,11 @@ class PyCollectorMixin(PyobjMixin, pytest.Collector):
         return self.ihook.pytest_pycollect_makeitem(
             collector=self, name=name, obj=obj)
 
-    def _istestclasscandidate(self, name, obj):
-        if self.classnamefilter(name) and \
-           inspect.isclass(obj):
-            if hasinit(obj):
-                # XXX WARN
-                return False
-            return True
-
     def _genfunctions(self, name, funcobj):
         module = self.getparent(Module).obj
         clscol = self.getparent(Class)
         cls = clscol and clscol.obj or None
+        transfer_markers(funcobj, cls, module)
         metafunc = Metafunc(funcobj, config=self.config,
             cls=cls, module=module)
         gentesthook = self.config.hook.pytest_generate_tests
@@ -213,6 +230,7 @@ class PyCollectorMixin(PyobjMixin, pytest.Collector):
             extra.append(cls())
         plugins = self.getplugins() + extra
         gentesthook.pcall(plugins, metafunc=metafunc)
+        Function = self._getcustomclass("Function")
         if not metafunc._calls:
             return Function(name, parent=self)
         l = []
@@ -223,6 +241,20 @@ class PyCollectorMixin(PyobjMixin, pytest.Collector):
             l.append(function)
         return l
 
+def transfer_markers(funcobj, cls, mod):
+    # XXX this should rather be code in the mark plugin or the mark
+    # plugin should merge with the python plugin.
+    for holder in (cls, mod):
+        try:
+            pytestmark = holder.pytestmark
+        except AttributeError:
+            continue
+        if isinstance(pytestmark, list):
+            for mark in pytestmark:
+                mark(funcobj)
+        else:
+            pytestmark(funcobj)
+
 class Module(pytest.File, PyCollectorMixin):
     def _getobj(self):
         return self._memoizedcall('_obj', self._importtestmodule)
@@ -242,7 +274,8 @@ class Module(pytest.File, PyCollectorMixin):
                 "  %s\n"
                 "which is not the same as the test file we want to collect:\n"
                 "  %s\n"
-                "HINT: use a unique basename for your test file modules"
+                "HINT: remove __pycache__ / .pyc files and/or use a "
+                "unique basename for your test file modules"
                  % e.args
             )
         #print "imported test module", mod
@@ -272,7 +305,7 @@ class Module(pytest.File, PyCollectorMixin):
 class Class(PyCollectorMixin, pytest.Collector):
 
     def collect(self):
-        return [Instance(name="()", parent=self)]
+        return [self._getcustomclass("Instance")(name="()", parent=self)]
 
     def setup(self):
         setup_class = getattr(self.obj, 'setup_class', None)
@@ -297,18 +330,19 @@ class Instance(PyCollectorMixin, pytest.Collector):
 class FunctionMixin(PyobjMixin):
     """ mixin for the code common to Function and Generator.
     """
-
     def setup(self):
         """ perform setup for this test function. """
-        if inspect.ismethod(self.obj):
-            name = 'setup_method'
-        else:
-            name = 'setup_function'
-        if isinstance(self.parent, Instance):
+        if hasattr(self, '_preservedparent'):
+            obj = self._preservedparent
+        elif isinstance(self.parent, Instance):
             obj = self.parent.newinstance()
             self.obj = self._getobj()
         else:
             obj = self.parent.obj
+        if inspect.ismethod(self.obj):
+            name = 'setup_method'
+        else:
+            name = 'setup_function'
         setup_func_or_method = getattr(obj, name, None)
         if setup_func_or_method is not None:
             setup_func_or_method(self.obj)
@@ -371,13 +405,15 @@ class FuncargLookupErrorRepr(TerminalRepr):
         tw.line()
         tw.line("%s:%d" % (self.filename, self.firstlineno+1))
 
+
 class Generator(FunctionMixin, PyCollectorMixin, pytest.Collector):
     def collect(self):
         # test generators are seen as collectors but they also
         # invoke setup/teardown on popular request
         # (induced by the common "test_*" naming shared with normal tests)
-        self.config._setupstate.prepare(self)
-
+        self.session._setupstate.prepare(self)
+        # see FunctionMixin.setup and test_setupstate_is_preserved_134
+        self._preservedparent = self.parent.obj
         l = []
         seen = {}
         for i, x in enumerate(self.obj()):
@@ -391,7 +427,7 @@ class Generator(FunctionMixin, PyCollectorMixin, pytest.Collector):
             if name in seen:
                 raise ValueError("%r generated tests with non-unique name %r" %(self, name))
             seen[name] = True
-            l.append(Function(name, self, args=args, callobj=call))
+            l.append(self.Function(name, self, args=args, callobj=call))
         return l
 
     def getcallargs(self, obj):
@@ -426,6 +462,7 @@ class Function(FunctionMixin, pytest.Item):
                 "yielded functions (deprecated) cannot have funcargs")
         else:
             if callspec is not None:
+                self.callspec = callspec
                 self.funcargs = callspec.funcargs or {}
                 self._genid = callspec.id
                 if hasattr(callspec, "param"):
@@ -484,10 +521,11 @@ def hasinit(obj):
             return True
 
 
-def getfuncargnames(function):
+def getfuncargnames(function, startindex=None):
     # XXX merge with main.py's varnames
     argnames = py.std.inspect.getargs(py.code.getrawcode(function))[0]
-    startindex = py.std.inspect.ismethod(function) and 1 or 0
+    if startindex is None:
+        startindex = py.std.inspect.ismethod(function) and 1 or 0
     defaults = getattr(function, 'func_defaults',
                        getattr(function, '__defaults__', None)) or ()
     numdefaults = len(defaults)
@@ -501,47 +539,137 @@ def fillfuncargs(function):
     request._fillfuncargs()
 
 _notexists = object()
-class CallSpec:
-    def __init__(self, funcargs, id, param):
-        self.funcargs = funcargs
-        self.id = id
+
+class CallSpec2(object):
+    def __init__(self, metafunc):
+        self.metafunc = metafunc
+        self.funcargs = {}
+        self._idlist = []
+        self.params = {}
+        self._globalid = _notexists
+        self._globalid_args = set()
+        self._globalparam = _notexists
+
+    def copy(self, metafunc):
+        cs = CallSpec2(self.metafunc)
+        cs.funcargs.update(self.funcargs)
+        cs.params.update(self.params)
+        cs._idlist = list(self._idlist)
+        cs._globalid = self._globalid
+        cs._globalid_args = self._globalid_args
+        cs._globalparam = self._globalparam
+        return cs
+
+    def _checkargnotcontained(self, arg):
+        if arg in self.params or arg in self.funcargs:
+            raise ValueError("duplicate %r" %(arg,))
+
+    def getparam(self, name):
+        try:
+            return self.params[name]
+        except KeyError:
+            if self._globalparam is _notexists:
+                raise ValueError(name)
+            return self._globalparam
+
+    @property
+    def id(self):
+        return "-".join(map(str, filter(None, self._idlist)))
+
+    def setmulti(self, valtype, argnames, valset, id):
+        for arg,val in zip(argnames, valset):
+            self._checkargnotcontained(arg)
+            getattr(self, valtype)[arg] = val
+        self._idlist.append(id)
+
+    def setall(self, funcargs, id, param):
+        for x in funcargs:
+            self._checkargnotcontained(x)
+        self.funcargs.update(funcargs)
+        if id is not _notexists:
+            self._idlist.append(id)
         if param is not _notexists:
-            self.param = param
-    def __repr__(self):
-        return "<CallSpec id=%r param=%r funcargs=%r>" %(
-            self.id, getattr(self, 'param', '?'), self.funcargs)
+            assert self._globalparam is _notexists
+            self._globalparam = param
+
 
 class Metafunc:
     def __init__(self, function, config=None, cls=None, module=None):
         self.config = config
         self.module = module
         self.function = function
-        self.funcargnames = getfuncargnames(function)
+        self.funcargnames = getfuncargnames(function,
+                                            startindex=int(cls is not None))
         self.cls = cls
         self.module = module
         self._calls = []
         self._ids = py.builtin.set()
 
+    def parametrize(self, argnames, argvalues, indirect=False, ids=None):
+        """ Add new invocations to the underlying test function using the list
+        of argvalues for the given argnames.  Parametrization is performed
+        during the collection phase.  If you need to setup expensive resources
+        you may pass indirect=True and implement a funcarg factory which can
+        perform the expensive setup just before a test is actually run.
+
+        :arg argnames: an argument name or a list of argument names
+
+        :arg argvalues: a list of values for the argname or a list of tuples of
+            values for the list of argument names.
+
+        :arg indirect: if True each argvalue corresponding to an argument will
+            be passed as request.param to its respective funcarg factory so
+            that it can perform more expensive setups during the setup phase of
+            a test rather than at collection time.
+
+        :arg ids: list of string ids each corresponding to the argvalues so
+            that they are part of the test id. If no ids are provided they will
+            be generated automatically from the argvalues.
+        """
+        if not isinstance(argnames, (tuple, list)):
+            argnames = (argnames,)
+            argvalues = [(val,) for val in argvalues]
+        if not indirect:
+            #XXX should we also check for the opposite case?
+            for arg in argnames:
+                if arg not in self.funcargnames:
+                    raise ValueError("%r has no argument %r" %(self.function, arg))
+        valtype = indirect and "params" or "funcargs"
+        if not ids:
+            idmaker = IDMaker()
+            ids = list(map(idmaker, argvalues))
+        newcalls = []
+        for callspec in self._calls or [CallSpec2(self)]:
+            for i, valset in enumerate(argvalues):
+                assert len(valset) == len(argnames)
+                newcallspec = callspec.copy(self)
+                newcallspec.setmulti(valtype, argnames, valset, ids[i])
+                newcalls.append(newcallspec)
+        self._calls = newcalls
+
     def addcall(self, funcargs=None, id=_notexists, param=_notexists):
-        """ add a new call to the underlying test function during the
-        collection phase of a test run.
-        
+        """ (deprecated, use parametrize) Add a new call to the underlying
+        test function during the collection phase of a test run.  Note that
+        request.addcall() is called during the test collection phase prior and
+        independently to actual test execution.  You should only use addcall()
+        if you need to specify multiple arguments of a test function.
+
         :arg funcargs: argument keyword dictionary used when invoking
             the test function.
-        
-        :arg id: used for reporting and identification purposes.  If you
-            don't supply an `id` the length of the currently
-            list of calls to the test function will be used.
 
-        :arg param: will be exposed to a later funcarg factory invocation
-            through the ``request.param`` attribute.  Setting it (instead of
-            directly providing a ``funcargs`` ditionary) is called
-            *indirect parametrization*.  Indirect parametrization is
-            preferable if test values are expensive to setup or can
-            only be created after certain fixtures or test-run related
-            initialization code has been run.
+        :arg id: used for reporting and identification purposes.  If you
+            don't supply an `id` an automatic unique id will be generated.
+
+        :arg param: a parameter which will be exposed to a later funcarg factory
+            invocation through the ``request.param`` attribute.
         """
         assert funcargs is None or isinstance(funcargs, dict)
+        if funcargs is not None:
+            for name in funcargs:
+                if name not in self.funcargnames:
+                    pytest.fail("funcarg %r not used in this function." % name)
+        else:
+            funcargs = {}
         if id is None:
             raise ValueError("id=None not allowed")
         if id is _notexists:
@@ -550,16 +678,37 @@ class Metafunc:
         if id in self._ids:
             raise ValueError("duplicate id %r" % id)
         self._ids.add(id)
-        self._calls.append(CallSpec(funcargs, id, param))
+
+        cs = CallSpec2(self)
+        cs.setall(funcargs, id, param)
+        self._calls.append(cs)
+
+class IDMaker:
+    def __init__(self):
+        self.counter = 0
+    def __call__(self, valset):
+        l = []
+        for val in valset:
+            if not isinstance(val, (int, str)):
+                val = "."+str(self.counter)
+            self.counter += 1
+            l.append(str(val))
+        return "-".join(l)
 
 class FuncargRequest:
-    """ A request for function arguments from a test function. """
+    """ A request for function arguments from a test function.
+
+        Note that there is an optional ``param`` attribute in case
+        there was an invocation to metafunc.addcall(param=...).
+        If no such call was done in a ``pytest_generate_tests``
+        hook, the attribute will not be present.
+    """
     _argprefix = "pytest_funcarg__"
     _argname = None
 
     class LookupError(LookupError):
         """ error on performing funcarg request. """
-   
+
     def __init__(self, pyfuncitem):
         self._pyfuncitem = pyfuncitem
         if hasattr(pyfuncitem, '_requestparam'):
@@ -587,7 +736,7 @@ class FuncargRequest:
     def module(self):
         """ module where the test function was collected. """
         return self._pyfuncitem.getparent(pytest.Module).obj
-       
+
     @property
     def cls(self):
         """ class (can be None) where the test function was collected. """
@@ -603,7 +752,7 @@ class FuncargRequest:
     def config(self):
         """ the pytest config object associated with this request. """
         return self._pyfuncitem.config
-    
+
     @property
     def fspath(self):
         """ the file system path of the test module which collected this test. """
@@ -620,7 +769,7 @@ class FuncargRequest:
 
 
     def applymarker(self, marker):
-        """ apply a marker to a single test function invocation.
+        """ Apply a marker to a single test function invocation.
         This method is useful if you don't want to have a keyword/marker
         on all function invocations.
 
@@ -632,7 +781,7 @@ class FuncargRequest:
         self._pyfuncitem.keywords[marker.markname] = marker
 
     def cached_setup(self, setup, teardown=None, scope="module", extrakey=None):
-        """ return a testing resource managed by ``setup`` &
+        """ Return a testing resource managed by ``setup`` &
         ``teardown`` calls.  ``scope`` and ``extrakey`` determine when the
         ``teardown`` function will be called so that subsequent calls to
         ``setup`` would recreate the resource.
@@ -681,11 +830,18 @@ class FuncargRequest:
             self._raiselookupfailed(argname)
         funcargfactory = self._name2factory[argname].pop()
         oldarg = self._currentarg
-        self._currentarg = argname
+        mp = monkeypatch()
+        mp.setattr(self, '_currentarg', argname)
+        try:
+            param = self._pyfuncitem.callspec.getparam(argname)
+        except (AttributeError, ValueError):
+            pass
+        else:
+            mp.setattr(self, 'param', param, raising=False)
         try:
             self._funcargs[argname] = res = funcargfactory(request=self)
         finally:
-            self._currentarg = oldarg
+            mp.undo()
         return res
 
     def _getscopeitem(self, scope):
@@ -709,7 +865,7 @@ class FuncargRequest:
 
     def _addfinalizer(self, finalizer, scope):
         colitem = self._getscopeitem(scope)
-        self.config._setupstate.addfinalizer(
+        self._pyfuncitem.session._setupstate.addfinalizer(
             finalizer=finalizer, colitem=colitem)
 
     def __repr__(self):
@@ -730,8 +886,10 @@ class FuncargRequest:
         raise self.LookupError(msg)
 
 def showfuncargs(config):
-    from _pytest.session import Session
-    session = Session(config)
+    from _pytest.main import wrap_session
+    return wrap_session(config, _showfuncargs_main)
+
+def _showfuncargs_main(config, session):
     session.perform_collect()
     if session.items:
         plugins = session.items[0].getplugins()
@@ -777,7 +935,7 @@ def getlocation(function, curdir):
 def raises(ExpectedException, *args, **kwargs):
     """ assert that a code block/function call raises @ExpectedException
     and raise a failure exception otherwise.
-        
+
     If using Python 2.5 or above, you may use this function as a
     context manager::
 
@@ -798,9 +956,8 @@ def raises(ExpectedException, *args, **kwargs):
         >>> raises(ZeroDivisionError, f, x=0)
         <ExceptionInfo ...>
 
-    A third possibility is to use a string which which will
-    be executed::
-    
+    A third possibility is to use a string to be executed::
+
         >>> raises(ZeroDivisionError, "f(0)")
         <ExceptionInfo ...>
     """
@@ -849,3 +1006,4 @@ class RaisesContext(object):
             pytest.fail("DID NOT RAISE")
         self.excinfo.__init__(tp)
         return issubclass(self.excinfo.type, self.ExpectedException)
+

_pytest/recwarn.py

@@ -8,6 +8,9 @@ def pytest_funcarg__recwarn(request):
 
     * ``pop(category=None)``: return last warning matching the category.
     * ``clear()``: clear list of warnings
+   
+    See http://docs.python.org/library/warnings.html for information
+    on warning categories.
     """
     if sys.version_info >= (2,7):
         import warnings

_pytest/resultlog.py

@@ -1,12 +1,12 @@
 """ (disabled by default) create result information in a plain text file. """
 
 import py
-from py.builtin import print_
 
 def pytest_addoption(parser):
-    group = parser.getgroup("resultlog", "resultlog plugin options")
-    group.addoption('--resultlog', action="store", dest="resultlog", metavar="path", default=None,
-           help="path for machine-readable result log.")
+    group = parser.getgroup("terminal reporting", "resultlog plugin options")
+    group.addoption('--resultlog', action="store", dest="resultlog",
+        metavar="path", default=None,
+        help="path for machine-readable result log.")
 
 def pytest_configure(config):
     resultlog = config.option.resultlog
@@ -52,9 +52,9 @@ class ResultLog(object):
         self.logfile = logfile # preferably line buffered
 
     def write_log_entry(self, testpath, lettercode, longrepr):
-        print_("%s %s" % (lettercode, testpath), file=self.logfile)
+        py.builtin.print_("%s %s" % (lettercode, testpath), file=self.logfile)
         for line in longrepr.splitlines():
-            print_(" %s" % line, file=self.logfile)
+            py.builtin.print_(" %s" % line, file=self.logfile)
 
     def log_outcome(self, report, lettercode, longrepr):
         testpath = getattr(report, 'nodeid', None)
@@ -63,6 +63,8 @@ class ResultLog(object):
         self.write_log_entry(testpath, lettercode, longrepr)
 
     def pytest_runtest_logreport(self, report):
+        if report.when != "call" and report.passed:
+            return
         res = self.config.hook.pytest_report_teststatus(report=report)
         code = res[1]
         if code == 'x':
@@ -89,5 +91,8 @@ class ResultLog(object):
             self.log_outcome(report, code, longrepr)
 
     def pytest_internalerror(self, excrepr):
-        path = excrepr.reprcrash.path
+        reprcrash = getattr(excrepr, 'reprcrash', None)
+        path = getattr(reprcrash, "path", None)
+        if path is None:
+            path = "cwd:%s" % py.path.local()
         self.write_log_entry(path, '!', str(excrepr))

_pytest/runner.py

@@ -1,6 +1,6 @@
 """ basic collect and runtest protocol implementations """
 
-import py, sys
+import py, sys, time
 from py._code.code import TerminalRepr
 
 def pytest_namespace():
@@ -14,53 +14,70 @@ def pytest_namespace():
 #
 # pytest plugin hooks
 
-# XXX move to pytest_sessionstart and fix py.test owns tests
-def pytest_configure(config):
-    config._setupstate = SetupState()
+def pytest_addoption(parser):
+    group = parser.getgroup("terminal reporting", "reporting", after="general")
+    group.addoption('--durations',
+         action="store", type="int", default=None, metavar="N",
+         help="show N slowest setup/test durations (N=0 for all)."),
 
-def pytest_sessionfinish(session, exitstatus):
-    if hasattr(session.config, '_setupstate'):
-        hook = session.config.hook
-        rep = hook.pytest__teardown_final(session=session)
-        if rep:
-            hook.pytest__teardown_final_logerror(session=session, report=rep)
-            session.exitstatus = 1
+def pytest_terminal_summary(terminalreporter):
+    durations = terminalreporter.config.option.durations
+    if durations is None:
+        return
+    tr = terminalreporter
+    dlist = []
+    for replist in tr.stats.values():
+        for rep in replist:
+            if hasattr(rep, 'duration'):
+                dlist.append(rep)
+    if not dlist:
+        return
+    dlist.sort(key=lambda x: x.duration)
+    dlist.reverse()
+    if not durations:
+        tr.write_sep("=", "slowest test durations")
+    else:
+        tr.write_sep("=", "slowest %s test durations" % durations)
+        dlist = dlist[:durations]
+
+    for rep in dlist:
+        nodeid = rep.nodeid.replace("::()::", "::")
+        tr.write_line("%02.2fs %-8s %s" %
+            (rep.duration, rep.when, nodeid))
+
+def pytest_sessionstart(session):
+    session._setupstate = SetupState()
+def pytest_sessionfinish(session):
+    session._setupstate.teardown_all()
 
 class NodeInfo:
     def __init__(self, location):
         self.location = location
 
-def pytest_runtest_protocol(item):
+def pytest_runtest_protocol(item, nextitem):
     item.ihook.pytest_runtest_logstart(
         nodeid=item.nodeid, location=item.location,
     )
-    runtestprotocol(item)
+    runtestprotocol(item, nextitem=nextitem)
     return True
 
-def runtestprotocol(item, log=True):
+def runtestprotocol(item, log=True, nextitem=None):
     rep = call_and_report(item, "setup", log)
     reports = [rep]
     if rep.passed:
         reports.append(call_and_report(item, "call", log))
-    reports.append(call_and_report(item, "teardown", log))
+    reports.append(call_and_report(item, "teardown", log,
+        nextitem=nextitem))
     return reports
 
 def pytest_runtest_setup(item):
-    item.config._setupstate.prepare(item)
+    item.session._setupstate.prepare(item)
 
 def pytest_runtest_call(item):
     item.runtest()
 
-def pytest_runtest_teardown(item):
-    item.config._setupstate.teardown_exact(item)
-
-def pytest__teardown_final(session):
-    call = CallInfo(session.config._setupstate.teardown_all, when="teardown")
-    if call.excinfo:
-        ntraceback = call.excinfo.traceback .cut(excludepath=py._pydir)
-        call.excinfo.traceback = ntraceback.filter()
-        longrepr = call.excinfo.getrepr(funcargs=True)
-        return TeardownErrorReport(longrepr)
+def pytest_runtest_teardown(item, nextitem):
+    item.session._setupstate.teardown_exact(item, nextitem)
 
 def pytest_report_teststatus(report):
     if report.when in ("setup", "teardown"):
@@ -76,18 +93,18 @@ def pytest_report_teststatus(report):
 #
 # Implementation
 
-def call_and_report(item, when, log=True):
-    call = call_runtest_hook(item, when)
+def call_and_report(item, when, log=True, **kwds):
+    call = call_runtest_hook(item, when, **kwds)
     hook = item.ihook
     report = hook.pytest_runtest_makereport(item=item, call=call)
-    if log and (when == "call" or not report.passed):
+    if log:
         hook.pytest_runtest_logreport(report=report)
     return report
 
-def call_runtest_hook(item, when):
+def call_runtest_hook(item, when, **kwds):
     hookname = "pytest_runtest_" + when
     ihook = getattr(item.ihook, hookname)
-    return CallInfo(lambda: ihook(item=item), when=when)
+    return CallInfo(lambda: ihook(item=item, **kwds), when=when)
 
 class CallInfo:
     """ Result/Exception info a function invocation. """
@@ -97,12 +114,16 @@ class CallInfo:
         #: context of invocation: one of "setup", "call",
         #: "teardown", "memocollect"
         self.when = when
+        self.start = time.time()
         try:
-            self.result = func()
-        except KeyboardInterrupt:
-            raise
-        except:
-            self.excinfo = py.code.ExceptionInfo()
+            try:
+                self.result = func()
+            except KeyboardInterrupt:
+                raise
+            except:
+                self.excinfo = py.code.ExceptionInfo()
+        finally:
+            self.stop = time.time()
 
     def __repr__(self):
         if self.excinfo:
@@ -122,6 +143,10 @@ def getslaveinfoline(node):
         return s
 
 class BaseReport(object):
+
+    def __init__(self, **kw):
+        self.__dict__.update(kw)
+
     def toterminal(self, out):
         longrepr = self.longrepr
         if hasattr(self, 'node'):
@@ -141,6 +166,7 @@ class BaseReport(object):
 
 def pytest_runtest_makereport(item, call):
     when = call.when
+    duration = call.stop-call.start
     keywords = dict([(x,1) for x in item.keywords])
     excinfo = call.excinfo
     if not call.excinfo:
@@ -153,7 +179,7 @@ def pytest_runtest_makereport(item, call):
             longrepr = excinfo
         elif excinfo.errisinstance(py.test.skip.Exception):
             outcome = "skipped"
-            r = item._repr_failure_py(excinfo, "line").reprcrash
+            r = excinfo._getreprcrash()
             longrepr = (str(r.path), r.lineno, r.message)
         else:
             outcome = "failed"
@@ -162,14 +188,15 @@ def pytest_runtest_makereport(item, call):
             else: # exception in setup or teardown
                 longrepr = item._repr_failure_py(excinfo)
     return TestReport(item.nodeid, item.location,
-        keywords, outcome, longrepr, when)
+                      keywords, outcome, longrepr, when,
+                      duration=duration)
 
 class TestReport(BaseReport):
     """ Basic test report object (also used for setup and teardown calls if
     they fail).
     """
     def __init__(self, nodeid, location,
-            keywords, outcome, longrepr, when):
+            keywords, outcome, longrepr, when, sections=(), duration=0, **extra):
         #: normalized collection node id
         self.nodeid = nodeid
 
@@ -181,16 +208,25 @@ class TestReport(BaseReport):
         #: a name -> value dictionary containing all keywords and
         #: markers associated with a test invocation.
         self.keywords = keywords
-        
+
         #: test outcome, always one of "passed", "failed", "skipped".
         self.outcome = outcome
 
         #: None or a failure representation.
         self.longrepr = longrepr
-        
+
         #: one of 'setup', 'call', 'teardown' to indicate runtest phase.
         self.when = when
 
+        #: list of (secname, data) extra information which needs to
+        #: marshallable
+        self.sections = list(sections)
+
+        #: time it took to run just the test
+        self.duration = duration
+
+        self.__dict__.update(extra)
+
     def __repr__(self):
         return "<TestReport %r when=%r outcome=%r>" % (
             self.nodeid, self.when, self.outcome)
@@ -198,8 +234,10 @@ class TestReport(BaseReport):
 class TeardownErrorReport(BaseReport):
     outcome = "failed"
     when = "teardown"
-    def __init__(self, longrepr):
+    def __init__(self, longrepr, **extra):
         self.longrepr = longrepr
+        self.sections = []
+        self.__dict__.update(extra)
 
 def pytest_make_collect_report(collector):
     call = CallInfo(collector._memocollect, "memocollect")
@@ -221,11 +259,13 @@ def pytest_make_collect_report(collector):
         getattr(call, 'result', None))
 
 class CollectReport(BaseReport):
-    def __init__(self, nodeid, outcome, longrepr, result):
+    def __init__(self, nodeid, outcome, longrepr, result, sections=(), **extra):
         self.nodeid = nodeid
         self.outcome = outcome
         self.longrepr = longrepr
         self.result = result or []
+        self.sections = list(sections)
+        self.__dict__.update(extra)
 
     @property
     def location(self):
@@ -279,20 +319,22 @@ class SetupState(object):
         self._teardown_with_finalization(None)
         assert not self._finalizers
 
-    def teardown_exact(self, item):
-        if self.stack and item == self.stack[-1]:
+    def teardown_exact(self, item, nextitem):
+        needed_collectors = nextitem and nextitem.listchain() or []
+        self._teardown_towards(needed_collectors)
+
+    def _teardown_towards(self, needed_collectors):
+        while self.stack:
+            if self.stack == needed_collectors[:len(self.stack)]:
+                break
             self._pop_and_teardown()
-        else:
-            self._callfinalizers(item)
 
     def prepare(self, colitem):
         """ setup objects along the collector chain to the test-method
             and teardown previously setup objects."""
         needed_collectors = colitem.listchain()
-        while self.stack:
-            if self.stack == needed_collectors[:len(self.stack)]:
-                break
-            self._pop_and_teardown()
+        self._teardown_towards(needed_collectors)
+
         # check if the last collection node has raised an error
         for col in self.stack:
             if hasattr(col, '_prepare_exc'):

_pytest/skipping.py

@@ -1,6 +1,7 @@
 """ support for skip/xfail functions and markers. """
 
 import py, pytest
+import sys
 
 def pytest_addoption(parser):
     group = parser.getgroup("general")
@@ -8,6 +9,21 @@ def pytest_addoption(parser):
            action="store_true", dest="runxfail", default=False,
            help="run tests even if they are marked xfail")
 
+def pytest_configure(config):
+    config.addinivalue_line("markers",
+        "skipif(*conditions): skip the given test function if evaluation "
+        "of all conditions has a True value.  Evaluation happens within the "
+        "module global context. Example: skipif('sys.platform == \"win32\"') "
+        "skips the test if we are on the win32 platform. "
+    )
+    config.addinivalue_line("markers",
+        "xfail(*conditions, reason=None, run=True): mark the the test function "
+        "as an expected failure. Optionally specify a reason and run=False "
+        "if you don't even want to execute the test function. Any positional "
+        "condition strings will be evaluated (like with skipif) and if one is "
+        "False the marker will not be applied."
+    )
+
 def pytest_namespace():
     return dict(xfail=xfail)
 
@@ -32,9 +48,39 @@ class MarkEvaluator:
         return bool(self.holder)
     __nonzero__ = __bool__
 
+    def wasvalid(self):
+        return not hasattr(self, 'exc')
+
     def istrue(self):
+        try:
+            return self._istrue()
+        except KeyboardInterrupt:
+            raise
+        except:
+            self.exc = sys.exc_info()
+            if isinstance(self.exc[1], SyntaxError):
+                msg = [" " * (self.exc[1].offset + 4) + "^",]
+                msg.append("SyntaxError: invalid syntax")
+            else:
+                msg = py.std.traceback.format_exception_only(*self.exc[:2])
+            pytest.fail("Error evaluating %r expression\n"
+                        "    %s\n"
+                        "%s"
+                        %(self.name, self.expr, "\n".join(msg)),
+                        pytrace=False)
+
+    def _getglobals(self):
+        d = {'os': py.std.os, 'sys': py.std.sys, 'config': self.item.config}
+        func = self.item.obj
+        try:
+            d.update(func.__globals__)
+        except AttributeError:
+            d.update(func.func_globals)
+        return d
+
+    def _istrue(self):
         if self.holder:
-            d = {'os': py.std.os, 'sys': py.std.sys, 'config': self.item.config}
+            d = self._getglobals()
             if self.holder.args:
                 self.result = False
                 for expr in self.holder.args:
@@ -42,7 +88,7 @@ class MarkEvaluator:
                     if isinstance(expr, str):
                         result = cached_eval(self.item.config, expr, d)
                     else:
-                        result = expr
+                        pytest.fail("expression is not a string")
                     if result:
                         self.result = True
                         self.expr = expr
@@ -60,7 +106,7 @@ class MarkEvaluator:
             if not hasattr(self, 'expr'):
                 return ""
             else:
-                return "condition: " + self.expr
+                return "condition: " + str(self.expr)
         return expl
 
 
@@ -97,19 +143,20 @@ def pytest_runtest_makereport(__multicall__, item, call):
             rep.keywords['xfail'] = "reason: " + call.excinfo.value.msg
             rep.outcome = "skipped"
             return rep
-    if call.when == "call":
-        rep = __multicall__.execute()
-        evalxfail = getattr(item, '_evalxfail')
-        if not item.config.getvalue("runxfail") and evalxfail.istrue():
+    rep = __multicall__.execute()
+    evalxfail = item._evalxfail
+    if not item.config.option.runxfail:
+        if evalxfail.wasvalid() and evalxfail.istrue():
             if call.excinfo:
                 rep.outcome = "skipped"
-            else:
+                rep.keywords['xfail'] = evalxfail.getexplanation()
+            elif call.when == "call":
                 rep.outcome = "failed"
-            rep.keywords['xfail'] = evalxfail.getexplanation()
-        else:
-            if 'xfail' in rep.keywords:
-                del rep.keywords['xfail']
-        return rep
+                rep.keywords['xfail'] = evalxfail.getexplanation()
+            return rep
+    if 'xfail' in rep.keywords:
+        del rep.keywords['xfail']
+    return rep
 
 # called by terminalreporter progress reporting
 def pytest_report_teststatus(report):
@@ -137,21 +184,23 @@ def pytest_terminal_summary(terminalreporter):
         elif char == "X":
             show_xpassed(terminalreporter, lines)
         elif char in "fF":
-            show_failed(terminalreporter, lines)
+            show_simple(terminalreporter, lines, 'failed', "FAIL %s")
         elif char in "sS":
             show_skipped(terminalreporter, lines)
+        elif char == "E":
+            show_simple(terminalreporter, lines, 'error', "ERROR %s")
     if lines:
         tr._tw.sep("=", "short test summary info")
         for line in lines:
             tr._tw.line(line)
 
-def show_failed(terminalreporter, lines):
+def show_simple(terminalreporter, lines, stat, format):
     tw = terminalreporter._tw
-    failed = terminalreporter.stats.get("failed")
+    failed = terminalreporter.stats.get(stat)
     if failed:
         for rep in failed:
             pos = rep.nodeid
-            lines.append("FAIL %s" %(pos, ))
+            lines.append(format %(pos, ))
 
 def show_xfailed(terminalreporter, lines):
     xfailed = terminalreporter.stats.get("xfailed")
@@ -179,7 +228,8 @@ def cached_eval(config, expr, d):
     except KeyError:
         #import sys
         #print >>sys.stderr, ("cache-miss: %r" % expr)
-        config._evalcache[expr] = x = eval(expr, d)
+        exprcode = py.code.compile(expr, mode="eval")
+        config._evalcache[expr] = x = eval(exprcode, d)
         return x
 
 

_pytest/terminal.py

@@ -15,7 +15,7 @@ def pytest_addoption(parser):
     group._addoption('-r',
          action="store", dest="reportchars", default=None, metavar="chars",
          help="show extra test summary info as specified by chars (f)ailed, "
-              "(s)skipped, (x)failed, (X)passed.")
+              "(E)error, (s)skipped, (x)failed, (X)passed.")
     group._addoption('-l', '--showlocals',
          action="store_true", dest="showlocals", default=False,
          help="show locals in tracebacks (disabled by default).")
@@ -25,29 +25,27 @@ def pytest_addoption(parser):
     group._addoption('--tb', metavar="style",
                action="store", dest="tbstyle", default='long',
                type="choice", choices=['long', 'short', 'no', 'line', 'native'],
-               help="traceback print mode (long/short/line/no).")
+               help="traceback print mode (long/short/line/native/no).")
     group._addoption('--fulltrace',
                action="store_true", dest="fulltrace", default=False,
                help="don't cut any tracebacks (default is to cut).")
 
 def pytest_configure(config):
     config.option.verbose -= config.option.quiet
-    if config.option.collectonly:
-        reporter = CollectonlyReporter(config)
-    else:
-        # we try hard to make printing resilient against
-        # later changes on FD level.
-        stdout = py.std.sys.stdout
-        if hasattr(os, 'dup') and hasattr(stdout, 'fileno'):
-            try:
-                newfd = os.dup(stdout.fileno())
-                #print "got newfd", newfd
-            except ValueError:
-                pass
-            else:
-                stdout = os.fdopen(newfd, stdout.mode, 1)
-                config._toclose = stdout
-        reporter = TerminalReporter(config, stdout)
+    # we try hard to make printing resilient against
+    # later changes on FD level.
+    stdout = py.std.sys.stdout
+    if hasattr(os, 'dup') and hasattr(stdout, 'fileno'):
+        try:
+            newfd = os.dup(stdout.fileno())
+            #print "got newfd", newfd
+        except ValueError:
+            pass
+        else:
+            stdout = os.fdopen(newfd, stdout.mode, 1)
+            config._cleanup.append(lambda: stdout.close())
+
+    reporter = TerminalReporter(config, stdout)
     config.pluginmanager.register(reporter, 'terminalreporter')
     if config.option.debug or config.option.traceconfig:
         def mywriter(tags, args):
@@ -55,11 +53,6 @@ def pytest_configure(config):
             reporter.write_line("[traceconfig] " + msg)
         config.trace.root.setprocessor("pytest:config", mywriter)
 
-def pytest_unconfigure(config):
-    if hasattr(config, '_toclose'):
-        #print "closing", config._toclose, config._toclose.fileno()
-        config._toclose.close()
-
 def getreportopt(config):
     reportopts = ""
     optvalue = config.option.report
@@ -136,6 +129,9 @@ class TerminalReporter:
             self._tw.line()
             self.currentfspath = None
 
+    def write(self, content, **markup):
+        self._tw.write(content, **markup)
+
     def write_line(self, line, **markup):
         line = str(line)
         self.ensure_newline()
@@ -165,9 +161,6 @@ class TerminalReporter:
     def pytest_deselected(self, items):
         self.stats.setdefault('deselected', []).extend(items)
 
-    def pytest__teardown_final_logerror(self, report):
-        self.stats.setdefault("error", []).append(report)
-
     def pytest_runtest_logstart(self, nodeid, location):
         # ensure that the path is printed before the
         # 1st test of a module starts running
@@ -215,7 +208,7 @@ class TerminalReporter:
 
     def pytest_collection(self):
         if not self.hasmarkup:
-            self.write_line("collecting ...", bold=True)
+            self.write("collecting ... ", bold=True)
 
     def pytest_collectreport(self, report):
         if report.failed:
@@ -259,7 +252,7 @@ class TerminalReporter:
         msg = "platform %s -- Python %s" % (sys.platform, verinfo)
         if hasattr(sys, 'pypy_version_info'):
             verinfo = ".".join(map(str, sys.pypy_version_info[:3]))
-            msg += "[pypy-%s]" % verinfo
+            msg += "[pypy-%s-%s]" % (verinfo, sys.pypy_version_info[3])
         msg += " -- pytest-%s" % (py.test.__version__)
         if self.verbosity > 0 or self.config.option.debug or \
            getattr(self.config.option, 'pastebin', None):
@@ -270,12 +263,53 @@ class TerminalReporter:
         for line in flatten(lines):
             self.write_line(line)
 
-    def pytest_collection_finish(self):
+    def pytest_collection_finish(self, session):
+        if self.config.option.collectonly:
+            self._printcollecteditems(session.items)
+            if self.stats.get('failed'):
+                self._tw.sep("!", "collection failures")
+                for rep in self.stats.get('failed'):
+                    rep.toterminal(self._tw)
+                return 1
+            return 0
         if not self.showheader:
             return
         #for i, testarg in enumerate(self.config.args):
         #    self.write_line("test path %d: %s" %(i+1, testarg))
 
+    def _printcollecteditems(self, items):
+        # to print out items and their parent collectors
+        # we take care to leave out Instances aka ()
+        # because later versions are going to get rid of them anyway
+        if self.config.option.verbose < 0:
+            if self.config.option.verbose < -1:
+                counts = {}
+                for item in items:
+                    name = item.nodeid.split('::', 1)[0]
+                    counts[name] = counts.get(name, 0) + 1
+                for name, count in sorted(counts.items()):
+                    self._tw.line("%s: %d" % (name, count))
+            else:
+                for item in items:
+                    nodeid = item.nodeid
+                    nodeid = nodeid.replace("::()::", "::")
+                    self._tw.line(nodeid)
+            return
+        stack = []
+        indent = ""
+        for item in items:
+            needed_collectors = item.listchain()[1:] # strip root node
+            while stack:
+                if stack == needed_collectors[:len(stack)]:
+                    break
+                stack.pop()
+            for col in needed_collectors[len(stack):]:
+                stack.append(col)
+                #if col.name == "()":
+                #    continue
+                indent = (len(stack)-1) * "  "
+                self._tw.line("%s%s" %(indent, col))
+
     def pytest_sessionfinish(self, exitstatus, __multicall__):
         __multicall__.execute()
         self._tw.line("")
@@ -285,12 +319,17 @@ class TerminalReporter:
             self.config.hook.pytest_terminal_summary(terminalreporter=self)
         if exitstatus == 2:
             self._report_keyboardinterrupt()
+            del self._keyboardinterrupt_memo
         self.summary_deselected()
         self.summary_stats()
 
     def pytest_keyboard_interrupt(self, excinfo):
         self._keyboardinterrupt_memo = excinfo.getrepr(funcargs=True)
 
+    def pytest_unconfigure(self):
+        if hasattr(self, '_keyboardinterrupt_memo'):
+            self._report_keyboardinterrupt()
+
     def _report_keyboardinterrupt(self):
         excrepr = self._keyboardinterrupt_memo
         msg = excrepr.reprcrash.message
@@ -302,19 +341,19 @@ class TerminalReporter:
                 excrepr.reprcrash.toterminal(self._tw)
 
     def _locationline(self, collect_fspath, fspath, lineno, domain):
-        if fspath and fspath != collect_fspath:
+        # collect_fspath comes from testid which has a "/"-normalized path
+        if fspath and fspath.replace("\\", "/") != collect_fspath:
             fspath = "%s <- %s" % (collect_fspath, fspath)
-        if lineno is not None:
-            lineno += 1
-        if fspath and lineno and domain:
-            line = "%(fspath)s:%(lineno)s: %(domain)s"
-        elif fspath and domain:
-            line = "%(fspath)s: %(domain)s"
-        elif fspath and lineno:
-            line = "%(fspath)s:%(lineno)s %(extrapath)s"
+        if fspath:
+            line = str(fspath)
+            if lineno is not None:
+                lineno += 1
+                line += ":" + str(lineno)
+            if domain:
+                line += ": " + str(domain)
         else:
-            line = "[nolocation]"
-        return line % locals() + " "
+            line = "[location]"
+        return line + " "
 
     def _getfailureheadline(self, rep):
         if hasattr(rep, 'location'):
@@ -355,7 +394,7 @@ class TerminalReporter:
                 else:
                     msg = self._getfailureheadline(rep)
                     self.write_sep("_", msg)
-                    rep.toterminal(self._tw)
+                    self._outrep_summary(rep)
 
     def summary_errors(self):
         if self.config.option.tbstyle != "no":
@@ -373,7 +412,15 @@ class TerminalReporter:
                 elif rep.when == "teardown":
                     msg = "ERROR at teardown of " + msg
                 self.write_sep("_", msg)
-                rep.toterminal(self._tw)
+                self._outrep_summary(rep)
+
+    def _outrep_summary(self, rep):
+        rep.toterminal(self._tw)
+        for secname, content in rep.sections:
+            self._tw.sep("-", secname)
+            if content[-1:] == "\n":
+                content = content[:-1]
+            self._tw.line(content)
 
     def summary_stats(self):
         session_duration = py.std.time.time() - self._sessionstarttime
@@ -384,9 +431,10 @@ class TerminalReporter:
                 keys.append(key)
         parts = []
         for key in keys:
-            val = self.stats.get(key, None)
-            if val:
-                parts.append("%d %s" %(len(val), key))
+            if key: # setup/teardown reports have an empty key, ignore them
+                val = self.stats.get(key, None)
+                if val:
+                    parts.append("%d %s" %(len(val), key))
         line = ", ".join(parts)
         # XXX coloring
         msg = "%s in %.2f seconds" %(line, session_duration)
@@ -397,54 +445,15 @@ class TerminalReporter:
 
     def summary_deselected(self):
         if 'deselected' in self.stats:
+            l = []
+            k = self.config.option.keyword
+            if k:
+                l.append("-k%s" % k)
+            m = self.config.option.markexpr
+            if m:
+                l.append("-m %r" % m)
             self.write_sep("=", "%d tests deselected by %r" %(
-                len(self.stats['deselected']), self.config.option.keyword), bold=True)
-
-
-class CollectonlyReporter:
-    INDENT = "  "
-
-    def __init__(self, config, out=None):
-        self.config = config
-        if out is None:
-            out = py.std.sys.stdout
-        self._tw = py.io.TerminalWriter(out)
-        self.indent = ""
-        self._failed = []
-
-    def outindent(self, line):
-        self._tw.line(self.indent + str(line))
-
-    def pytest_internalerror(self, excrepr):
-        for line in str(excrepr).split("\n"):
-            self._tw.line("INTERNALERROR> " + line)
-
-    def pytest_collectstart(self, collector):
-        if collector.session != collector:
-            self.outindent(collector)
-            self.indent += self.INDENT
-
-    def pytest_itemcollected(self, item):
-        self.outindent(item)
-
-    def pytest_collectreport(self, report):
-        if not report.passed:
-            if hasattr(report.longrepr, 'reprcrash'):
-                msg = report.longrepr.reprcrash.message
-            else:
-                # XXX unify (we have CollectErrorRepr here)
-                msg = str(report.longrepr[2])
-            self.outindent("!!! %s !!!" % msg)
-            #self.outindent("!!! error !!!")
-            self._failed.append(report)
-        self.indent = self.indent[:-len(self.INDENT)]
-
-    def pytest_collection_finish(self):
-        if self._failed:
-            self._tw.sep("!", "collection failures")
-        for rep in self._failed:
-            rep.toterminal(self._tw)
-        return self._failed and 1 or 0
+                len(self.stats['deselected']), " ".join(l)), bold=True)
 
 def repr_pythonversion(v=None):
     if v is None:

_pytest/tmpdir.py

@@ -46,20 +46,17 @@ class TempdirHandler:
 
     def finish(self):
         self.trace("finish")
-        
+
 def pytest_configure(config):
-    config._mp = mp = monkeypatch()
+    mp = monkeypatch()
     t = TempdirHandler(config)
+    config._cleanup.extend([mp.undo, t.finish])
     mp.setattr(config, '_tmpdirhandler', t, raising=False)
     mp.setattr(pytest, 'ensuretemp', t.ensuretemp, raising=False)
 
-def pytest_unconfigure(config):
-    config._tmpdirhandler.finish()
-    config._mp.undo()
-
 def pytest_funcarg__tmpdir(request):
     """return a temporary directory path object
-    unique to each test function invocation,
+    which is unique to each test function invocation,
     created as a sub directory of the base temporary
     directory.  The returned object is a `py.path.local`_
     path object.
@@ -67,5 +64,5 @@ def pytest_funcarg__tmpdir(request):
     name = request._pyfuncitem.name
     name = py.std.re.sub("[\W]", "_", name)
     x = request.config._tmpdirhandler.mktemp(name, numbered=True)
-    return x.realpath()
+    return x
 

_pytest/unittest.py

@@ -2,6 +2,9 @@
 import pytest, py
 import sys, pdb
 
+# for transfering markers
+from _pytest.python import transfer_markers
+
 def pytest_pycollect_makeitem(collector, name, obj):
     unittest = sys.modules.get('unittest')
     if unittest is None:
@@ -19,7 +22,14 @@ def pytest_pycollect_makeitem(collector, name, obj):
 class UnitTestCase(pytest.Class):
     def collect(self):
         loader = py.std.unittest.TestLoader()
+        module = self.getparent(pytest.Module).obj
+        cls = self.obj
         for name in loader.getTestCaseNames(self.obj):
+            x = getattr(self.obj, name)
+            funcobj = getattr(x, 'im_func', x)
+            transfer_markers(funcobj, cls, module)
+            if hasattr(funcobj, 'todo'):
+                pytest.mark.xfail(reason=str(funcobj.todo))(funcobj)
             yield TestCaseFunction(name, parent=self)
 
     def setup(self):
@@ -37,15 +47,13 @@ class UnitTestCase(pytest.Class):
 class TestCaseFunction(pytest.Function):
     _excinfo = None
 
-    def __init__(self, name, parent):
-        super(TestCaseFunction, self).__init__(name, parent)
-        if hasattr(self._obj, 'todo'):
-            getattr(self._obj, 'im_func', self._obj).xfail = \
-                pytest.mark.xfail(reason=str(self._obj.todo))
-
     def setup(self):
         self._testcase = self.parent.obj(self.name)
         self._obj = getattr(self._testcase, self.name)
+        if hasattr(self._testcase, 'skip'):
+            pytest.skip(self._testcase.skip)
+        if hasattr(self._obj, 'skip'):
+            pytest.skip(self._obj.skip)
         if hasattr(self._testcase, 'setup_method'):
             self._testcase.setup_method(self._obj)
 
@@ -102,6 +110,13 @@ class TestCaseFunction(pytest.Function):
     def runtest(self):
         self._testcase(result=self)
 
+    def _prunetraceback(self, excinfo):
+        pytest.Function._prunetraceback(self, excinfo)
+        traceback = excinfo.traceback.filter(
+            lambda x:not x.frame.f_globals.get('__unittest'))
+        if traceback:
+            excinfo.traceback = traceback
+
 @pytest.mark.tryfirst
 def pytest_runtest_makereport(item, call):
     if isinstance(item, TestCaseFunction):
@@ -116,14 +131,19 @@ def pytest_runtest_protocol(item, __multicall__):
             ut = sys.modules['twisted.python.failure']
             Failure__init__ = ut.Failure.__init__.im_func
             check_testcase_implements_trial_reporter()
-            def excstore(self, exc_value=None, exc_type=None, exc_tb=None):
+            def excstore(self, exc_value=None, exc_type=None, exc_tb=None,
+                captureVars=None):
                 if exc_value is None:
                     self._rawexcinfo = sys.exc_info()
                 else:
                     if exc_type is None:
                         exc_type = type(exc_value)
                     self._rawexcinfo = (exc_type, exc_value, exc_tb)
-                Failure__init__(self, exc_value, exc_type, exc_tb)
+                try:
+                    Failure__init__(self, exc_value, exc_type, exc_tb,
+                        captureVars=captureVars)
+                except TypeError:
+                    Failure__init__(self, exc_value, exc_type, exc_tb)
             ut.Failure.__init__ = excstore
             try:
                 return __multicall__.execute()

distribute_setup.py

@@ -46,7 +46,7 @@ except ImportError:
             args = [quote(arg) for arg in args]
         return os.spawnl(os.P_WAIT, sys.executable, *args) == 0
 
-DEFAULT_VERSION = "0.6.14"
+DEFAULT_VERSION = "0.6.24"
 DEFAULT_URL = "http://pypi.python.org/packages/source/d/distribute/"
 SETUPTOOLS_FAKED_VERSION = "0.6c11"
 

doc/Makefile

@@ -15,7 +15,7 @@ ALLSPHINXOPTS   = -d $(BUILDDIR)/doctrees $(PAPEROPT_$(PAPER)) $(SPHINXOPTS) .
 .PHONY: help clean html dirhtml singlehtml pickle json htmlhelp qthelp devhelp epub latex latexpdf text man changes linkcheck doctest
 
 regen:
-	COLUMNS=76 regendoc --update *.txt */*.txt
+	PYTHONDONTWRITEBYTECODE=1 COLUMNS=76 regendoc --update *.txt */*.txt
     
 help:
 	@echo "Please use \`make <target>' where <target> is one of"
@@ -39,8 +39,14 @@ help:
 clean:
 	-rm -rf $(BUILDDIR)/*
 
-install: clean html
-	rsync -avz _build/html/ code:www-pytest/2.0.0
+install: html
+	 rsync -avz _build/html/ pytest.org:/www/pytest.org/latest
+
+installpdf: latexpdf
+	@scp $(BUILDDIR)/latex/pytest.pdf pytest.org:/www/pytest.org/latest
+
+installall: clean install installpdf
+	@echo "done"
 
 html:
 	$(SPHINXBUILD) -b html $(ALLSPHINXOPTS) $(BUILDDIR)/html

doc/_static/sphinxdoc.css

@@ -48,7 +48,7 @@ div.body {
 }
 
 div.related {
-    font-size: 1em;
+    font-size: 0.8em;
 }
 
 div.related ul {

doc/_templates/indexsidebar.html

@@ -7,10 +7,10 @@
   released versions in the <a href="http://pypi.python.org/pypi/pytest">Python
     Package Index</a>.</p>
 {% else %}
-<p><b><a href="{{ pathto('announce/index')}}">{{ version }} release</a></b>
+<p><b><a href="{{ pathto('announce/index')}}">{{ release }} release</a></b>
 [<a href="{{ pathto('changelog') }}">Changelog</a>]</p>
 <p>
-<a href="http://pypi.python.org/pypi/pytest">pytest on PyPI</a>
+<a href="http://pypi.python.org/pypi/pytest">pytest/PyPI</a>
 </p>
 <pre>easy_install pytest</pre>
 <pre>pip install pytest</pre>

doc/_templates/layout.html

@@ -1,27 +1,8 @@
 {% extends "!layout.html" %}
 
-{% block relbar1 %}
-{% endblock %}
-{% block relbar2 %}
-{% endblock %}
-
-{% block rootrellink %}
-{% endblock %}
-{% block sidebarrel %}
-{% endblock %}
-
-{% block header %}
-<div style="background-color: white; text-align: left; padding: 10px 10px 15px 15px">
-    <h1>pytest: rapid no-boilerplate testing with Python</h1>
-    <div style="text-align: left; font-size: 130%; vertical-align: middle;">
-            <a href="{{ pathto('index') }}">home</a>&nbsp;|&nbsp;
-            <a href="{{ pathto('contents') }}">all docs</a>&nbsp;|&nbsp;
-            <a href="{{ pathto('getting-started') }}">install</a>&nbsp;|&nbsp;
-            <a href="{{ pathto('example/index') }}">examples</a>&nbsp;|&nbsp;
-            <a href="{{ pathto('customize') }}">customize</a>&nbsp;|&nbsp;
-            <a href="{{ pathto('contact') }}">contact</a>&nbsp;
-    </div>
-</div>
+{% block relbaritems %}
+{{ super() }}
+<g:plusone></g:plusone>
 {% endblock %}
 
 {% block footer %}
@@ -39,4 +20,5 @@
   })();
 
 </script>
+<script type="text/javascript" src="https://apis.google.com/js/plusone.js"></script>
 {% endblock %}

doc/_templates/localtoc.html

@@ -0,0 +1,37 @@
+
+{%- if pagename != "search" %}
+<div id="searchbox" style="display: none">
+    <form class="search" action="{{ pathto('search') }}" method="get">
+      <input type="text" name="q" size="18" />
+      <input type="submit" value="{{ _('Search') }}" />
+      <input type="hidden" name="check_keywords" value="yes" />
+      <input type="hidden" name="area" value="default" />
+    </form>
+</div>
+<script type="text/javascript">$('#searchbox').show(0);</script>
+{%- endif %}
+
+<h3>quicklinks</h3>
+<div style="text-align: left; font-size: 100%; vertical-align: middle;">
+<table>
+<tr>
+<td>
+        <a href="{{ pathto('index') }}">home</a>
+</td><td>
+        <a href="{{ pathto('contents') }}">TOC/contents</a>
+</td></tr><tr><td>
+        <a href="{{ pathto('getting-started') }}">install</a>
+</td><td>
+        <a href="{{ pathto('changelog') }}">changelog</a>
+</td></tr><tr><td>
+        <a href="{{ pathto('example/index') }}">examples</a>
+</td><td>
+        <a href="{{ pathto('customize') }}">customize</a>
+</td></tr><tr><td>
+        <a href="https://bitbucket.org/hpk42/pytest/issues?status=new&status=open">issues[bb]</a>
+</td><td>
+        <a href="{{ pathto('contact') }}">contact</a>
+</td></tr></table>
+</div>
+{% extends "basic/localtoc.html" %}
+

doc/announce/index.txt

@@ -5,5 +5,15 @@ Release announcements
 .. toctree::
    :maxdepth: 2
 
+   release-2.2.2
+   release-2.2.1
+   release-2.2.0
+   release-2.1.3
+   release-2.1.2
+   release-2.1.1
+   release-2.1.0
+   release-2.0.3
+   release-2.0.2
+   release-2.0.1
    release-2.0.0
 

doc/announce/release-2.0.0.txt

@@ -117,7 +117,7 @@ Important Notes
 - py.test.collect.Directory does not exist anymore and it
   is not possible to provide an own "Directory" object.
   If you have used this and don't know what to do, get
-  in contact.  We'll figure someting out.
+  in contact.  We'll figure something out.
 
   Note that pytest_collect_directory() is still called but
   any return value will be ignored.  This allows to keep

doc/announce/release-2.0.1.txt

@@ -0,0 +1,67 @@
+py.test 2.0.1: bug fixes
+===========================================================================
+
+Welcome to pytest-2.0.1, a maintenance and bug fix release of pytest,
+a mature testing tool for Python, supporting CPython 2.4-3.2, Jython
+and latest PyPy interpreters.  See extensive docs with tested examples here:
+
+    http://pytest.org/
+
+If you want to install or upgrade pytest, just type one of::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+Many thanks to all issue reporters and people asking questions or
+complaining.  Particular thanks to Floris Bruynooghe and Ronny Pfannschmidt
+for their great coding contributions and many others for feedback and help.
+
+best,
+holger krekel
+
+Changes between 2.0.0 and 2.0.1
+----------------------------------------------
+
+- refine and unify initial capturing so that it works nicely
+  even if the logging module is used on an early-loaded conftest.py
+  file or plugin.
+- fix issue12 - show plugin versions with "--version" and
+  "--traceconfig" and also document how to add extra information
+  to reporting test header
+- fix issue17 (import-* reporting issue on python3) by
+  requiring py>1.4.0 (1.4.1 is going to include it)
+- fix issue10 (numpy arrays truth checking) by refining
+  assertion interpretation in py lib
+- fix issue15: make nose compatibility tests compatible
+  with python3 (now that nose-1.0 supports python3)
+- remove somewhat surprising "same-conftest" detection because
+  it ignores conftest.py when they appear in several subdirs.
+- improve assertions ("not in"), thanks Floris Bruynooghe
+- improve behaviour/warnings when running on top of "python -OO"
+  (assertions and docstrings are turned off, leading to potential
+  false positives)
+- introduce a pytest_cmdline_processargs(args) hook
+  to allow dynamic computation of command line arguments.
+  This fixes a regression because py.test prior to 2.0
+  allowed to set command line options from conftest.py
+  files which so far pytest-2.0 only allowed from ini-files now.
+- fix issue7: assert failures in doctest modules.
+  unexpected failures in doctests will not generally
+  show nicer, i.e. within the doctest failing context.
+- fix issue9: setup/teardown functions for an xfail-marked
+  test will report as xfail if they fail but report as normally
+  passing (not xpassing) if they succeed.  This only is true
+  for "direct" setup/teardown invocations because teardown_class/
+  teardown_module cannot closely relate to a single test.
+- fix issue14: no logging errors at process exit
+- refinements to "collecting" output on non-ttys
+- refine internal plugin registration and --traceconfig output
+- introduce a mechanism to prevent/unregister plugins from the
+  command line, see http://pytest.org/plugins.html#cmdunregister
+- activate resultlog plugin by default
+- fix regression wrt yielded tests which due to the
+  collection-before-running semantics were not
+  setup as with pytest 1.3.4.  Note, however, that
+  the recommended and much cleaner way to do test
+  parametrization remains the "pytest_generate_tests"
+  mechanism, see the docs.

doc/announce/release-2.0.2.txt

@@ -0,0 +1,73 @@
+py.test 2.0.2: bug fixes, improved xfail/skip expressions, speed ups
+===========================================================================
+
+Welcome to pytest-2.0.2, a maintenance and bug fix release of pytest,
+a mature testing tool for Python, supporting CPython 2.4-3.2, Jython
+and latest PyPy interpreters.  See the extensive docs with tested examples here:
+
+    http://pytest.org/
+
+If you want to install or upgrade pytest, just type one of::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+Many thanks to all issue reporters and people asking questions
+or complaining, particularly Jurko for his insistence,
+Laura, Victor and Brianna for helping with improving
+and Ronny for his general advise.
+
+best,
+holger krekel
+
+Changes between 2.0.1 and 2.0.2
+----------------------------------------------
+
+- tackle issue32 - speed up test runs of very quick test functions
+  by reducing the relative overhead
+
+- fix issue30 - extended xfail/skipif handling and improved reporting.
+  If you have a syntax error in your skip/xfail
+  expressions you now get nice error reports.
+
+  Also you can now access module globals from xfail/skipif
+  expressions so that this for example works now::
+
+    import pytest
+    import mymodule
+    @pytest.mark.skipif("mymodule.__version__[0] == "1")
+    def test_function():
+        pass
+
+  This will not run the test function if the module's version string
+  does not start with a "1".  Note that specifying a string instead
+  of a boolean expressions allows py.test to report meaningful information
+  when summarizing a test run as to what conditions lead to skipping
+  (or xfail-ing) tests.
+
+- fix issue28 - setup_method and pytest_generate_tests work together
+  The setup_method fixture method now gets called also for
+  test function invocations generated from the pytest_generate_tests
+  hook.
+
+- fix issue27 - collectonly and keyword-selection (-k) now work together
+  Also, if you do "py.test --collectonly -q" you now get a flat list
+  of test ids that you can use to paste to the py.test commandline
+  in order to execute a particular test.
+
+- fix issue25 avoid reported problems with --pdb and python3.2/encodings output
+
+- fix issue23 - tmpdir argument now works on Python3.2 and WindowsXP
+  Starting with Python3.2 os.symlink may be supported. By requiring
+  a newer py lib version the py.path.local() implementation acknowledges
+  this.
+
+- fixed typos in the docs (thanks Victor Garcia, Brianna Laugher) and particular
+  thanks to Laura Creighton who also revieved parts of the documentation.
+
+- fix slighly wrong output of verbose progress reporting for classes
+  (thanks Amaury)
+
+- more precise (avoiding of) deprecation warnings for node.Class|Function accesses
+
+- avoid std unittest assertion helper code in tracebacks (thanks Ronny)

doc/announce/release-2.0.3.txt

@@ -0,0 +1,40 @@
+py.test 2.0.3: bug fixes and speed ups 
+===========================================================================
+
+Welcome to pytest-2.0.3, a maintenance and bug fix release of pytest,
+a mature testing tool for Python, supporting CPython 2.4-3.2, Jython
+and latest PyPy interpreters.  See the extensive docs with tested examples here:
+
+    http://pytest.org/
+
+If you want to install or upgrade pytest, just type one of::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+There also is a bugfix release 1.6 of pytest-xdist, the plugin
+that enables seemless distributed and "looponfail" testing for Python.
+
+best,
+holger krekel
+
+Changes between 2.0.2 and 2.0.3
+----------------------------------------------
+
+- fix issue38: nicer tracebacks on calls to hooks, particularly early
+  configure/sessionstart ones
+
+- fix missing skip reason/meta information in junitxml files, reported
+  via http://lists.idyll.org/pipermail/testing-in-python/2011-March/003928.html
+
+- fix issue34: avoid collection failure with "test" prefixed classes
+  deriving from object.
+
+- don't require zlib (and other libs) for genscript plugin without
+  --genscript actually being used.
+
+- speed up skips (by not doing a full traceback represenation
+  internally)
+
+- fix issue37: avoid invalid characters in junitxml's output
+

doc/announce/release-2.1.0.txt

@@ -0,0 +1,47 @@
+py.test 2.1.0: perfected assertions and bug fixes
+===========================================================================
+
+Welcome to the release of pytest-2.1, a mature testing tool for Python,
+supporting CPython 2.4-3.2, Jython and latest PyPy interpreters.  See
+the improved extensive docs (now also as PDF!) with tested examples here:
+
+     http://pytest.org/
+
+The single biggest news about this release are **perfected assertions**
+courtesy of Benjamin Peterson.  You can now safely use ``assert``
+statements in test modules without having to worry about side effects
+or python optimization ("-OO") options.  This is achieved by rewriting
+assert statements in test modules upon import, using a PEP302 hook.
+See http://pytest.org/assert.html#advanced-assertion-introspection for
+detailed information.  The work has been partly sponsored by my company,
+merlinux GmbH.
+
+For further details on bug fixes and smaller enhancements see below.
+
+If you want to install or upgrade pytest, just type one of::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+best,
+holger krekel / http://merlinux.eu
+
+Changes between 2.0.3 and 2.1.0
+----------------------------------------------
+
+- fix issue53 call nosestyle setup functions with correct ordering
+- fix issue58 and issue59: new assertion code fixes
+- merge Benjamin's assertionrewrite branch: now assertions
+  for test modules on python 2.6 and above are done by rewriting
+  the AST and saving the pyc file before the test module is imported.
+  see doc/assert.txt for more info.
+- fix issue43: improve doctests with better traceback reporting on
+  unexpected exceptions
+- fix issue47: timing output in junitxml for test cases is now correct
+- fix issue48: typo in MarkInfo repr leading to exception
+- fix issue49: avoid confusing error when initialization partially fails
+- fix issue44: env/username expansion for junitxml file path
+- show releaselevel information in test runs for pypy
+- reworked doc pages for better navigation and PDF generation
+- report KeyboardInterrupt even if interrupted during session startup
+- fix issue 35 - provide PDF doc version and download link from index page

doc/announce/release-2.1.1.txt

@@ -0,0 +1,37 @@
+py.test 2.1.1: assertion fixes and improved junitxml output
+===========================================================================
+
+pytest-2.1.1 is a backward compatible maintenance release of the
+popular py.test testing tool.  See extensive docs with examples here:
+
+     http://pytest.org/
+
+Most bug fixes address remaining issues with the perfected assertions
+introduced with 2.1.0 - many thanks to the bug reporters and to Benjamin
+Peterson for helping to fix them.  Also, junitxml output now produces
+system-out/err tags which lead to better displays of tracebacks with Jenkins.
+
+Also a quick note to package maintainers and others interested: there now
+is a "pytest" man page which can be generated with "make man" in doc/.
+
+If you want to install or upgrade pytest, just type one of::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+best,
+holger krekel / http://merlinux.eu
+
+Changes between 2.1.0 and 2.1.1
+----------------------------------------------
+
+- fix issue64 / pytest.set_trace now works within pytest_generate_tests hooks
+- fix issue60 / fix error conditions involving the creation of __pycache__
+- fix issue63 / assertion rewriting on inserts involving strings containing '%'
+- fix assertion rewriting on calls with a ** arg
+- don't cache rewritten modules if bytecode generation is disabled
+- fix assertion rewriting in read-only directories
+- fix issue59: provide system-out/err tags for junitxml output
+- fix issue61: assertion rewriting on boolean operations with 3 or more operands
+- you can now build a man page with "cd doc ; make man"
+

doc/announce/release-2.1.2.txt

@@ -0,0 +1,33 @@
+py.test 2.1.2: bug fixes and fixes for jython
+===========================================================================
+
+pytest-2.1.2 is a minor backward compatible maintenance release of the
+popular py.test testing tool.  pytest is commonly used for unit,
+functional- and integration testing.  See extensive docs with examples
+here:
+
+     http://pytest.org/
+
+Most bug fixes address remaining issues with the perfected assertions
+introduced in the 2.1 series - many thanks to the bug reporters and to Benjamin
+Peterson for helping to fix them.  pytest should also work better with
+Jython-2.5.1 (and Jython trunk).
+
+If you want to install or upgrade pytest, just type one of::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+best,
+holger krekel / http://merlinux.eu
+
+Changes between 2.1.1 and 2.1.2
+----------------------------------------
+
+- fix assertion rewriting on files with windows newlines on some Python versions
+- refine test discovery by package/module name (--pyargs), thanks Florian Mayer
+- fix issue69 / assertion rewriting fixed on some boolean operations
+- fix issue68 / packages now work with assertion rewriting
+- fix issue66: use different assertion rewriting caches when the -O option is passed
+- don't try assertion rewriting on Jython, use reinterp
+

doc/announce/release-2.1.3.txt

@@ -0,0 +1,32 @@
+py.test 2.1.3: just some more fixes
+===========================================================================
+
+pytest-2.1.3 is a minor backward compatible maintenance release of the
+popular py.test testing tool.  It is commonly used for unit, functional-
+and integration testing.  See extensive docs with examples here:
+
+     http://pytest.org/
+
+The release contains another fix to the perfected assertions introduced
+with the 2.1 series as well as the new possibility to customize reporting
+for assertion expressions on a per-directory level.  
+
+If you want to install or upgrade pytest, just type one of::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+Thanks to the bug reporters and to Ronny Pfannschmidt, Benjamin Peterson
+and Floris Bruynooghe who implemented the fixes.
+
+best,
+holger krekel
+
+Changes between 2.1.2 and 2.1.3
+----------------------------------------
+
+- fix issue79: assertion rewriting failed on some comparisons in boolops,
+- correctly handle zero length arguments (a la pytest '')
+- fix issue67 / junitxml now contains correct test durations
+- fix issue75 / skipping test failure on jython
+- fix issue77 / Allow assertrepr_compare hook to apply to a subset of tests

doc/announce/release-2.2.0.txt

@@ -0,0 +1,95 @@
+py.test 2.2.0: test marking++, parametrization++ and duration profiling
+===========================================================================
+
+pytest-2.2.0 is a test-suite compatible release of the popular
+py.test testing tool.  Plugins might need upgrades. It comes
+with these improvements:
+
+* easier and more powerful parametrization of tests:
+
+  - new @pytest.mark.parametrize decorator to run tests with different arguments
+  - new metafunc.parametrize() API for parametrizing arguments independently
+  - see examples at http://pytest.org/latest/example/parametrize.html
+  - NOTE that parametrize() related APIs are still a bit experimental
+    and might change in future releases.
+
+* improved handling of test markers and refined marking mechanism:
+
+  - "-m markexpr" option for selecting tests according to their mark
+  - a new "markers" ini-variable for registering test markers for your project
+  - the new "--strict" bails out with an error if using unregistered markers.
+  - see examples at http://pytest.org/latest/example/markers.html
+
+* duration profiling: new "--duration=N" option showing the N slowest test
+  execution or setup/teardown calls. This is most useful if you want to
+  find out where your slowest test code is.
+
+* also 2.2.0 performs more eager calling of teardown/finalizers functions
+  resulting in better and more accurate reporting when they fail
+
+Besides there is the usual set of bug fixes along with a cleanup of
+pytest's own test suite allowing it to run on a wider range of environments.
+
+For general information, see extensive docs with examples here:
+
+     http://pytest.org/
+
+If you want to install or upgrade pytest you might just type::
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+Thanks to Ronny Pfannschmidt, David Burns, Jeff Donner, Daniel Nouri, Alfredo Deza and all who gave feedback or sent bug reports.
+
+best,
+holger krekel
+
+
+notes on incompatibility
+------------------------------
+
+While test suites should work unchanged you might need to upgrade plugins:
+
+* You need a new version of the pytest-xdist plugin (1.7) for distributing
+  test runs.
+
+* Other plugins might need an upgrade if they implement
+  the ``pytest_runtest_logreport`` hook which now is called unconditionally
+  for the setup/teardown fixture phases of a test. You may choose to
+  ignore setup/teardown failures by inserting "if rep.when != 'call': return"
+  or something similar. Note that most code probably "just" works because
+  the hook was already called for failing setup/teardown phases of a test
+  so a plugin should have been ready to grok such reports already.
+
+
+Changes between 2.1.3 and 2.2.0
+----------------------------------------
+
+- fix issue90: introduce eager tearing down of test items so that
+  teardown function are called earlier.
+- add an all-powerful metafunc.parametrize function which allows to
+  parametrize test function arguments in multiple steps and therefore
+  from independent plugins and places.
+- add a @pytest.mark.parametrize helper which allows to easily
+  call a test function with different argument values.
+- Add examples to the "parametrize" example page, including a quick port
+  of Test scenarios and the new parametrize function and decorator.
+- introduce registration for "pytest.mark.*" helpers via ini-files
+  or through plugin hooks.  Also introduce a "--strict" option which
+  will treat unregistered markers as errors
+  allowing to avoid typos and maintain a well described set of markers
+  for your test suite.  See examples at http://pytest.org/latest/mark.html
+  and its links.
+- issue50: introduce "-m marker" option to select tests based on markers
+  (this is a stricter and more predictable version of "-k" in that "-m"
+  only matches complete markers and has more obvious rules for and/or
+  semantics.
+- new feature to help optimizing the speed of your tests:
+  --durations=N option for displaying N slowest test calls
+  and setup/teardown methods.
+- fix issue87: --pastebin now works with python3
+- fix issue89: --pdb with unexpected exceptions in doctest work more sensibly
+- fix and cleanup pytest's own test suite to not leak FDs
+- fix issue83: link to generated funcarg list
+- fix issue74: pyarg module names are now checked against imp.find_module false positives
+- fix compatibility with twisted/trial-11.1.0 use cases

doc/announce/release-2.2.1.txt

@@ -0,0 +1,41 @@
+pytest-2.2.1: bug fixes, perfect teardowns
+===========================================================================
+
+
+pytest-2.2.1 is a minor backward-compatible release of the the py.test
+testing tool.   It contains bug fixes and little improvements, including
+documentation fixes.  If you are using the distributed testing
+pluginmake sure to upgrade it to pytest-xdist-1.8.
+
+For general information see here:
+
+     http://pytest.org/
+
+To install or upgrade pytest:
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+Special thanks for helping on this release to Ronny Pfannschmidt, Jurko
+Gospodnetic and Ralf Schmitt.
+
+best,
+holger krekel
+
+
+Changes between 2.2.0 and 2.2.1
+----------------------------------------
+
+- fix issue99 (in pytest and py) internallerrors with resultlog now
+  produce better output - fixed by normalizing pytest_internalerror 
+  input arguments.
+- fix issue97 / traceback issues (in pytest and py) improve traceback output
+  in conjunction with jinja2 and cython which hack tracebacks
+- fix issue93 (in pytest and pytest-xdist) avoid "delayed teardowns":
+  the final test in a test node will now run its teardown directly
+  instead of waiting for the end of the session. Thanks Dave Hunt for
+  the good reporting and feedback.  The pytest_runtest_protocol as well
+  as the pytest_runtest_teardown hooks now have "nextitem" available 
+  which will be None indicating the end of the test run.
+- fix collection crash due to unknown-source collected items, thanks
+  to Ralf Schmitt (fixed by depending on a more recent pylib)

doc/announce/release-2.2.2.txt

@@ -0,0 +1,42 @@
+pytest-2.2.2: bug fixes
+===========================================================================
+
+pytest-2.2.2 is a minor backward-compatible release of the versatile py.test
+testing tool.   It contains bug fixes and a few refinements particularly
+to reporting with "--collectonly", see below for betails.  
+
+For general information see here:
+
+     http://pytest.org/
+
+To install or upgrade pytest:
+
+    pip install -U pytest # or
+    easy_install -U pytest
+
+Special thanks for helping on this release to Ronny Pfannschmidt
+and Ralf Schmitt and the contributors of issues.
+
+best,
+holger krekel
+
+
+Changes between 2.2.1 and 2.2.2
+----------------------------------------
+
+- fix issue101: wrong args to unittest.TestCase test function now
+  produce better output
+- fix issue102: report more useful errors and hints for when a 
+  test directory was renamed and some pyc/__pycache__ remain
+- fix issue106: allow parametrize to be applied multiple times
+  e.g. from module, class and at function level.
+- fix issue107: actually perform session scope finalization
+- don't check in parametrize if indirect parameters are funcarg names
+- add chdir method to monkeypatch funcarg
+- fix crash resulting from calling monkeypatch undo a second time
+- fix issue115: make --collectonly robust against early failure
+  (missing files/directories)
+- "-qq --collectonly" now shows only files and the number of tests in them
+- "-q --collectonly" now shows test ids 
+- allow adding of attributes to test reports such that it also works
+  with distributed testing (no upgrade of pytest-xdist needed)

doc/assert.txt

@@ -1,15 +1,15 @@
 
-Writing and reporting of assertions in tests
-============================================
+The writing and reporting of assertions in tests
+==================================================
 
 .. _`assert with the assert statement`:
 
-assert with the ``assert`` statement
+Asserting with the ``assert`` statement
 ---------------------------------------------------------
 
-``py.test`` allows to use the standard python ``assert`` for verifying
+``py.test`` allows you to use the standard python ``assert`` for verifying
 expectations and values in Python tests.  For example, you can write the
-following in your tests::
+following::
 
     # content of test_assert1.py
     def f():
@@ -18,49 +18,49 @@ following in your tests::
     def test_function():
         assert f() == 4
 
-to state that your object has a certain ``attribute``. In case this
-assertion fails you will see the value of ``x``::
+to assert that your function returns a certain value. If this assertion fails
+you will see the return value of the function call::
 
     $ py.test test_assert1.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_assert1.py
-
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
+    
     test_assert1.py F
-
+    
     ================================= FAILURES =================================
     ______________________________ test_function _______________________________
-
+    
         def test_function():
     >       assert f() == 4
     E       assert 3 == 4
     E        +  where 3 = f()
-
+    
     test_assert1.py:5: AssertionError
-    ========================= 1 failed in 0.03 seconds =========================
+    ========================= 1 failed in 0.02 seconds =========================
 
-Reporting details about the failing assertion is achieved by re-evaluating
-the assert expression and recording intermediate values.
+py.test has support for showing the values of the most common subexpressions
+including calls, attributes, comparisons, and binary and unary
+operators. (See :ref:`tbreportdemo`).  This allows you to use the
+idiomatic python constructs without boilerplate code while not losing
+introspection information.
 
-Note: If evaluating the assert expression has side effects you may get a
-warning that the intermediate values could not be determined safely.  A
-common example for this issue is reading from a file and comparing in one
-line::
+However, if you specify a message with the assertion like this::
 
-        assert f.read() != '...'
+    assert a % 2 == 0, "value was odd, should be even"
 
-This might fail but when re-interpretation comes along it might pass.
-You can rewrite this (and any other expression with side effects) easily, though:
+then no assertion introspection takes places at all and the message
+will be simply shown in the traceback.
 
-        content = f.read()
-        assert content != '...'
+See :ref:`assert-details` for more information on assertion introspection.
 
-assertions about expected exceptions
+Assertions about expected exceptions
 ------------------------------------------
 
 In order to write assertions about raised exceptions, you can use
 ``pytest.raises`` as a context manager like this::
 
+    import pytest
     with pytest.raises(ZeroDivisionError):
         1 / 0
 
@@ -73,7 +73,7 @@ and if you need to have access to the actual exception info you may use::
 
     # do checks related to excinfo.type, excinfo.value, excinfo.traceback
 
-If you want to write test code that works on Python2.4 as well,
+If you want to write test code that works on Python 2.4 as well,
 you may also use two other ways to test for an expected exception::
 
     pytest.raises(ExpectedException, func, *args, **kwargs)
@@ -91,7 +91,7 @@ Making use of context-sensitive comparisons
 
 .. versionadded:: 2.0
 
-py.test has rich support for providing context-sensitive informations
+py.test has rich support for providing context-sensitive information
 when it encounters comparisons.  For example::
 
     # content of test_assert2.py
@@ -105,14 +105,14 @@ if you run this module::
 
     $ py.test test_assert2.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_assert2.py
-
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
+    
     test_assert2.py F
-
+    
     ================================= FAILURES =================================
     ___________________________ test_set_comparison ____________________________
-
+    
         def test_set_comparison():
             set1 = set("1308")
             set2 = set("8035")
@@ -122,9 +122,9 @@ if you run this module::
     E         '1'
     E         Extra items in the right set:
     E         '5'
-
+    
     test_assert2.py:5: AssertionError
-    ========================= 1 failed in 0.02 seconds =========================
+    ========================= 1 failed in 0.03 seconds =========================
 
 Special comparisons are done for a number of cases:
 
@@ -134,8 +134,117 @@ Special comparisons are done for a number of cases:
 
 See the :ref:`reporting demo <tbreportdemo>` for many more examples.
 
-..
-    Defining your own comparison
-    ----------------------------------------------
+Defining your own assertion comparison
+----------------------------------------------
+
+It is possible to add your own detailed explanations by implementing
+the ``pytest_assertrepr_compare`` hook.
+
+.. autofunction:: _pytest.hookspec.pytest_assertrepr_compare
+
+As an example consider adding the following hook in a conftest.py which
+provides an alternative explanation for ``Foo`` objects::
+
+   # content of conftest.py
+   from test_foocompare import Foo
+   def pytest_assertrepr_compare(op, left, right):
+       if isinstance(left, Foo) and isinstance(right, Foo) and op == "==":
+        return ['Comparing Foo instances:',
+                  '   vals: %s != %s' % (left.val, right.val)]
+
+now, given this test module::
+
+   # content of test_foocompare.py
+   class Foo:
+       def __init__(self, val):
+            self.val = val
+
+   def test_compare():
+       f1 = Foo(1)
+       f2 = Foo(2)
+       assert f1 == f2
+
+you can run the test module and get the custom output defined in 
+the conftest file::
+
+   $ py.test -q test_foocompare.py
+   collecting ... collected 1 items
+   F
+   ================================= FAILURES =================================
+   _______________________________ test_compare _______________________________
+   
+       def test_compare():
+           f1 = Foo(1)
+           f2 = Foo(2)
+   >       assert f1 == f2
+   E       assert Comparing Foo instances:
+   E            vals: 1 != 2
+   
+   test_foocompare.py:8: AssertionError
+   1 failed in 0.02 seconds
+
+.. _assert-details:
+.. _`assert introspection`:
+
+Advanced assertion introspection
+----------------------------------
+
+.. versionadded:: 2.1
 
 
+Reporting details about a failing assertion is achieved either by rewriting
+assert statements before they are run or re-evaluating the assert expression and
+recording the intermediate values. Which technique is used depends on the
+location of the assert, py.test's configuration, and Python version being used
+to run py.test.  Note that for assert statements with a manually provided
+message, i.e. ``assert expr, message``, no assertion introspection takes place
+and the manually provided message will be rendered in tracebacks.
+
+By default, if the Python version is greater than or equal to 2.6, py.test
+rewrites assert statements in test modules. Rewritten assert statements put
+introspection information into the assertion failure message. py.test only
+rewrites test modules directly discovered by its test collection process, so
+asserts in supporting modules which are not themselves test modules will not be
+rewritten.
+
+.. note::
+
+   py.test rewrites test modules on import. It does this by using an import hook
+   to write a new pyc files. Most of the time this works transparently. However,
+   if you are messing with import yourself, the import hook may interfere. If
+   this is the case, simply use ``--assert=reinterp`` or
+   ``--assert=plain``. Additionally, rewriting will fail silently if it cannot
+   write new pycs, i.e. in a read-only filesystem or a zipfile.
+
+If an assert statement has not been rewritten or the Python version is less than
+2.6, py.test falls back on assert reinterpretation. In assert reinterpretation,
+py.test walks the frame of the function containing the assert statement to
+discover sub-expression results of the failing assert statement. You can force
+py.test to always use assertion reinterpretation by passing the
+``--assert=reinterp`` option.
+
+Assert reinterpretation has a caveat not present with assert rewriting: If
+evaluating the assert expression has side effects you may get a warning that the
+intermediate values could not be determined safely.  A common example of this
+issue is an assertion which reads from a file::
+
+        assert f.read() != '...'
+
+If this assertion fails then the re-evaluation will probably succeed!
+This is because ``f.read()`` will return an empty string when it is
+called the second time during the re-evaluation.  However, it is
+easy to rewrite the assertion and avoid any trouble::
+
+        content = f.read()
+        assert content != '...'
+
+All assert introspection can be turned off by passing ``--assert=plain``.
+
+For further information, Benjamin Peterson wrote up `Behind the scenes of py.test's new assertion rewriting <http://pybites.blogspot.com/2011/07/behind-scenes-of-pytests-new-assertion.html>`_.
+
+.. versionadded:: 2.1
+   Add assert rewriting as an alternate introspection technique.
+
+.. versionchanged:: 2.1
+   Introduce the ``--assert`` option. Deprecate ``--no-assert`` and
+   ``--nomagic``.

doc/builtin.txt

@@ -1,7 +1,7 @@
 
 .. _`pytest helpers`:
 
-pytest builtin helpers
+Pytest builtin helpers
 ================================================
 
 builtin pytest.* functions and helping objects
@@ -12,43 +12,47 @@ You can always use an interactive Python prompt and type::
     import pytest
     help(pytest)
 
-to get an overview on available globally available helpers.
+to get an overview on the globally available helpers.
 
 .. automodule:: pytest
     :members:
 
-builtin function arguments
+
+.. _builtinfuncargs:
+
+Builtin function arguments
 -----------------------------------------------------
 
 You can ask for available builtin or project-custom
 :ref:`function arguments <funcargs>` by typing::
 
     $ py.test --funcargs
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collected 0 items
     pytestconfig
         the pytest config object with access to command line opts.
     capsys
-        captures writes to sys.stdout/sys.stderr and makes
-        them available successively via a ``capsys.readouterr()`` method
-        which returns a ``(out, err)`` tuple of captured snapshot strings.
-
+        enables capturing of writes to sys.stdout/sys.stderr and makes
+        captured output available via ``capsys.readouterr()`` method calls
+        which return a ``(out, err)`` tuple.
+        
     capfd
-        captures writes to file descriptors 1 and 2 and makes
-        snapshotted ``(out, err)`` string tuples available
-        via the ``capsys.readouterr()`` method.  If the underlying
-        platform does not have ``os.dup`` (e.g. Jython) tests using
-        this funcarg will automatically skip.
-
+        enables capturing of writes to file descriptors 1 and 2 and makes
+        captured output available via ``capsys.readouterr()`` method calls
+        which return a ``(out, err)`` tuple.
+        
     tmpdir
         return a temporary directory path object
-        unique to each test function invocation,
+        which is unique to each test function invocation,
         created as a sub directory of the base temporary
         directory.  The returned object is a `py.path.local`_
         path object.
-
+        
     monkeypatch
         The returned ``monkeypatch`` funcarg provides these
         helper methods to modify objects, dictionaries or os.environ::
-
+        
         monkeypatch.setattr(obj, name, value, raising=True)
         monkeypatch.delattr(obj, name, raising=True)
         monkeypatch.setitem(mapping, name, value)
@@ -56,15 +60,23 @@ You can ask for available builtin or project-custom
         monkeypatch.setenv(name, value, prepend=False)
         monkeypatch.delenv(name, value, raising=True)
         monkeypatch.syspath_prepend(path)
-
-        All modifications will be undone when the requesting
-        test function finished its execution.  The ``raising``
+        monkeypatch.chdir(path)
+        
+        All modifications will be undone after the requesting
+        test function has finished. The ``raising``
         parameter determines if a KeyError or AttributeError
         will be raised if the set/deletion operation has no target.
-
+        
     recwarn
         Return a WarningsRecorder instance that provides these methods:
-
+        
         * ``pop(category=None)``: return last warning matching the category.
         * ``clear()``: clear list of warnings
-
+        
+        See http://docs.python.org/library/warnings.html for information
+        on warning categories.
+        
+    cov
+        A pytest funcarg that provides access to the underlying coverage object.
+    
+    =============================  in 0.01 seconds =============================

doc/capture.txt

@@ -1,31 +1,36 @@
 
 .. _`captures`:
 
-Capturing of stdout/stderr output
+Capturing of the stdout/stderr output
 =========================================================
 
-By default ``stdout`` and ``stderr`` output is captured separately for
-setup and test execution code.  If a test or a setup method fails its
-according output will usually be shown along with the failure traceback.
-In addition, ``stdin`` is set to a "null" object which will fail all
-attempts to read from it.  This is important if some code paths in
-test otherwise might lead to waiting for input - which is usually
-not desired when running automated tests.
+Default stdout/stderr/stdin capturing behaviour
+---------------------------------------------------------
+
+During test execution any output sent to ``stdout`` and ``stderr`` is
+captured.  If a test or a setup method fails its according captured
+output will usually be shown along with the failure traceback.
+
+In addition, ``stdin`` is set to a "null" object which will
+fail on attempts to read from it because it is rarely desired
+to wait for interactive input when running automated tests.
+
+By default capturing is done by intercepting writes to low level
+file descriptors.  This allows to capture output from simple
+print statements as well as output from a subprocess started by
+a test.
 
 Setting capturing methods or disabling capturing
 -------------------------------------------------
 
 There are two ways in which ``py.test`` can perform capturing:
 
-* ``fd`` level capturing (default): All writes going to the operating
-  system file descriptors 1 and 2 will be captured, for example writes such
-  as ``os.write(1, 'hello')``.  Capturing on ``fd``-level also includes
-  **output from subprocesses**.
+* file descriptor (FD) level capturing (default): All writes going to the
+  operating system file descriptors 1 and 2 will be captured.
 
-* ``sys`` level capturing: The ``sys.stdout`` and ``sys.stderr`` will
-  will be replaced with in-memory files and the ``print`` builtin or
-  output from code like ``sys.stderr.write(...)`` will be captured with
-  this method.
+* ``sys`` level capturing: Only writes to Python files ``sys.stdout``
+  and ``sys.stderr`` will be captured.  No capturing of writes to
+  filedescriptors is performed.
 
 .. _`disable capturing`:
 
@@ -35,14 +40,46 @@ You can influence output capturing mechanisms from the command line::
     py.test --capture=sys # replace sys.stdout/stderr with in-mem files
     py.test --capture=fd  # also point filedescriptors 1 and 2 to temp file
 
-If you set capturing values in a conftest file like this::
+.. _printdebugging:
 
-    # conftest.py
-    option_capture = 'fd'
+Using print statements for debugging
+---------------------------------------------------
 
-then all tests in that directory will execute with "fd" style capturing.
+One primary benefit of the default capturing of stdout/stderr output
+is that you can use print statements for debugging::
 
-_ `printdebugging`:
+    # content of test_module.py
+   
+    def setup_function(function):
+        print ("setting up %s" % function)
+
+    def test_func1():
+        assert True
+
+    def test_func2():
+        assert False
+
+and running this module will show you precisely the output
+of the failing function and hide the other one::
+
+    $ py.test
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
+    
+    test_module.py .F
+    
+    ================================= FAILURES =================================
+    ________________________________ test_func2 ________________________________
+    
+        def test_func2():
+    >       assert False
+    E       assert False
+    
+    test_module.py:9: AssertionError
+    ----------------------------- Captured stdout ------------------------------
+    setting up <function test_func2 at 0x1013230c8>
+    ==================== 1 failed, 1 passed in 0.03 seconds ====================
 
 Accessing captured output from a test function
 ---------------------------------------------------

doc/conf.py

@@ -38,22 +38,21 @@ source_suffix = '.txt'
 #source_encoding = 'utf-8-sig'
 
 # The master toctree document.
-master_doc = 'index'
+master_doc = 'contents'
 
 # General information about the project.
 project = u'pytest'
-copyright = u'2010, holger krekel et aliter'
+copyright = u'2011, holger krekel et alii'
 
 # The version info for the project you're documenting, acts as replacement for
 # |version| and |release|, also used in various other places throughout the
 # built documents.
 #
-# The short X.Y version.
-version = '2.0.0'
 # The full version, including alpha/beta/rc tags.
-import py, pytest
-assert py.path.local().relto(py.path.local(pytest.__file__).dirpath().dirpath())
+import pytest
 release = pytest.__version__
+# The short X.Y version.
+version = ".".join(release.split(".")[:2])
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.
@@ -67,7 +66,10 @@ release = pytest.__version__
 
 # List of patterns, relative to source directory, that match files and
 # directories to ignore when looking for source files.
-exclude_patterns = ['links.inc', '_build', 'test', ] # XXX
+exclude_patterns = ['links.inc', '_build', 'naming20.txt', 'test/*',
+    'example/attic.txt',
+    ]
+
 
 # The reST default role (used for this markup: `text`) to use for all documents.
 #default_role = None
@@ -106,10 +108,10 @@ html_theme_options = {}
 
 # The name for this set of Sphinx documents.  If None, it defaults to
 # "<project> v<release> documentation".
-#html_title = None
+html_title = None
 
 # A shorter title for the navigation bar.  Default is the same as html_title.
-#html_short_title = None
+html_short_title = "pytest-%s" % release
 
 # The name of an image file (relative to this directory) to place at the top
 # of the sidebar.
@@ -135,7 +137,7 @@ html_static_path = ['_static']
 
 # Custom sidebar templates, maps document names to template names.
 #html_sidebars = {}
-html_sidebars = {'index': 'indexsidebar.html'}
+#html_sidebars = {'index': 'indexsidebar.html'}
 
 # Additional templates that should be rendered to pages, maps page names to
 # template names.
@@ -144,16 +146,16 @@ html_sidebars = {'index': 'indexsidebar.html'}
 
 
 # If false, no module index is generated.
-#html_domain_indices = True
+html_domain_indices = True
 
 # If false, no index is generated.
-#html_use_index = True
+html_use_index = False
 
 # If true, the index is split into individual pages for each letter.
 #html_split_index = False
 
 # If true, links to the reST sources are added to the pages.
-#html_show_sourcelink = True
+html_show_sourcelink = False
 
 # If true, "Created using Sphinx" is shown in the HTML footer. Default is True.
 #html_show_sphinx = True
@@ -184,8 +186,8 @@ htmlhelp_basename = 'pytestdoc'
 # Grouping the document tree into LaTeX files. List of tuples
 # (source start file, target name, title, author, documentclass [howto/manual]).
 latex_documents = [
-  ('index', 'pytest.tex', u'pytest Documentation',
-   u'holger krekel et aliter', 'manual'),
+  ('contents', 'pytest.tex', u'pytest Documentation',
+   u'holger krekel et alii', 'manual'),
 ]
 
 # The name of an image file (relative to this directory) to place at the top of
@@ -209,16 +211,15 @@ latex_documents = [
 #latex_appendices = []
 
 # If false, no module index is generated.
-#latex_domain_indices = True
-
+latex_domain_indices = False
 
 # -- Options for manual page output --------------------------------------------
 
 # One entry per manual page. List of tuples
 # (source start file, name, description, authors, manual section).
 man_pages = [
-    ('index', 'pytest', u'pytest Documentation',
-     [u'holger krekel et aliter'], 1)
+    ('usage', 'pytest', u'pytest usage',
+     [u'holger krekel at merlinux eu'], 1)
 ]
 
 
@@ -226,9 +227,9 @@ man_pages = [
 
 # Bibliographic Dublin Core info.
 epub_title = u'pytest'
-epub_author = u'holger krekel et aliter'
-epub_publisher = u'holger krekel et aliter'
-epub_copyright = u'2010, holger krekel et aliter'
+epub_author = u'holger krekel at merlinux eu'
+epub_publisher = u'holger krekel at merlinux eu'
+epub_copyright = u'2011, holger krekel et alii'
 
 # The language of the text. It defaults to the language option
 # or en if the language is not set.

doc/contact.txt

@@ -15,7 +15,7 @@ Contact channels
 
 - #pylib on irc.freenode.net IRC channel for random questions.
 
-- private mail to Holger.Krekel at gmail com if you have sensitive issues to communicate
+- private mail to Holger.Krekel at gmail com if you want to communicate sensitive issues
 
 - `commit mailing list`_
 

doc/contents.txt

@@ -1,19 +1,20 @@
 
 .. _toc:
 
-Table of Contents
-========================
+Full pytest documentation
+===========================
 
-.. note::
-   version 2.0 introduces :ref:`pytest as the main Python import name <naming20>`
+`Download latest version as PDF <pytest.pdf>`_
+
+.. `Download latest version as EPUB <http://media.readthedocs.org/epub/pytest/latest/pytest.epub>`_
 
 .. toctree::
    :maxdepth: 2
 
    overview
+   example/index
    apiref
    plugins
-   example/index
    talks
    develop
    announce/index
@@ -22,12 +23,4 @@ Table of Contents
    :hidden:
 
    changelog.txt
-   naming20.txt
-   example/attic
 
-Indices and tables
-==================
-
-* :ref:`genindex`
-* :ref:`modindex`
-* :ref:`search`

doc/customize.txt

@@ -1,23 +1,23 @@
-basic test configuration
+Basic test configuration
 ===================================
 
 Command line options and configuration file settings
 -----------------------------------------------------------------
 
-You can get help on options and ini-config values by running::
+You can get help on command line options and values in INI-style
+configurations files by using the general help option::
 
     py.test -h   # prints options _and_ config file settings
 
 This will display command line and configuration file settings
 which were registered by installed plugins.
 
+How test configuration is read from configuration INI-files
+-------------------------------------------------------------
 
-how test configuration is read from setup/tox ini-files
---------------------------------------------------------
-
-py.test searched for the first matching ini-style configuration file
+py.test searches for the first matching ini-style configuration file
 in the directories of command line argument and the directories above.
-It looks for filenames in this order::
+It looks for file basenames in this order::
 
     pytest.ini
     tox.ini
@@ -44,35 +44,33 @@ is used to start the search.
 .. _`how to change command line options defaults`:
 .. _`adding default options`:
 
-how to change command line options defaults
+How to change command line options defaults
 ------------------------------------------------
 
-py.test provides a simple way to set some default
-command line options.  For example, if you want
-to always see detailed info on skipped and xfailed
-tests, as well as have terser "dot progress output",
-you can add this to your root directory::
+It can be tedious to type the same series of command line options
+every time you use py.test .  For example, if you always want to see
+detailed info on skipped and xfailed tests, as well as have terser "dot"
+progress output, you can write it into a configuration file::
 
     # content of pytest.ini
     # (or tox.ini or setup.cfg)
     [pytest]
     addopts = -rsxX -q
 
-From now on, running ``py.test`` will implicitely add
-the specified options.
+From now on, running ``py.test`` will add the specified options.
 
-builtin configuration file options
+Builtin configuration file options
 ----------------------------------------------
 
 .. confval:: minversion
 
-   specifies a minimal pytest version needed for running tests.
+   Specifies a minimal pytest version required for running tests.
 
         minversion = 2.1  # will fail if we run with pytest-2.0
 
 .. confval:: addopts
 
-   add the specified ``OPTS`` to the set of command line arguments as if they
+   Add the specified ``OPTS`` to the set of command line arguments as if they
    had been specified by the user. Example: if you have this ini file content::
 
        [pytest]
@@ -96,15 +94,15 @@ builtin configuration file options
         [seq]   matches any character in seq
         [!seq]  matches any char not in seq
 
-   Default patterns are ``.* _* CVS {args}``. Setting a ``norecurse``
-   replaces the default.  Here is a customizing example for avoiding
-   a different set of directories::
+   Default patterns are ``.* _* CVS {args}``. Setting a ``norecursedir``
+   replaces the default.  Here is an example of how to avoid
+   certain directories::
 
     # content of setup.cfg
     [pytest]
     norecursedirs = .svn _build tmp*
 
-   This would tell py.test to not recurse into typical subversion or
+   This would tell py.test to not look into typical subversion or
    sphinx-build directories or into any ``tmp`` prefixed directory.
 
 .. confval:: python_files
@@ -123,4 +121,3 @@ builtin configuration file options
    and methods are considered as test modules.
 
    See :ref:`change naming conventions` for examples.
-

doc/develop.txt

@@ -22,7 +22,7 @@ download and unpack a TAR file::
 
     http://pypi.python.org/pypi/pytest/
 
-activating a checkout with setuptools
+Activating a checkout with setuptools
 --------------------------------------------
 
 With a working Distribute_ or setuptools_ installation you can type::
@@ -31,4 +31,10 @@ With a working Distribute_ or setuptools_ installation you can type::
 
 in order to work inline with the tools and the lib of your checkout.
 
+If this command complains that it could not find the required version
+of "py" then you need to use the development pypi repository::
+
+    python setup.py develop -i http://pypi.testrun.org
+
+
 .. include:: links.inc

doc/doctest.txt

@@ -1,5 +1,5 @@
 
-doctest integration for modules and test files.
+Doctest integration for modules and test files
 =========================================================
 
 By default all files matching the ``test*.txt`` pattern will
@@ -15,11 +15,11 @@ python test modules)::
     py.test --doctest-modules
 
 You can make these changes permanent in your project by
-putting them into a conftest.py file like this::
+putting them into a pytest.ini file like this::
 
-    # content of conftest.py
-    option_doctestmodules = True
-    option_doctestglob = "*.rst"
+    # content of pytest.ini
+    [pytest]
+    addopts = --doctest-modules
 
 If you then have a text file like this::
 
@@ -35,7 +35,7 @@ and another like this::
     # content of mymodule.py
     def something():
         """ a doctest in a docstring
-        >>> something()  
+        >>> something()
         42
         """
         return 42
@@ -44,7 +44,10 @@ then you can just invoke ``py.test`` without command line options::
 
     $ py.test
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: /tmp/doc-exec-66
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
     
-    =============================  in 0.00 seconds =============================
+    mymodule.py .
+    
+    ========================= 1 passed in 0.51 seconds =========================
+    [?1034h

doc/example/assertion/failure_demo.py

@@ -77,6 +77,22 @@ class TestSpecialisedExplanations(object):
     def test_in_list(self):
         assert 1 in [0, 2, 3, 4, 5]
 
+    def test_not_in_text_multiline(self):
+        text = 'some multiline\ntext\nwhich\nincludes foo\nand a\ntail'
+        assert 'foo' not in text
+
+    def test_not_in_text_single(self):
+        text = 'single foo line'
+        assert 'foo' not in text
+
+    def test_not_in_text_single_long(self):
+        text = 'head ' * 50 + 'foo ' + 'tail ' * 20
+        assert 'foo' not in text
+
+    def test_not_in_text_single_long_term(self):
+        text = 'head ' * 50 + 'f'*70 + 'tail ' * 20
+        assert 'f'*70 not in text
+
 
 def test_attribute():
     class Foo(object):

doc/example/assertion/test_failures.py

@@ -9,6 +9,6 @@ def test_failure_demo_fails_properly(testdir):
     failure_demo.copy(testdir.tmpdir.join(failure_demo.basename))
     result = testdir.runpytest(target)
     result.stdout.fnmatch_lines([
-        "*35 failed*"
+        "*39 failed*"
     ])
     assert result.ret != 0

doc/example/attic.txt

@@ -48,7 +48,7 @@ the output.
 example: decorating a funcarg in a test module
 --------------------------------------------------------------
 
-For larger scale setups it's sometimes useful to decorare
+For larger scale setups it's sometimes useful to decorate
 a funcarg just for a particular test module.  We can
 extend the `accept example`_ by putting this in our test module:
 

doc/example/index.txt

@@ -7,12 +7,17 @@ Usages and Examples
 Here is a (growing) list of examples. :ref:`Contact <contact>` us if you
 need more examples or have questions. Also take a look at the :ref:`comprehensive documentation <toc>` which contains many example snippets as well.
 
+.. note::
+
+   see :doc:`../getting-started` for basic introductory examples
+
 .. toctree::
    :maxdepth: 2
 
    reportingdemo.txt
    simple.txt
-   pythoncollection.txt
    mysetup.txt
    parametrize.txt
+   markers.txt
+   pythoncollection.txt
    nonpython.txt

doc/example/markers.txt

@@ -0,0 +1,297 @@
+
+.. _`mark examples`:
+
+Working with custom markers
+=================================================
+
+Here are some example using the :ref:`mark` mechanism.
+
+Marking test functions and selecting them for a run
+----------------------------------------------------
+
+You can "mark" a test function with custom metadata like this::
+
+    # content of test_server.py
+
+    import pytest
+    @pytest.mark.webtest
+    def test_send_http():
+        pass # perform some webtest test for your app
+    def test_something_quick():
+        pass
+
+.. versionadded:: 2.2
+
+You can then restrict a test run to only run tests marked with ``webtest``::
+
+    $ py.test -v -m webtest
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2 -- /Users/hpk/venv/0/bin/python
+    collecting ... collected 2 items
+    
+    test_server.py:3: test_send_http PASSED
+    
+    =================== 1 tests deselected by "-m 'webtest'" ===================
+    ================== 1 passed, 1 deselected in 0.01 seconds ==================
+
+Or the inverse, running all tests except the webtest ones::
+    
+    $ py.test -v -m "not webtest"
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2 -- /Users/hpk/venv/0/bin/python
+    collecting ... collected 2 items
+    
+    test_server.py:6: test_something_quick PASSED
+    
+    ================= 1 tests deselected by "-m 'not webtest'" =================
+    ================== 1 passed, 1 deselected in 0.02 seconds ==================
+
+Registering markers
+-------------------------------------
+
+.. versionadded:: 2.2
+
+.. ini-syntax for custom markers:
+
+Registering markers for your test suite is simple::
+
+    # content of pytest.ini
+    [pytest]
+    markers =
+        webtest: mark a test as a webtest.
+
+You can ask which markers exist for your test suite - the list includes our just defined ``webtest`` markers::
+
+    $ py.test --markers
+    @pytest.mark.webtest: mark a test as a webtest.
+    
+    @pytest.mark.skipif(*conditions): skip the given test function if evaluation of all conditions has a True value.  Evaluation happens within the module global context. Example: skipif('sys.platform == "win32"') skips the test if we are on the win32 platform. 
+    
+    @pytest.mark.xfail(*conditions, reason=None, run=True): mark the the test function as an expected failure. Optionally specify a reason and run=False if you don't even want to execute the test function. Any positional condition strings will be evaluated (like with skipif) and if one is False the marker will not be applied.
+    
+    @pytest.mark.parametrize(argnames, argvalues): call a test function multiple times passing in multiple different argument value sets. Example: @parametrize('arg1', [1,2]) would lead to two calls of the decorated test function, one with arg1=1 and another with arg1=2.
+    
+    @pytest.mark.tryfirst: mark a hook implementation function such that the plugin machinery will try to call it first/as early as possible.
+    
+    @pytest.mark.trylast: mark a hook implementation function such that the plugin machinery will try to call it last/as late as possible.
+    
+
+For an example on how to add and work with markers from a plugin, see
+:ref:`adding a custom marker from a plugin`.
+
+.. note::
+
+    It is recommended to explicitely register markers so that:
+
+    * there is one place in your test suite defining your markers
+
+    * asking for existing markers via ``py.test --markers`` gives good output
+
+    * typos in function markers are treated as an error if you use
+      the ``--strict`` option. Later versions of py.test are probably
+      going to treat non-registered markers as an error.
+
+.. _`scoped-marking`:
+
+Marking whole classes or modules
+----------------------------------------------------
+
+If you are programming with Python 2.6 or later you may use ``pytest.mark``
+decorators with classes to apply markers to all of its test methods::
+
+    # content of test_mark_classlevel.py
+    import pytest
+    @pytest.mark.webtest
+    class TestClass:
+        def test_startup(self):
+            pass
+        def test_startup_and_more(self):
+            pass
+
+This is equivalent to directly applying the decorator to the
+two test functions.
+
+To remain backward-compatible with Python 2.4 you can also set a
+``pytestmark`` attribute on a TestClass like this::
+
+    import pytest
+
+    class TestClass:
+        pytestmark = pytest.mark.webtest
+
+or if you need to use multiple markers you can use a list::
+
+    import pytest
+
+    class TestClass:
+        pytestmark = [pytest.mark.webtest, pytest.mark.slowtest]
+
+You can also set a module level marker::
+
+    import pytest
+    pytestmark = pytest.mark.webtest
+
+in which case it will be applied to all functions and
+methods defined in the module.
+
+
+Using ``-k TEXT`` to select tests
+----------------------------------------------------
+
+You can use the ``-k`` command line option to only run tests with names matching
+the given argument::
+
+    $ py.test -k send_http  # running with the above defined examples
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 4 items
+    
+    test_server.py .
+    
+    =================== 3 tests deselected by '-ksend_http' ====================
+    ================== 1 passed, 3 deselected in 0.02 seconds ==================
+
+And you can also run all tests except the ones that match the keyword::
+
+    $ py.test -k-send_http
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 4 items
+    
+    test_mark_classlevel.py ..
+    test_server.py .
+    
+    =================== 1 tests deselected by '-k-send_http' ===================
+    ================== 3 passed, 1 deselected in 0.03 seconds ==================
+
+Or to only select the class::
+
+    $ py.test -kTestClass
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 4 items
+    
+    test_mark_classlevel.py ..
+    
+    =================== 2 tests deselected by '-kTestClass' ====================
+    ================== 2 passed, 2 deselected in 0.03 seconds ==================
+
+.. _`adding a custom marker from a plugin`:
+
+Custom marker and command line option to control test runs
+----------------------------------------------------------
+
+.. regendoc:wipe
+
+Plugins can provide custom markers and implement specific behaviour
+based on it. This is a self-contained example which adds a command
+line option and a parametrized test function marker to run tests
+specifies via named environments::
+
+    # content of conftest.py
+
+    import pytest
+    def pytest_addoption(parser):
+        parser.addoption("-E", dest="env", action="store", metavar="NAME",
+            help="only run tests matching the environment NAME.")
+
+    def pytest_configure(config):
+        # register an additional marker
+        config.addinivalue_line("markers",
+            "env(name): mark test to run only on named environment")
+
+    def pytest_runtest_setup(item):
+        if not isinstance(item, item.Function):
+            return
+        if hasattr(item.obj, 'env'):
+            envmarker = getattr(item.obj, 'env')
+            envname = envmarker.args[0]
+            if envname != item.config.option.env:
+                pytest.skip("test requires env %r" % envname)
+
+A test file using this local plugin::
+
+    # content of test_someenv.py
+
+    import pytest
+    @pytest.mark.env("stage1")
+    def test_basic_db_operation():
+        pass
+
+and an example invocations specifying a different environment than what
+the test needs::
+
+    $ py.test -E stage2
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
+    
+    test_someenv.py s
+    
+    ======================== 1 skipped in 0.02 seconds =========================
+  
+and here is one that specifies exactly the environment needed::
+
+    $ py.test -E stage1
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
+    
+    test_someenv.py .
+    
+    ========================= 1 passed in 0.02 seconds =========================
+
+The ``--markers`` option always gives you a list of available markers::
+
+    $ py.test --markers
+    @pytest.mark.env(name): mark test to run only on named environment
+    
+    @pytest.mark.skipif(*conditions): skip the given test function if evaluation of all conditions has a True value.  Evaluation happens within the module global context. Example: skipif('sys.platform == "win32"') skips the test if we are on the win32 platform. 
+    
+    @pytest.mark.xfail(*conditions, reason=None, run=True): mark the the test function as an expected failure. Optionally specify a reason and run=False if you don't even want to execute the test function. Any positional condition strings will be evaluated (like with skipif) and if one is False the marker will not be applied.
+    
+    @pytest.mark.parametrize(argnames, argvalues): call a test function multiple times passing in multiple different argument value sets. Example: @parametrize('arg1', [1,2]) would lead to two calls of the decorated test function, one with arg1=1 and another with arg1=2.
+    
+    @pytest.mark.tryfirst: mark a hook implementation function such that the plugin machinery will try to call it first/as early as possible.
+    
+    @pytest.mark.trylast: mark a hook implementation function such that the plugin machinery will try to call it last/as late as possible.
+
+    
+Reading markers which were set from multiple places
+----------------------------------------------------
+
+.. versionadded: 2.2.2
+
+If you are heavily using markers in your test suite you may encounter the case where a marker is applied several times to a test function.  From plugin
+code you can read over all such settings.  Example::
+
+    # content of test_mark_three_times.py
+    import pytest
+    pytestmark = pytest.mark.glob("module", x=1)
+
+    @pytest.mark.glob("class", x=2)
+    class TestClass:
+        @pytest.mark.glob("function", x=3)
+        def test_something(self):
+            pass
+
+Here we have the marker "glob" applied three times to the same
+test function.  From a conftest file we can read it like this::
+
+    # content of conftest.py
+
+    def pytest_runtest_setup(item):
+        g = getattr(item.obj, 'glob', None)
+        if g is not None:
+            for info in g:
+                print ("glob args=%s kwargs=%s" %(info.args, info.kwargs))
+
+Let's run this without capturing output and see what we get::
+
+    $ py.test -q -s 
+    collecting ... collected 2 items
+    ..
+    2 passed in 0.02 seconds
+    glob args=('function',) kwargs={'x': 3}
+    glob args=('class',) kwargs={'x': 2}
+    glob args=('module',) kwargs={'x': 1}

doc/example/multipython.py

@@ -2,20 +2,20 @@
 module containing a parametrized tests testing cross-python
 serialization via the pickle module.
 """
-import py
+import py, pytest
 
 pythonlist = ['python2.4', 'python2.5', 'python2.6', 'python2.7', 'python2.8']
 
 def pytest_generate_tests(metafunc):
-    if 'python1' in metafunc.funcargnames:
-        assert 'python2' in metafunc.funcargnames
-        for obj in metafunc.function.multiarg.kwargs['obj']:
-            for py1 in pythonlist:
-                for py2 in pythonlist:
-                    metafunc.addcall(id="%s-%s-%s" % (py1, py2, obj),
-                        param=(py1, py2, obj))
+    # we parametrize all "python1" and "python2" arguments to iterate
+    # over the python interpreters of our list above - the actual
+    # setup and lookup of interpreters in the python1/python2 factories
+    # respectively.
+    for arg in metafunc.funcargnames:
+        if arg in ("python1", "python2"):
+            metafunc.parametrize(arg, pythonlist, indirect=True)
 
-@py.test.mark.multiarg(obj=[42, {}, {1:3},])
+@pytest.mark.parametrize("obj", [42, {}, {1:3},])
 def test_basic_objects(python1, python2, obj):
     python1.dumps(obj)
     python2.load_and_is_true("obj == %s" % obj)
@@ -23,14 +23,11 @@ def test_basic_objects(python1, python2, obj):
 def pytest_funcarg__python1(request):
     tmpdir = request.getfuncargvalue("tmpdir")
     picklefile = tmpdir.join("data.pickle")
-    return Python(request.param[0], picklefile)
+    return Python(request.param, picklefile)
 
 def pytest_funcarg__python2(request):
     python1 = request.getfuncargvalue("python1")
-    return Python(request.param[1], python1.picklefile)
-
-def pytest_funcarg__obj(request):
-    return request.param[2]
+    return Python(request.param, python1.picklefile)
 
 class Python:
     def __init__(self, version, picklefile):

doc/example/mysetup.txt

@@ -3,7 +3,7 @@
 
 .. _mysetup:
 
-mysetup pattern: application specific test fixtures
+Mysetup pattern: application specific test fixtures
 ==========================================================
 
 Here is a basic useful step-by-step example for managing and interacting
@@ -12,8 +12,8 @@ where we have the glue and test support code for bootstrapping and
 configuring application objects and allow test modules and test
 functions to stay ignorant of involved details.
 
-step1: implementing the test/app-specific ``mysetup`` pattern
--------------------------------------------------------------
+Step 1: Implementing the test/app-specific ``mysetup`` pattern
+--------------------------------------------------------------
 
 Let's write a simple test function using a ``mysetup`` funcarg::
 
@@ -26,7 +26,7 @@ Let's write a simple test function using a ``mysetup`` funcarg::
 To run this test py.test needs to find and call a factory to
 obtain the required ``mysetup`` function argument.  To make
 an according factory findable we write down a specifically named factory
-method in a :ref:`local plugin`::
+method in a :ref:`local plugin <localplugin>` ::
 
     # content of conftest.py
     from myapp import MyApp
@@ -38,7 +38,7 @@ method in a :ref:`local plugin`::
         def myapp(self):
             return MyApp()
 
-To run the example we stub out a a simple ``MyApp`` application object::
+To run the example we stub out a simple ``MyApp`` application object::
 
     # content of myapp.py
     class MyApp:
@@ -49,15 +49,15 @@ You can now run the test::
 
     $ py.test test_sample.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_sample.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
     
     test_sample.py F
     
     ================================= FAILURES =================================
     _______________________________ test_answer ________________________________
     
-    mysetup = <conftest.MySetup instance at 0x16f5998>
+    mysetup = <conftest.MySetup instance at 0x101322fc8>
     
         def test_answer(mysetup):
             app = mysetup.myapp()
@@ -66,7 +66,7 @@ You can now run the test::
     E       assert 54 == 42
     
     test_sample.py:4: AssertionError
-    ========================= 1 failed in 0.02 seconds =========================
+    ========================= 1 failed in 0.72 seconds =========================
 
 This means that our ``mysetup`` object was successfully instantiated
 and ``mysetup.app()`` returned an initialized ``MyApp`` instance.
@@ -76,7 +76,7 @@ the concrete question or answers actually mean, please see here_.
 .. _here: http://uncyclopedia.wikia.com/wiki/The_Hitchhiker's_Guide_to_the_Galaxy
 .. _`tut-cmdlineoption`:
 
-step 2: checking a command line option and skipping tests
+Step 2: Checking a command line option and skipping tests
 -----------------------------------------------------------
 
 To add a command line option we update the ``conftest.py`` of
@@ -122,12 +122,12 @@ Running it yields::
 
     $ py.test test_ssh.py -rs
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_ssh.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
     
     test_ssh.py s
     ========================= short test summary info ==========================
-    SKIP [1] /tmp/doc-exec-107/conftest.py:22: specify ssh host with --ssh
+    SKIP [1] /Users/hpk/tmp/doc-exec-153/conftest.py:22: specify ssh host with --ssh
     
     ======================== 1 skipped in 0.02 seconds =========================
 

doc/example/nonpython.txt

@@ -6,7 +6,7 @@ Working with non-python tests
 
 .. _`yaml plugin`:
 
-a basic example for specifying tests in Yaml files
+A basic example for specifying tests in Yaml files
 --------------------------------------------------------------
 
 .. _`pytest-yamlwsgi`: http://bitbucket.org/aafshar/pytest-yamlwsgi/src/tip/pytest_yamlwsgi.py
@@ -27,8 +27,8 @@ now execute the test specification::
 
     nonpython $ py.test test_simple.yml
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_simple.yml
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
     
     test_simple.yml .F
     
@@ -37,9 +37,7 @@ now execute the test specification::
     usecase execution failed
        spec failed: 'some': 'other'
        no further details known at this point.
-    ========================= short test summary info ==========================
-    FAIL test_simple.yml::hello
-    ==================== 1 failed, 1 passed in 0.06 seconds ====================
+    ==================== 1 failed, 1 passed in 0.48 seconds ====================
 
 You get one dot for the passing ``sub1: sub1`` check and one failure.
 Obviously in the above ``conftest.py`` you'll want to implement a more
@@ -53,13 +51,13 @@ your own domain specific testing language this way.
     representation string of your choice.  It
     will be reported as a (red) string.
 
-``reportinfo()`` is used for representing the test location and is also consulted for
-reporting in ``verbose`` mode::
+``reportinfo()`` is used for representing the test location and is also
+consulted when reporting in ``verbose`` mode::
 
     nonpython $ py.test -v
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30 -- /home/hpk/venv/0/bin/python
-    test path 1: /home/hpk/p/pytest/doc/example/nonpython
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2 -- /Users/hpk/venv/0/bin/python
+    collecting ... collected 2 items
     
     test_simple.yml:1: usecase: ok PASSED
     test_simple.yml:1: usecase: hello FAILED
@@ -69,14 +67,17 @@ reporting in ``verbose`` mode::
     usecase execution failed
        spec failed: 'some': 'other'
        no further details known at this point.
-    ========================= short test summary info ==========================
-    FAIL test_simple.yml::hello
-    ==================== 1 failed, 1 passed in 0.06 seconds ====================
+    ==================== 1 failed, 1 passed in 0.10 seconds ====================
 
 While developing your custom test collection and execution it's also
 interesting to just look at the collection tree::
 
     nonpython $ py.test --collectonly
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
     <YamlFile 'test_simple.yml'>
       <YamlItem 'ok'>
       <YamlItem 'hello'>
+    
+    =============================  in 0.18 seconds =============================

doc/example/parametrize.txt

@@ -1,34 +1,283 @@
 
-parametrizing tests
+.. _paramexamples:
+
+Parametrizing tests
 =================================================
 
-py.test allows to easily implement your own custom
-parametrization scheme for tests.  Here we provide
-some examples for inspiration and re-use.
+.. currentmodule:: _pytest.python
+
+py.test allows to easily parametrize test functions.
+In the following we provide some examples using
+the builtin mechanisms.
+
+.. _parametrizemark:
+
+Simple "decorator" parametrization of a test function
+----------------------------------------------------------------------------
+
+.. versionadded:: 2.2
+
+The builtin ``pytest.mark.parametrize`` decorator directly enables
+parametrization of arguments for a test function.  Here is an example
+of a test function that wants to compare that processing some input
+results in expected output::
+
+    # content of test_expectation.py
+    import pytest
+    @pytest.mark.parametrize(("input", "expected"), [
+        ("3+5", 8),
+        ("2+4", 6),
+        ("6*9", 42),
+    ])
+    def test_eval(input, expected):
+        assert eval(input) == expected
+
+we parametrize two arguments of the test function so that the test
+function is called three times.  Let's run it::
+
+    $ py.test -q 
+    collecting ... collected 3 items
+    ..F
+    ================================= FAILURES =================================
+    ____________________________ test_eval[6*9-42] _____________________________
+    
+    input = '6*9', expected = 42
+    
+        @pytest.mark.parametrize(("input", "expected"), [
+            ("3+5", 8),
+            ("2+4", 6),
+            ("6*9", 42),
+        ])
+        def test_eval(input, expected):
+    >       assert eval(input) == expected
+    E       assert 54 == 42
+    E        +  where 54 = eval('6*9')
+    
+    test_expectation.py:8: AssertionError
+    1 failed, 2 passed in 0.03 seconds
+
+As expected only one pair of input/output values fails the simple test function.
+
+Note that there are various ways how you can mark groups of functions,
+see :ref:`mark`.
+
+Generating parameters combinations, depending on command line
+----------------------------------------------------------------------------
+
+.. regendoc:wipe
+
+Let's say we want to execute a test with different computation
+parameters and the parameter range shall be determined by a command
+line argument.  Let's first write a simple (do-nothing) computation test::
+
+    # content of test_compute.py
+
+    def test_compute(param1):
+        assert param1 < 4
+
+Now we add a test configuration like this::
+
+    # content of conftest.py
+
+    def pytest_addoption(parser):
+        parser.addoption("--all", action="store_true",
+            help="run all combinations")
+
+    def pytest_generate_tests(metafunc):
+        if 'param1' in metafunc.funcargnames:
+            if metafunc.config.option.all:
+                end = 5
+            else:
+                end = 2
+            metafunc.parametrize("param1", range(end))
+
+This means that we only run 2 tests if we do not pass ``--all``::
+
+    $ py.test -q test_compute.py
+    collecting ... collected 2 items
+    ..
+    2 passed in 0.03 seconds
+
+We run only two computations, so we see two dots.
+let's run the full monty::
+
+    $ py.test -q --all
+    collecting ... collected 5 items
+    ....F
+    ================================= FAILURES =================================
+    _____________________________ test_compute[4] ______________________________
+    
+    param1 = 4
+    
+        def test_compute(param1):
+    >       assert param1 < 4
+    E       assert 4 < 4
+    
+    test_compute.py:3: AssertionError
+    1 failed, 4 passed in 0.05 seconds
+
+As expected when running the full range of ``param1`` values
+we'll get an error on the last one.
+
+A quick port of "testscenarios"
+------------------------------------
+
+.. _`test scenarios`: http://bazaar.launchpad.net/~lifeless/testscenarios/trunk/annotate/head%3A/doc/example.py
+
+Here is a quick port to run tests configured with `test scenarios`_,
+an add-on from Robert Collins for the standard unittest framework. We
+only have to work a bit to construct the correct arguments for pytest's
+:py:func:`Metafunc.parametrize`::
+
+    # content of test_scenarios.py
+
+    def pytest_generate_tests(metafunc):
+        idlist = []
+        argvalues = []
+        for scenario in metafunc.cls.scenarios:
+            idlist.append(scenario[0])
+            items = scenario[1].items()
+            argnames = [x[0] for x in items]
+            argvalues.append(([x[1] for x in items]))
+        metafunc.parametrize(argnames, argvalues, ids=idlist)
+
+    scenario1 = ('basic', {'attribute': 'value'})
+    scenario2 = ('advanced', {'attribute': 'value2'})
+
+    class TestSampleWithScenarios:
+        scenarios = [scenario1, scenario2]
+
+        def test_demo(self, attribute):
+            assert isinstance(attribute, str)
+
+this is a fully self-contained example which you can run with::
+
+    $ py.test test_scenarios.py
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
+    
+    test_scenarios.py ..
+    
+    ========================= 2 passed in 0.02 seconds =========================
+
+If you just collect tests you'll also nicely see 'advanced' and 'basic' as variants for the test function::
+
+
+    $ py.test --collectonly test_scenarios.py
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
+    <Module 'test_scenarios.py'>
+      <Class 'TestSampleWithScenarios'>
+        <Instance '()'>
+          <Function 'test_demo[basic]'>
+          <Function 'test_demo[advanced]'>
+    
+    =============================  in 0.05 seconds =============================
+
+Deferring the setup of parametrized resources
+---------------------------------------------------
+
+.. regendoc:wipe
+
+The parametrization of test functions happens at collection
+time.  It is a good idea to setup expensive resources like DB
+connections or subprocess only when the actual test is run.
+Here is a simple example how you can achieve that, first
+the actual test requiring a ``db`` object::
+
+    # content of test_backends.py
+
+    import pytest
+    def test_db_initialized(db):
+        # a dummy test
+        if db.__class__.__name__ == "DB2":
+            pytest.fail("deliberately failing for demo purposes")
+
+We can now add a test configuration that generates two invocations of
+the ``test_db_initialized`` function and also implements a factory that
+creates a database object for the actual test invocations::
+
+    # content of conftest.py
+
+    def pytest_generate_tests(metafunc):
+        if 'db' in metafunc.funcargnames:
+            metafunc.parametrize("db", ['d1', 'd2'], indirect=True)
+
+    class DB1:
+        "one database object"
+    class DB2:
+        "alternative database object"
+
+    def pytest_funcarg__db(request):
+        if request.param == "d1":
+            return DB1()
+        elif request.param == "d2":
+            return DB2()
+        else:
+            raise ValueError("invalid internal test config")
+
+Let's first see how it looks like at collection time::
+
+    $ py.test test_backends.py --collectonly
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
+    <Module 'test_backends.py'>
+      <Function 'test_db_initialized[d1]'>
+      <Function 'test_db_initialized[d2]'>
+    
+    =============================  in 0.01 seconds =============================
+
+And then when we run the test::
+
+    $ py.test -q test_backends.py
+    collecting ... collected 2 items
+    .F
+    ================================= FAILURES =================================
+    _________________________ test_db_initialized[d2] __________________________
+    
+    db = <conftest.DB2 instance at 0x101323710>
+    
+        def test_db_initialized(db):
+            # a dummy test
+            if db.__class__.__name__ == "DB2":
+    >           pytest.fail("deliberately failing for demo purposes")
+    E           Failed: deliberately failing for demo purposes
+    
+    test_backends.py:6: Failed
+    1 failed, 1 passed in 0.03 seconds
+
+The first invocation with ``db == "DB1"`` passed while the second with ``db == "DB2"`` failed.  Our ``pytest_funcarg__db`` factory has instantiated each of the DB values during the setup phase while the ``pytest_generate_tests`` generated two according calls to the ``test_db_initialized`` during the collection phase.
+
+.. regendoc:wipe
 
 Parametrizing test methods through per-class configuration
 --------------------------------------------------------------
 
 .. _`unittest parameterizer`: http://code.google.com/p/unittest-ext/source/browse/trunk/params.py
 
+
 Here is an example ``pytest_generate_function`` function implementing a
-parametrization scheme similar to Michael Foords `unittest
-parameterizer`_ in a lot less code::
+parametrization scheme similar to Michael Foord's `unittest
+parameterizer`_ but in a lot less code::
 
     # content of ./test_parametrize.py
     import pytest
 
     def pytest_generate_tests(metafunc):
         # called once per each test function
-        for funcargs in metafunc.cls.params[metafunc.function.__name__]:
-            # schedule a new test function run with applied **funcargs
-            metafunc.addcall(funcargs=funcargs)
+        funcarglist = metafunc.cls.params[metafunc.function.__name__]
+        argnames = list(funcarglist[0])
+        metafunc.parametrize(argnames, [[funcargs[name] for name in argnames]
+                for funcargs in funcarglist])
 
     class TestClass:
         # a map specifying multiple argument sets for a test method
         params = {
             'test_equals': [dict(a=1, b=2), dict(a=3, b=3), ],
-            'test_zerodivision': [dict(a=1, b=0), dict(a=3, b=2)],
+            'test_zerodivision': [dict(a=1, b=0), ],
         }
 
         def test_equals(self, a, b):
@@ -37,106 +286,44 @@ parameterizer`_ in a lot less code::
         def test_zerodivision(self, a, b):
             pytest.raises(ZeroDivisionError, "a/b")
 
-Running it means we are two tests for each test functions, using
-the respective settings::
+Our test generator looks up a class-level definition which specifies which
+argument sets to use for each test function.  Let's run it::
 
     $ py.test -q
-    F..F
+    collecting ... collected 3 items
+    F..
     ================================= FAILURES =================================
-    _________________________ TestClass.test_equals[0] _________________________
+    ________________________ TestClass.test_equals[1-2] ________________________
     
-    self = <test_parametrize.TestClass instance at 0x128a638>, a = 1, b = 2
+    self = <test_parametrize.TestClass instance at 0x101326368>, a = 1, b = 2
     
         def test_equals(self, a, b):
     >       assert a == b
     E       assert 1 == 2
     
-    test_parametrize.py:17: AssertionError
-    ______________________ TestClass.test_zerodivision[1] ______________________
-    
-    self = <test_parametrize.TestClass instance at 0x1296440>, a = 3, b = 2
-    
-        def test_zerodivision(self, a, b):
-    >       pytest.raises(ZeroDivisionError, "a/b")
-    E       Failed: DID NOT RAISE
-    
-    test_parametrize.py:20: Failed
-    2 failed, 2 passed in 0.03 seconds
+    test_parametrize.py:18: AssertionError
+    1 failed, 2 passed in 0.03 seconds
 
-Parametrizing test methods through a decorator
---------------------------------------------------------------
-
-Modifying the previous example we can also allow decorators
-for parametrizing test methods::
-
-    # content of test_parametrize2.py
-
-    import pytest
-
-    # test support code
-    def params(funcarglist):
-        def wrapper(function):
-            function.funcarglist = funcarglist
-            return function
-        return wrapper
-
-    def pytest_generate_tests(metafunc):
-        for funcargs in getattr(metafunc.function, 'funcarglist', ()):
-            metafunc.addcall(funcargs=funcargs)
-
-    # actual test code
-    class TestClass:
-        @params([dict(a=1, b=2), dict(a=3, b=3), ])
-        def test_equals(self, a, b):
-            assert a == b
-
-        @params([dict(a=1, b=0), dict(a=3, b=2)])
-        def test_zerodivision(self, a, b):
-            pytest.raises(ZeroDivisionError, "a/b")
-
-Running it gives similar results as before::
-
-    $ py.test -q test_parametrize2.py
-    F..F
-    ================================= FAILURES =================================
-    _________________________ TestClass.test_equals[0] _________________________
-    
-    self = <test_parametrize2.TestClass instance at 0x1dbcc68>, a = 1, b = 2
-    
-        @params([dict(a=1, b=2), dict(a=3, b=3), ])
-        def test_equals(self, a, b):
-    >       assert a == b
-    E       assert 1 == 2
-    
-    test_parametrize2.py:19: AssertionError
-    ______________________ TestClass.test_zerodivision[1] ______________________
-    
-    self = <test_parametrize2.TestClass instance at 0x1dd0488>, a = 3, b = 2
-    
-        @params([dict(a=1, b=0), dict(a=3, b=2)])
-        def test_zerodivision(self, a, b):
-    >       pytest.raises(ZeroDivisionError, "a/b")
-    E       Failed: DID NOT RAISE
-    
-    test_parametrize2.py:23: Failed
-    2 failed, 2 passed in 0.03 seconds
-
-checking serialization between Python interpreters
+Indirect parametrization with multiple resources
 --------------------------------------------------------------
 
 Here is a stripped down real-life example of using parametrized
-testing for testing serialization betwee different interpreters.
+testing for testing serialization, invoking different python interpreters.
 We define a ``test_basic_objects`` function which is to be run
-with different sets of arguments for its three arguments::
+with different sets of arguments for its three arguments:
 
-* ``python1``: first python interpreter
-* ``python2``: second python interpreter
-* ``obj``: object to be dumped from first interpreter and loaded into second interpreter
+* ``python1``: first python interpreter, run to pickle-dump an object to a file
+* ``python2``: second interpreter, run to pickle-load an object from a file
+* ``obj``: object to be dumped/loaded
 
 .. literalinclude:: multipython.py
 
-Running it (with Python-2.4 through to Python2.7 installed)::
+Running it results in some skips if we don't have all the python interpreters installed and otherwise runs all combinations (5 interpreters times 5 interpreters times 3 objects to serialize/deserialize)::
 
-   . $ py.test -q multipython.py
-   ....s....s....s....ssssss....s....s....s....ssssss....s....s....s....ssssss
-   48 passed, 27 skipped in 2.55 seconds
+   . $ py.test -rs -q multipython.py
+   collecting ... collected 75 items
+   ssssssssssssssssss.........ssssss.........ssssss.........ssssssssssssssssss
+   ========================= short test summary info ==========================
+   SKIP [24] /Users/hpk/p/pytest/doc/example/multipython.py:36: 'python2.8' not found
+   SKIP [24] /Users/hpk/p/pytest/doc/example/multipython.py:36: 'python2.4' not found
+   27 passed, 48 skipped in 7.76 seconds

doc/example/pythoncollection.txt

@@ -1,7 +1,7 @@
 Changing standard (Python) test discovery
 ===============================================
 
-changing directory recursion
+Changing directory recursion
 -----------------------------------------------------
 
 You can set the :confval:`norecursedirs` option in an ini-file, for example your ``setup.cfg`` in the project root directory::
@@ -14,7 +14,7 @@ This would tell py.test to not recurse into typical subversion or sphinx-build d
 
 .. _`change naming conventions`:
 
-change naming conventions
+Changing naming conventions
 -----------------------------------------------------
 
 You can configure different naming conventions by setting
@@ -42,13 +42,18 @@ in functions and classes.  For example, if we have::
 then the test collection looks like this::
 
     $ py.test --collectonly
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
     <Module 'check_myapp.py'>
       <Class 'CheckMyApp'>
         <Instance '()'>
           <Function 'check_simple'>
           <Function 'check_complex'>
+    
+    =============================  in 0.01 seconds =============================
 
-interpret cmdline arguments as Python packages
+Interpreting cmdline arguments as Python packages
 -----------------------------------------------------
 
 You can use the ``--pyargs`` option to make py.test try
@@ -70,15 +75,20 @@ Now a simple invocation of ``py.test NAME`` will check
 if NAME exists as an importable package/module and otherwise
 treat it as a filesystem path.
 
-finding out what is collected
+Finding out what is collected
 -----------------------------------------------
 
 You can always peek at the collection tree without running tests like this::
 
     . $ py.test --collectonly pythoncollection.py
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 3 items
     <Module 'pythoncollection.py'>
       <Function 'test_function'>
       <Class 'TestClass'>
         <Instance '()'>
           <Function 'test_method'>
           <Function 'test_anothermethod'>
+    
+    =============================  in 0.01 seconds =============================

doc/example/reportingdemo.txt

@@ -13,11 +13,10 @@ get on the terminal - we are working on that):
 
     assertion $ py.test failure_demo.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev38
-    collecting ...
-    collected 35 items
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 39 items
     
-    failure_demo.py FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
+    failure_demo.py FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
     
     ================================= FAILURES =================================
     ____________________________ test_generative[0] ____________________________
@@ -31,7 +30,7 @@ get on the terminal - we are working on that):
     failure_demo.py:15: AssertionError
     _________________________ TestFailing.test_simple __________________________
     
-    self = <failure_demo.TestFailing object at 0x2c9da90>
+    self = <failure_demo.TestFailing object at 0x101490690>
     
         def test_simple(self):
             def f():
@@ -41,45 +40,45 @@ get on the terminal - we are working on that):
         
     >       assert f() == g()
     E       assert 42 == 43
-    E        +  where 42 = <function f at 0x2c447d0>()
-    E        +  and   43 = <function g at 0x2c44cf8>()
+    E        +  where 42 = <function f at 0x101462b90>()
+    E        +  and   43 = <function g at 0x101462c08>()
     
     failure_demo.py:28: AssertionError
     ____________________ TestFailing.test_simple_multiline _____________________
     
-    self = <failure_demo.TestFailing object at 0x2c9dc90>
+    self = <failure_demo.TestFailing object at 0x101490b10>
     
         def test_simple_multiline(self):
             otherfunc_multi(
                       42,
     >                 6*9)
     
-    failure_demo.py:33:
-    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+    failure_demo.py:33: 
+    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
     
     a = 42, b = 54
     
         def otherfunc_multi(a,b):
-            assert (a ==
-    >               b)
+    >       assert (a ==
+                    b)
     E       assert 42 == 54
     
-    failure_demo.py:12: AssertionError
+    failure_demo.py:11: AssertionError
     ___________________________ TestFailing.test_not ___________________________
     
-    self = <failure_demo.TestFailing object at 0x2c93f10>
+    self = <failure_demo.TestFailing object at 0x101490210>
     
         def test_not(self):
             def f():
                 return 42
     >       assert not f()
     E       assert not 42
-    E        +  where 42 = <function f at 0x2ca1050>()
+    E        +  where 42 = <function f at 0x101462aa0>()
     
     failure_demo.py:38: AssertionError
     _________________ TestSpecialisedExplanations.test_eq_text _________________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2c9d9d0>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x101490a10>
     
         def test_eq_text(self):
     >       assert 'spam' == 'eggs'
@@ -90,7 +89,7 @@ get on the terminal - we are working on that):
     failure_demo.py:42: AssertionError
     _____________ TestSpecialisedExplanations.test_eq_similar_text _____________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2a04e90>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148d9d0>
     
         def test_eq_similar_text(self):
     >       assert 'foo 1 bar' == 'foo 2 bar'
@@ -103,7 +102,7 @@ get on the terminal - we are working on that):
     failure_demo.py:45: AssertionError
     ____________ TestSpecialisedExplanations.test_eq_multiline_text ____________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2c9d710>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148d590>
     
         def test_eq_multiline_text(self):
     >       assert 'foo\nspam\nbar' == 'foo\neggs\nbar'
@@ -116,13 +115,13 @@ get on the terminal - we are working on that):
     failure_demo.py:48: AssertionError
     ______________ TestSpecialisedExplanations.test_eq_long_text _______________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2c9db10>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148dc90>
     
         def test_eq_long_text(self):
             a = '1'*100 + 'a' + '2'*100
             b = '1'*100 + 'b' + '2'*100
     >       assert a == b
-    E       assert '111111111111...2222222222222' == '111111111111...2222222222222'
+    E       assert '111111111111...2222222222222' == '1111111111111...2222222222222'
     E         Skipping 90 identical leading characters in diff
     E         Skipping 91 identical trailing characters in diff
     E         - 1111111111a222222222
@@ -133,13 +132,13 @@ get on the terminal - we are working on that):
     failure_demo.py:53: AssertionError
     _________ TestSpecialisedExplanations.test_eq_long_text_multiline __________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2caf950>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148d910>
     
         def test_eq_long_text_multiline(self):
             a = '1\n'*100 + 'a' + '2\n'*100
             b = '1\n'*100 + 'b' + '2\n'*100
     >       assert a == b
-    E       assert '1\n1\n1\n1\n...n2\n2\n2\n2\n' == '1\n1\n1\n1\n...n2\n2\n2\n2\n'
+    E       assert '1\n1\n1\n1\n...n2\n2\n2\n2\n' == '1\n1\n1\n1\n1...n2\n2\n2\n2\n'
     E         Skipping 190 identical leading characters in diff
     E         Skipping 191 identical trailing characters in diff
     E           1
@@ -157,7 +156,7 @@ get on the terminal - we are working on that):
     failure_demo.py:58: AssertionError
     _________________ TestSpecialisedExplanations.test_eq_list _________________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2caf590>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148b9d0>
     
         def test_eq_list(self):
     >       assert [0, 1, 2] == [0, 1, 3]
@@ -167,7 +166,7 @@ get on the terminal - we are working on that):
     failure_demo.py:61: AssertionError
     ______________ TestSpecialisedExplanations.test_eq_list_long _______________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2c9e310>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148b750>
     
         def test_eq_list_long(self):
             a = [0]*100 + [1] + [3]*100
@@ -179,7 +178,7 @@ get on the terminal - we are working on that):
     failure_demo.py:66: AssertionError
     _________________ TestSpecialisedExplanations.test_eq_dict _________________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2c9dc50>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148bdd0>
     
         def test_eq_dict(self):
     >       assert {'a': 0, 'b': 1} == {'a': 0, 'b': 2}
@@ -192,7 +191,7 @@ get on the terminal - we are working on that):
     failure_demo.py:69: AssertionError
     _________________ TestSpecialisedExplanations.test_eq_set __________________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2cafc10>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148b1d0>
     
         def test_eq_set(self):
     >       assert set([0, 10, 11, 12]) == set([0, 20, 21])
@@ -208,7 +207,7 @@ get on the terminal - we are working on that):
     failure_demo.py:72: AssertionError
     _____________ TestSpecialisedExplanations.test_eq_longer_list ______________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2cba890>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148bf10>
     
         def test_eq_longer_list(self):
     >       assert [1,2] == [1,2,3]
@@ -218,13 +217,70 @@ get on the terminal - we are working on that):
     failure_demo.py:75: AssertionError
     _________________ TestSpecialisedExplanations.test_in_list _________________
     
-    self = <failure_demo.TestSpecialisedExplanations object at 0x2cba6d0>
+    self = <failure_demo.TestSpecialisedExplanations object at 0x10148b390>
     
         def test_in_list(self):
     >       assert 1 in [0, 2, 3, 4, 5]
     E       assert 1 in [0, 2, 3, 4, 5]
     
     failure_demo.py:78: AssertionError
+    __________ TestSpecialisedExplanations.test_not_in_text_multiline __________
+    
+    self = <failure_demo.TestSpecialisedExplanations object at 0x101483e50>
+    
+        def test_not_in_text_multiline(self):
+            text = 'some multiline\ntext\nwhich\nincludes foo\nand a\ntail'
+    >       assert 'foo' not in text
+    E       assert 'foo' not in 'some multiline\ntext\nw...ncludes foo\nand a\ntail'
+    E         'foo' is contained here:
+    E           some multiline
+    E           text
+    E           which
+    E           includes foo
+    E         ?          +++
+    E           and a
+    E           tail
+    
+    failure_demo.py:82: AssertionError
+    ___________ TestSpecialisedExplanations.test_not_in_text_single ____________
+    
+    self = <failure_demo.TestSpecialisedExplanations object at 0x101483c10>
+    
+        def test_not_in_text_single(self):
+            text = 'single foo line'
+    >       assert 'foo' not in text
+    E       assert 'foo' not in 'single foo line'
+    E         'foo' is contained here:
+    E           single foo line
+    E         ?        +++
+    
+    failure_demo.py:86: AssertionError
+    _________ TestSpecialisedExplanations.test_not_in_text_single_long _________
+    
+    self = <failure_demo.TestSpecialisedExplanations object at 0x101483ed0>
+    
+        def test_not_in_text_single_long(self):
+            text = 'head ' * 50 + 'foo ' + 'tail ' * 20
+    >       assert 'foo' not in text
+    E       assert 'foo' not in 'head head head head hea...ail tail tail tail tail '
+    E         'foo' is contained here:
+    E           head head foo tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail 
+    E         ?           +++
+    
+    failure_demo.py:90: AssertionError
+    ______ TestSpecialisedExplanations.test_not_in_text_single_long_term _______
+    
+    self = <failure_demo.TestSpecialisedExplanations object at 0x101483310>
+    
+        def test_not_in_text_single_long_term(self):
+            text = 'head ' * 50 + 'f'*70 + 'tail ' * 20
+    >       assert 'f'*70 not in text
+    E       assert 'fffffffffff...ffffffffffff' not in 'head head he...l tail tail '
+    E         'ffffffffffffffffff...fffffffffffffffffff' is contained here:
+    E           head head fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffftail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail tail 
+    E         ?           ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+    
+    failure_demo.py:94: AssertionError
     ______________________________ test_attribute ______________________________
     
         def test_attribute():
@@ -233,9 +289,9 @@ get on the terminal - we are working on that):
             i = Foo()
     >       assert i.b == 2
     E       assert 1 == 2
-    E        +  where 1 = <failure_demo.Foo object at 0x2c9d750>.b
+    E        +  where 1 = <failure_demo.Foo object at 0x101483f50>.b
     
-    failure_demo.py:85: AssertionError
+    failure_demo.py:101: AssertionError
     _________________________ test_attribute_instance __________________________
     
         def test_attribute_instance():
@@ -243,10 +299,10 @@ get on the terminal - we are working on that):
                 b = 1
     >       assert Foo().b == 2
     E       assert 1 == 2
-    E        +  where 1 = <failure_demo.Foo object at 0x2cafdd0>.b
-    E        +    where <failure_demo.Foo object at 0x2cafdd0> = <class 'failure_demo.Foo'>()
+    E        +  where 1 = <failure_demo.Foo object at 0x101483210>.b
+    E        +    where <failure_demo.Foo object at 0x101483210> = <class 'failure_demo.Foo'>()
     
-    failure_demo.py:91: AssertionError
+    failure_demo.py:107: AssertionError
     __________________________ test_attribute_failure __________________________
     
         def test_attribute_failure():
@@ -257,16 +313,16 @@ get on the terminal - we are working on that):
             i = Foo()
     >       assert i.b == 2
     
-    failure_demo.py:100:
-    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+    failure_demo.py:116: 
+    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
     
-    self = <failure_demo.Foo object at 0x2cba790>
+    self = <failure_demo.Foo object at 0x101483450>
     
         def _get_b(self):
     >       raise Exception('Failed to get attrib')
     E       Exception: Failed to get attrib
     
-    failure_demo.py:97: Exception
+    failure_demo.py:113: Exception
     _________________________ test_attribute_multiple __________________________
     
         def test_attribute_multiple():
@@ -276,57 +332,57 @@ get on the terminal - we are working on that):
                 b = 2
     >       assert Foo().b == Bar().b
     E       assert 1 == 2
-    E        +  where 1 = <failure_demo.Foo object at 0x2cba210>.b
-    E        +    where <failure_demo.Foo object at 0x2cba210> = <class 'failure_demo.Foo'>()
-    E        +  and   2 = <failure_demo.Bar object at 0x2cba850>.b
-    E        +    where <failure_demo.Bar object at 0x2cba850> = <class 'failure_demo.Bar'>()
+    E        +  where 1 = <failure_demo.Foo object at 0x101483150>.b
+    E        +    where <failure_demo.Foo object at 0x101483150> = <class 'failure_demo.Foo'>()
+    E        +  and   2 = <failure_demo.Bar object at 0x101483350>.b
+    E        +    where <failure_demo.Bar object at 0x101483350> = <class 'failure_demo.Bar'>()
     
-    failure_demo.py:108: AssertionError
+    failure_demo.py:124: AssertionError
     __________________________ TestRaises.test_raises __________________________
     
-    self = <failure_demo.TestRaises instance at 0x2cc2560>
+    self = <failure_demo.TestRaises instance at 0x1014a6758>
     
         def test_raises(self):
             s = 'qwe'
     >       raises(TypeError, "int(s)")
     
-    failure_demo.py:117:
-    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+    failure_demo.py:133: 
+    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
     
     >   int(s)
     E   ValueError: invalid literal for int() with base 10: 'qwe'
     
-    <0-codegen /home/hpk/p/pytest/_pytest/python.py:819>:1: ValueError
+    <0-codegen /Users/hpk/p/pytest/_pytest/python.py:976>:1: ValueError
     ______________________ TestRaises.test_raises_doesnt _______________________
     
-    self = <failure_demo.TestRaises instance at 0x2cb6bd8>
+    self = <failure_demo.TestRaises instance at 0x1014b03f8>
     
         def test_raises_doesnt(self):
     >       raises(IOError, "int('3')")
     E       Failed: DID NOT RAISE
     
-    failure_demo.py:120: Failed
+    failure_demo.py:136: Failed
     __________________________ TestRaises.test_raise ___________________________
     
-    self = <failure_demo.TestRaises instance at 0x2cc4830>
+    self = <failure_demo.TestRaises instance at 0x1014a8998>
     
         def test_raise(self):
     >       raise ValueError("demo error")
     E       ValueError: demo error
     
-    failure_demo.py:123: ValueError
+    failure_demo.py:139: ValueError
     ________________________ TestRaises.test_tupleerror ________________________
     
-    self = <failure_demo.TestRaises instance at 0x2cc5560>
+    self = <failure_demo.TestRaises instance at 0x1014a27a0>
     
         def test_tupleerror(self):
     >       a,b = [1]
     E       ValueError: need more than 1 value to unpack
     
-    failure_demo.py:126: ValueError
+    failure_demo.py:142: ValueError
     ______ TestRaises.test_reinterpret_fails_with_print_for_the_fun_of_it ______
     
-    self = <failure_demo.TestRaises instance at 0x2cc6248>
+    self = <failure_demo.TestRaises instance at 0x1014a5518>
     
         def test_reinterpret_fails_with_print_for_the_fun_of_it(self):
             l = [1,2,3]
@@ -334,18 +390,18 @@ get on the terminal - we are working on that):
     >       a,b = l.pop()
     E       TypeError: 'int' object is not iterable
     
-    failure_demo.py:131: TypeError
+    failure_demo.py:147: TypeError
     ----------------------------- Captured stdout ------------------------------
     l is [1, 2, 3]
     ________________________ TestRaises.test_some_error ________________________
     
-    self = <failure_demo.TestRaises instance at 0x2cc6f38>
+    self = <failure_demo.TestRaises instance at 0x1014a1320>
     
         def test_some_error(self):
     >       if namenotexi:
     E   NameError: global name 'namenotexi' is not defined
     
-    failure_demo.py:134: NameError
+    failure_demo.py:150: NameError
     ____________________ test_dynamic_compile_shows_nicely _____________________
     
         def test_dynamic_compile_shows_nicely():
@@ -357,17 +413,17 @@ get on the terminal - we are working on that):
             py.std.sys.modules[name] = module
     >       module.foo()
     
-    failure_demo.py:149:
-    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+    failure_demo.py:165: 
+    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
     
         def foo():
     >    assert 1 == 0
     E    assert 1 == 0
     
-    <2-codegen 'abc-123' /home/hpk/p/pytest/doc/example/assertion/failure_demo.py:146>:2: AssertionError
+    <2-codegen 'abc-123' /Users/hpk/p/pytest/doc/example/assertion/failure_demo.py:162>:2: AssertionError
     ____________________ TestMoreErrors.test_complex_error _____________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2cc4050>
+    self = <failure_demo.TestMoreErrors instance at 0x1014a6638>
     
         def test_complex_error(self):
             def f():
@@ -376,16 +432,16 @@ get on the terminal - we are working on that):
                 return 43
     >       somefunc(f(), g())
     
-    failure_demo.py:159:
-    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+    failure_demo.py:175: 
+    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
     
     x = 44, y = 43
     
         def somefunc(x,y):
     >       otherfunc(x,y)
     
-    failure_demo.py:8:
-    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
+    failure_demo.py:8: 
+    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
     
     a = 44, b = 43
     
@@ -396,39 +452,39 @@ get on the terminal - we are working on that):
     failure_demo.py:5: AssertionError
     ___________________ TestMoreErrors.test_z1_unpack_error ____________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2cc7ab8>
+    self = <failure_demo.TestMoreErrors instance at 0x1014a42d8>
     
         def test_z1_unpack_error(self):
             l = []
     >       a,b  = l
     E       ValueError: need more than 0 values to unpack
     
-    failure_demo.py:163: ValueError
+    failure_demo.py:179: ValueError
     ____________________ TestMoreErrors.test_z2_type_error _____________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2ccb8c0>
+    self = <failure_demo.TestMoreErrors instance at 0x1014a0128>
     
         def test_z2_type_error(self):
             l = 3
     >       a,b  = l
     E       TypeError: 'int' object is not iterable
     
-    failure_demo.py:167: TypeError
+    failure_demo.py:183: TypeError
     ______________________ TestMoreErrors.test_startswith ______________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2ccd5f0>
+    self = <failure_demo.TestMoreErrors instance at 0x1014a0ef0>
     
         def test_startswith(self):
             s = "123"
             g = "456"
     >       assert s.startswith(g)
-    E       assert <built-in method startswith of str object at 0x2c321b0>('456')
-    E        +  where <built-in method startswith of str object at 0x2c321b0> = '123'.startswith
+    E       assert <built-in method startswith of str object at 0x1014951c0>('456')
+    E        +  where <built-in method startswith of str object at 0x1014951c0> = '123'.startswith
     
-    failure_demo.py:172: AssertionError
+    failure_demo.py:188: AssertionError
     __________________ TestMoreErrors.test_startswith_nested ___________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2ccbc20>
+    self = <failure_demo.TestMoreErrors instance at 0x1014a4170>
     
         def test_startswith_nested(self):
             def f():
@@ -436,47 +492,46 @@ get on the terminal - we are working on that):
             def g():
                 return "456"
     >       assert f().startswith(g())
-    E       assert <built-in method startswith of str object at 0x2c321b0>('456')
-    E        +  where <built-in method startswith of str object at 0x2c321b0> = '123'.startswith
-    E        +    where '123' = <function f at 0x2c2d140>()
-    E        +  and   '456' = <function g at 0x2cb00c8>()
+    E       assert <built-in method startswith of str object at 0x1014951c0>('456')
+    E        +  where <built-in method startswith of str object at 0x1014951c0> = '123'.startswith
+    E        +    where '123' = <function f at 0x1014aea28>()
+    E        +  and   '456' = <function g at 0x101477c80>()
     
-    failure_demo.py:179: AssertionError
+    failure_demo.py:195: AssertionError
     _____________________ TestMoreErrors.test_global_func ______________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2cb69e0>
+    self = <failure_demo.TestMoreErrors instance at 0x1014b3ab8>
     
         def test_global_func(self):
     >       assert isinstance(globf(42), float)
     E       assert isinstance(43, float)
     E        +  where 43 = globf(42)
     
-    failure_demo.py:182: AssertionError
+    failure_demo.py:198: AssertionError
     _______________________ TestMoreErrors.test_instance _______________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2cc6440>
+    self = <failure_demo.TestMoreErrors instance at 0x1014a2878>
     
         def test_instance(self):
             self.x = 6*7
     >       assert self.x != 42
     E       assert 42 != 42
-    E        +  where 42 = 42
-    E        +    where 42 = <failure_demo.TestMoreErrors instance at 0x2cc6440>.x
+    E        +  where 42 = <failure_demo.TestMoreErrors instance at 0x1014a2878>.x
     
-    failure_demo.py:186: AssertionError
+    failure_demo.py:202: AssertionError
     _______________________ TestMoreErrors.test_compare ________________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2dcc200>
+    self = <failure_demo.TestMoreErrors instance at 0x10149da70>
     
         def test_compare(self):
     >       assert globf(10) < 5
     E       assert 11 < 5
     E        +  where 11 = globf(10)
     
-    failure_demo.py:189: AssertionError
+    failure_demo.py:205: AssertionError
     _____________________ TestMoreErrors.test_try_finally ______________________
     
-    self = <failure_demo.TestMoreErrors instance at 0x2dce0e0>
+    self = <failure_demo.TestMoreErrors instance at 0x101493908>
     
         def test_try_finally(self):
             x = 1
@@ -484,5 +539,5 @@ get on the terminal - we are working on that):
     >           assert x == 0
     E           assert 1 == 0
     
-    failure_demo.py:194: AssertionError
-    ======================== 35 failed in 0.19 seconds =========================
+    failure_demo.py:210: AssertionError
+    ======================== 39 failed in 1.05 seconds =========================

doc/example/simple.txt

@@ -1,12 +1,14 @@
 
 .. highlightlang:: python
 
-basic patterns and examples
+Basic patterns and examples
 ==========================================================
 
-pass different values to a test function, depending on command line options
+Pass different values to a test function, depending on command line options
 ----------------------------------------------------------------------------
 
+.. regendoc:wipe
+
 Suppose we want to write a test that depends on a command line option.
 Here is a basic pattern how to achieve this::
 
@@ -32,7 +34,8 @@ provide the ``cmdopt`` through a :ref:`function argument <funcarg>` factory::
 
 Let's run this without supplying our new command line option::
 
-    $ py.test -q
+    $ py.test -q test_sample.py
+    collecting ... collected 1 items
     F
     ================================= FAILURES =================================
     _______________________________ test_answer ________________________________
@@ -50,11 +53,12 @@ Let's run this without supplying our new command line option::
     test_sample.py:6: AssertionError
     ----------------------------- Captured stdout ------------------------------
     first
-    1 failed in 0.02 seconds
+    1 failed in 0.50 seconds
 
 And now with supplying a command line option::
 
     $ py.test -q --cmdopt=type2
+    collecting ... collected 1 items
     F
     ================================= FAILURES =================================
     _______________________________ test_answer ________________________________
@@ -80,67 +84,46 @@ rather pass in different or more complex objects.  See the
 next example or refer to :ref:`mysetup` for more information
 on real-life examples.
 
-generating parameters combinations, depending on command line
-----------------------------------------------------------------------------
 
-Let's say we want to execute a test with different parameters
-and the parameter range shall be determined by a command
-line argument.  Let's first write a simple computation test::
+Dynamically adding command line options
+--------------------------------------------------------------
 
-    # content of test_compute.py
+.. regendoc:wipe
 
-    def test_compute(param1):
-        assert param1 < 4
-
-Now we add a test configuration like this::
+Through :confval:`addopts` you can statically add command line
+options for your project.  You can also dynamically modify
+the command line arguments before they get processed::
 
     # content of conftest.py
+    import sys
+    def pytest_cmdline_preparse(args):
+        if 'xdist' in sys.modules: # pytest-xdist plugin
+            import multiprocessing
+            num = max(multiprocessing.cpu_count() / 2, 1)
+            args[:] = ["-n", str(num)] + args
 
-    def pytest_addoption(parser):
-        parser.addoption("--all", action="store_true",
-            help="run all combinations")
+If you have the :ref:`xdist plugin <xdist>` installed
+you will now always perform test runs using a number
+of subprocesses close to your CPU. Running in an empty
+directory with the above conftest.py::
 
-    def pytest_generate_tests(metafunc):
-        if 'param1' in metafunc.funcargnames:
-            if metafunc.config.option.all:
-                end = 5
-            else:
-                end = 2
-            for i in range(end):
-                metafunc.addcall(funcargs={'param1': i})
-
-This means that we only run 2 tests if we do not pass ``--all``::
-
-    $ py.test -q test_compute.py
-    ..
-    2 passed in 0.01 seconds
-
-We run only two computations, so we see two dots.
-let's run the full monty::
-
-    $ py.test -q --all test_compute.py
-    ....F
-    ================================= FAILURES =================================
-    _____________________________ test_compute[4] ______________________________
+    $ py.test
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    gw0 I
+    gw0 [0]
     
-    param1 = 4
+    scheduling tests via LoadScheduling
     
-        def test_compute(param1):
-    >       assert param1 < 4
-    E       assert 4 < 4
-    
-    test_compute.py:3: AssertionError
-    1 failed, 4 passed in 0.03 seconds
+    =============================  in 5.12 seconds =============================
 
-As expected when running the full range of ``param1`` values
-we'll get an error on the last one.
+.. _`excontrolskip`:
 
-
-.. _`retrieved by hooks as item keywords`:
-
-control skipping of tests according to command line option
+Control skipping of tests according to command line option
 --------------------------------------------------------------
 
+.. regendoc:wipe
+
 Here is a ``conftest.py`` file adding a ``--runslow`` command
 line option to control skipping of ``slow`` marked tests::
 
@@ -171,32 +154,33 @@ We can now write a test module like this::
 
 and when running it will see a skipped "slow" test::
 
-    $ py.test test_module.py -rs    # "-rs" means report details on the little 's'
+    $ py.test -rs    # "-rs" means report details on the little 's'
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_module.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
     
     test_module.py .s
     ========================= short test summary info ==========================
-    SKIP [1] /tmp/doc-exec-104/conftest.py:9: need --runslow option to run
+    SKIP [1] /Users/hpk/tmp/doc-exec-158/conftest.py:9: need --runslow option to run
     
-    =================== 1 passed, 1 skipped in 0.02 seconds ====================
+    =================== 1 passed, 1 skipped in 0.09 seconds ====================
 
 Or run it including the ``slow`` marked test::
 
-    $ py.test test_module.py --runslow
+    $ py.test --runslow
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_module.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 2 items
     
     test_module.py ..
     
-    ========================= 2 passed in 0.01 seconds =========================
+    ========================= 2 passed in 0.02 seconds =========================
 
-
-writing well integrated assertion helpers
+Writing well integrated assertion helpers
 --------------------------------------------------
 
+.. regendoc:wipe
+
 If you have a test helper function called from a test you can
 use the ``pytest.fail`` marker to fail a test with a certain message.
 The test support function will not show up in the traceback if you
@@ -218,7 +202,8 @@ of tracebacks: the ``checkconfig`` function will not be shown
 unless the ``--fulltrace`` command line option is specified.
 Let's run our little function::
 
-    $ py.test -q
+    $ py.test -q test_checkconfig.py
+    collecting ... collected 1 items
     F
     ================================= FAILURES =================================
     ______________________________ test_something ______________________________
@@ -228,18 +213,19 @@ Let's run our little function::
     E       Failed: not configured: 42
     
     test_checkconfig.py:8: Failed
-    1 failed in 0.02 seconds
-
+    1 failed in 0.07 seconds
 
 Detect if running from within a py.test run
 --------------------------------------------------------------
 
+.. regendoc:wipe
+
 Usually it is a bad idea to make application code
 behave differently if called from a test.  But if you
 absolutely must find out if your application code is
 running from a test you can do something like this::
 
-    # content of conftest.py in your testing directory
+    # content of conftest.py
 
     def pytest_configure(config):
         import sys
@@ -254,8 +240,100 @@ and then check for the ``sys._called_from_test`` flag::
         # called from within a test run
     else:
         # called "normally"
-    
+
 accordingly in your application.  It's also a good idea
-to rather use your own application module rather than ``sys``
+to use your own application module rather than ``sys``
 for handling flag.
 
+Adding info to test report header
+--------------------------------------------------------------
+
+.. regendoc:wipe
+
+It's easy to present extra information in a py.test run::
+
+    # content of conftest.py
+    
+    def pytest_report_header(config):
+        return "project deps: mylib-1.1"
+
+which will add the string to the test header accordingly::
+
+    $ py.test
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    project deps: mylib-1.1
+    collecting ... collected 0 items
+    
+    =============================  in 0.01 seconds =============================
+
+.. regendoc:wipe
+
+You can also return a list of strings which will be considered as several
+lines of information.  You can of course also make the amount of reporting
+information on e.g. the value of ``config.option.verbose`` so that
+you present more information appropriately::
+
+    # content of conftest.py
+
+    def pytest_report_header(config):
+        if config.option.verbose > 0:
+            return ["info1: did you know that ...", "did you?"]
+
+which will add info only when run with "--v"::
+
+    $ py.test -v
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2 -- /Users/hpk/venv/0/bin/python
+    info1: did you know that ...
+    did you?
+    collecting ... collected 0 items
+    
+    =============================  in 0.03 seconds =============================
+
+and nothing when run plainly::
+
+    $ py.test
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 0 items
+    
+    =============================  in 0.01 seconds =============================
+
+profiling test duration
+--------------------------
+
+.. regendoc:wipe
+
+.. versionadded: 2.2
+
+If you have a slow running large test suite you might want to find
+out which tests are the slowest. Let's make an artifical test suite::
+
+    # content of test_some_are_slow.py
+
+    import time
+
+    def test_funcfast():
+        pass
+
+    def test_funcslow1():
+        time.sleep(0.1)
+
+    def test_funcslow2():
+        time.sleep(0.2)
+
+Now we can profile which test functions execute the slowest::
+
+    $ py.test --durations=3
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 3 items
+    
+    test_some_are_slow.py ...
+    
+    ========================= slowest 3 test durations =========================
+    0.20s call     test_some_are_slow.py::test_funcslow2
+    0.10s call     test_some_are_slow.py::test_funcslow1
+    0.00s call     test_some_are_slow.py::test_funcfast
+    ========================= 3 passed in 0.33 seconds =========================

doc/example/xfail_demo.py

@@ -17,5 +17,9 @@ def test_hello3():
 def test_hello4():
     assert 0
 
+@xfail('pytest.__version__[0] != "17"')
 def test_hello5():
+    assert 0
+
+def test_hello6():
     pytest.xfail("reason")

doc/extracol

@@ -1,32 +0,0 @@
-changing Python test discovery patterns
---------------------------------------------------
-
-You can influence python test file, function and class prefixes through
-the :confval:`python_patterns` configuration valueto determine which
-files are checked and which test functions are found.  Example for using
-a scheme that builds on ``check`` rather than on ``test`` prefixes::
-
-
-    # content of setup.cfg
-    [pytest]
-    python_patterns =
-        files: check_*.py
-        functions: check_
-        classes: Check
-
-See 
-    :confval:`python_funcprefixes` and :confval:`python_classprefixes`
-
-
-    changing test file discovery
-    -----------------------------------------------------
-
-    You can specify patterns where python tests are found::
-     
-        python_testfilepatterns =
-            testing/**/{purebasename}.py
-            testing/*.py
-
-    .. note::
-
-       conftest.py files are never considered for test discovery

doc/faq.txt

@@ -12,25 +12,26 @@ On naming, nosetests, licensing and magic
 Why a ``py.test`` instead of a ``pytest`` command?
 ++++++++++++++++++++++++++++++++++++++++++++++++++
 
-Some historic, some practical reasons: ``py.test`` used to be part of
-the ``py`` package which provided several developer utitilities,
-all starting with ``py.<TAB>``, providing nice TAB-completion. If
+Some of the reasons are historic, others are practical.  ``py.test``
+used to be part of the ``py`` package which provided several developer
+utilities, all starting with ``py.<TAB>``, thus providing nice
+TAB-completion. If
 you install ``pip install pycmd`` you get these tools from a separate
 package.  These days the command line tool could be called ``pytest``
-but then again many people have gotten used to the old name and there
-is another tool named "pytest" so we just decided to stick with
+but since many people have gotten used to the old name and there
+is another tool named "pytest" we just decided to stick with
 ``py.test``.
 
-What's the relation to nose and unittest?
+How does py.test relate to nose and unittest?
 +++++++++++++++++++++++++++++++++++++++++++++++++
 
 py.test and nose_ share basic philosophy when it comes
-to running Python tests.  In fact, you can run many tests
-written nose with py.test.  nose_ was originally created
+to running and writing Python tests.  In fact, you can run many tests
+written for nose with py.test.  nose_ was originally created
 as a clone of ``py.test`` when py.test was in the ``0.8`` release
-cycle.  As of version 2.0 support for running unittest test
-suites is majorly improved and you should be able to run
-many Django and Twisted test suites.
+cycle.  Note that starting with pytest-2.0 support for running unittest
+test suites is majorly improved and you should be able to run
+many Django and Twisted test suites without modification.
 
 .. _features: test/features.html
 
@@ -39,28 +40,30 @@ What's this "magic" with py.test?
 ++++++++++++++++++++++++++++++++++++++++++
 
 Around 2007 (version ``0.8``) some people claimed that py.test
-was using too much "magic".  It has been refactored a lot.  Thrown
-out old code.  Deprecated unused approaches and code.  And it is today
-probably one of the smallest, most universally runnable and most
-customizable testing frameworks for Python.   It's true that
-``py.test`` uses metaprogramming techniques, i.e. it views
-test code similar to how compilers view programs, using a
-somewhat abstract internal model.
+was using too much "magic".  Partly this has been fixed by removing
+unused, deprecated or complicated code.  It is today probably one
+of the smallest, most universally runnable and most
+customizable testing frameworks for Python.   However,
+``py.test`` still uses many metaprogramming techniques and
+reading its source is thus likely not something for Python beginners.
 
-It's also true that the no-boilerplate testing is implemented by making
-use of the Python assert statement through "re-interpretation":
-When an ``assert`` statement fails, py.test re-interprets the expression
-to show intermediate values if a test fails.  If your expression
-has side effects the intermediate values may not be the same, obfuscating
-the initial error (this is also explained at the command line if it happens).
-``py.test --no-assert`` turns off assert re-intepretation.
-Sidenote: it is good practise to avoid asserts with side effects.
+A second "magic" issue is arguably the assert statement debugging feature. When
+loading test modules py.test rewrites the source code of assert statements. When
+a rewritten assert statement fails, its error message has more information than
+the original. py.test also has a second assert debugging technique. When an
+``assert`` statement that was missed by the rewriter fails, py.test
+re-interprets the expression to show intermediate values if a test fails. This
+second technique suffers from a caveat that the rewriting does not: If your
+expression has side effects (better to avoid them anyway!) the intermediate
+values may not be the same, confusing the reinterpreter and obfuscating the
+initial error (this is also explained at the command line if it happens).
+You can turn off all assertion debugging with ``py.test --assertmode=off``.
 
 .. _`py namespaces`: index.html
 .. _`py/__init__.py`: http://bitbucket.org/hpk42/py-trunk/src/trunk/py/__init__.py
 
 
-function arguments, parametrized tests and setup
+Function arguments, parametrized tests and setup
 -------------------------------------------------------
 
 .. _funcargs: test/funcargs.html
@@ -69,7 +72,7 @@ Is using funcarg- versus xUnit setup a style question?
 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 
 For simple applications and for people experienced with nose_ or
-unittest-style test setup using `xUnit style setup`_ often
+unittest-style test setup using `xUnit style setup`_ probably
 feels natural.  For larger test suites, parametrized testing
 or setup of complex test resources using funcargs_ may feel more natural.
 Moreover, funcargs are ideal for writing advanced test support
@@ -86,13 +89,11 @@ in a managed class/module/function scope.
 Why the ``pytest_funcarg__*`` name for funcarg factories?
 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 
-We alternatively implemented an explicit registration mechanism for
-function argument factories.  But lacking a good use case for this
-indirection and flexibility we decided to go for `Convention over
-Configuration`_ and rather have factories specified by convention.
-Besides removing the need for an registration indirection it allows to
-"grep" for ``pytest_funcarg__MYARG`` and will safely find all factory
-functions for the ``MYARG`` function argument.
+We like `Convention over Configuration`_ and didn't see much point
+in allowing a more flexible or abstract mechanism.  Moreover,
+it is nice to be able to search for ``pytest_funcarg__MYARG`` in
+source code and safely find all factory functions for
+the ``MYARG`` function argument.
 
 .. _`Convention over Configuration`: http://en.wikipedia.org/wiki/Convention_over_Configuration
 
@@ -125,8 +126,8 @@ Issues with py.test, multiprocess and setuptools?
 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 
 On windows the multiprocess package will instantiate sub processes
-by pickling and thus implicitely re-import a lot of local modules.
-Unfortuantely, setuptools-0.6.11 does not ``if __name__=='__main__'``
+by pickling and thus implicitly re-import a lot of local modules.
+Unfortunately, setuptools-0.6.11 does not ``if __name__=='__main__'``
 protect its generated command line script.  This leads to infinite
 recursion when running a test that instantiates Processes.
 

doc/funcargs.txt

@@ -1,5 +1,5 @@
 ==============================================================
-creating and managing test function arguments
+Injecting objects into test functions (funcargs)
 ==============================================================
 
 .. currentmodule:: _pytest.python
@@ -11,16 +11,26 @@ creating and managing test function arguments
 Dependency injection through function arguments
 =================================================
 
-py.test allows to inject values into test functions through the *funcarg
-mechanism*: For each argument name in a test function signature a factory is
-looked up and called to create the value.  The factory can live in the
-same test class, test module, in a per-directory ``confest.py`` file or
-in an external plugin.  It has full access to the requesting test
-function, can register finalizers and invoke lifecycle-caching
-helpers.  As can be expected from a systematic dependency
-injection mechanism, this allows full de-coupling of resource and
-fixture setup from test code, enabling more maintainable and
-easy-to-modify test suites.
+py.test lets you inject objects into test functions and precisely
+control their life cycle in relation to the test execution. It is
+also possible to run a test function multiple times with different objects.
+
+The basic mechanism for injecting objects is also called the
+*funcarg mechanism* because objects are ultimately injected
+by calling a test function with it as an argument.  Unlike the
+classical xUnit approach *funcargs* relate more to `Dependency Injection`_
+because they help to de-couple test code from objects required for
+them to execute.
+
+.. _`Dependency injection`: http://en.wikipedia.org/wiki/Dependency_injection
+
+To create a value with which to call a test function a factory function
+is called which gets full access to the test function context and can
+register finalizers or invoke lifecycle-caching helpers.  The factory
+can be implemented in same test class or test module, or in a
+per-directory ``conftest.py`` file or even in an external plugin.  This
+allows full de-coupling of test code and objects needed for test
+execution.
 
 A test function may be invoked multiple times in which case we
 speak of :ref:`parametrized testing <parametrizing-tests>`. This can be
@@ -28,11 +38,14 @@ very useful if you want to test e.g. against different database backends
 or with multiple numerical arguments sets and want to reuse the same set
 of test functions.
 
-Basic funcarg example
------------------------
+py.test comes with :ref:`builtinfuncargs` and there are some refined usages in the examples section.
 
-Let's look at a simple self-contained example that you can put
-into a test module::
+.. _funcarg:
+
+Basic injection example
+--------------------------------
+
+Let's look at a simple self-contained test module::
 
     # content of ./test_simplefactory.py
     def pytest_funcarg__myfuncarg(request):
@@ -41,12 +54,16 @@ into a test module::
     def test_function(myfuncarg):
         assert myfuncarg == 17
 
+This test function needs an injected object named ``myfuncarg``.
+py.test will discover and call the factory named
+``pytest_funcarg__myfuncarg`` within the same module in this case.
+
 Running the test looks like this::
 
     $ py.test test_simplefactory.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_simplefactory.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
     
     test_simplefactory.py F
     
@@ -60,10 +77,10 @@ Running the test looks like this::
     E       assert 42 == 17
     
     test_simplefactory.py:5: AssertionError
-    ========================= 1 failed in 0.02 seconds =========================
+    ========================= 1 failed in 0.03 seconds =========================
 
-This means that the test function was called with a ``myfuncarg`` value
-of ``42`` and the assert fails accordingly.  Here is how py.test
+This means that indeed the test function was called with a ``myfuncarg``
+argument value of ``42`` and the assert fails.  Here is how py.test
 comes to call the test function this way:
 
 1. py.test :ref:`finds <test discovery>` the ``test_function`` because
@@ -74,14 +91,15 @@ comes to call the test function this way:
 2. ``pytest_funcarg__myfuncarg(request)`` is called and
    returns the value for ``myfuncarg``.
 
-3. the test function can now be called: ``test_function(42)``
-   and results in the above exception because of the assertion
+3. the test function can now be called: ``test_function(42)``.
+   This results in the above exception because of the assertion
    mismatch.
 
 Note that if you misspell a function argument or want
 to use one that isn't available, you'll see an error
-with a list of available function arguments. You can
-also issue::
+with a list of available function arguments.
+
+You can always issue::
 
     py.test --funcargs test_simplefactory.py
 
@@ -101,9 +119,9 @@ think of as "resources").
 The funcarg **request** object
 =============================================
 
-Each funcarg factory receives a **request** object which is tied to a
-specific test function call.  A request object is passed to a funcarg
-factory and provides access to test configuration and context:
+Each funcarg factory receives a **request** object tied to a specific test
+function call.  A request object is passed to a funcarg factory and provides
+access to test configuration and context:
 
 .. autoclass:: _pytest.python.FuncargRequest()
     :members: function,cls,module,keywords,config
@@ -140,18 +158,17 @@ hook to generate several calls to the same test function::
     # content of test_example.py
     def pytest_generate_tests(metafunc):
         if "numiter" in metafunc.funcargnames:
-            for i in range(10):
-                metafunc.addcall(funcargs=dict(numiter=i))
+            metafunc.parametrize("numiter", range(10))
 
     def test_func(numiter):
         assert numiter < 9
 
-Running this::
+Running this will generate ten invocations of ``test_func`` passing in each of the items in the list of ``range(10)``::
 
     $ py.test test_example.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_example.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 10 items
     
     test_example.py .........F
     
@@ -164,14 +181,17 @@ Running this::
     >       assert numiter < 9
     E       assert 9 < 9
     
-    test_example.py:7: AssertionError
-    ==================== 1 failed, 9 passed in 0.03 seconds ====================
+    test_example.py:6: AssertionError
+    ==================== 1 failed, 9 passed in 0.07 seconds ====================
 
-Note that the ``pytest_generate_tests(metafunc)`` hook is called during
+Obviously, only when ``numiter`` has the value of ``9`` does the test fail.  Note that the ``pytest_generate_tests(metafunc)`` hook is called during
 the test collection phase which is separate from the actual test running.
 Let's just look at what is collected::
 
     $ py.test --collectonly test_example.py
+    =========================== test session starts ============================
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 10 items
     <Module 'test_example.py'>
       <Function 'test_func[0]'>
       <Function 'test_func[1]'>
@@ -183,19 +203,23 @@ Let's just look at what is collected::
       <Function 'test_func[7]'>
       <Function 'test_func[8]'>
       <Function 'test_func[9]'>
+    
+    =============================  in 0.01 seconds =============================
 
 If you want to select only the run with the value ``7`` you could do::
 
     $ py.test -v -k 7 test_example.py  # or -k test_func[7]
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30 -- /home/hpk/venv/0/bin/python
-    test path 1: test_example.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2 -- /Users/hpk/venv/0/bin/python
+    collecting ... collected 10 items
     
-    test_example.py:6: test_func[7] PASSED
+    test_example.py:5: test_func[7] PASSED
     
-    ======================== 9 tests deselected by '7' =========================
+    ======================= 9 tests deselected by '-k7' ========================
     ================== 1 passed, 9 deselected in 0.01 seconds ==================
 
+You might want to look at :ref:`more parametrization examples <paramexamples>`.
+
 .. _`metafunc object`:
 
 The **metafunc** object
@@ -216,4 +240,5 @@ in the class or module where a test function is defined:
 
 ``metafunc.config``: access to command line opts and general config
 
-.. automethod:: Metafunc.addcall(funcargs=None, id=_notexists, param=_notexists)
+.. automethod:: Metafunc.parametrize
+.. automethod:: Metafunc.addcall(funcargs=None,id=_notexists,param=_notexists)

doc/getting-started.txt

@@ -1,7 +1,13 @@
 Installation and Getting Started
 ===================================
 
-**Compatibility**: Python 2.4-3.2, Jython, PyPy on Unix/Posix and Windows
+**Pythons**: Python 2.4-3.2, Jython, PyPy
+
+**Platforms**: Unix/Posix and Windows
+
+**PyPI package name**: `pytest <http://pypi.python.org/pypi/pytest>`_
+
+**documentation as PDF**: `download latest <http://pytest.org/latest/pytest.pdf>`_
 
 .. _`getstarted`:
 
@@ -10,13 +16,16 @@ Installation
 
 Installation options::
 
-    easy_install -U pytest  # or
-    pip install -U pytest
+    pip install -U pytest # or
+    easy_install -U pytest
 
 To check your installation has installed the correct version::
 
     $ py.test --version
-    This is py.test version 2.0.0.dev30, imported from /home/hpk/p/pytest/pytest.py
+    This is py.test version 2.2.2, imported from /Users/hpk/p/pytest/pytest.pyc
+    setuptools registered plugins:
+      pytest-xdist-1.8 at /Users/hpk/p/pytest-xdist/xdist/plugin.pyc
+      pytest-cov-1.4 at /Users/hpk/venv/0/lib/python2.7/site-packages/pytest_cov.pyc
 
 If you get an error checkout :ref:`installation issues`.
 
@@ -38,48 +47,39 @@ That's it. You can execute the test function now::
 
     $ py.test
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: /tmp/doc-exec-70
-
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
+    
     test_sample.py F
-
+    
     ================================= FAILURES =================================
     _______________________________ test_answer ________________________________
-
+    
         def test_answer():
     >       assert func(3) == 5
     E       assert 4 == 5
     E        +  where 4 = func(3)
-
+    
     test_sample.py:5: AssertionError
     ========================= 1 failed in 0.02 seconds =========================
 
-py.test found the ``test_answer`` function by following :ref:`standard test discovery rules <test discovery>`, basically detecting the ``test_`` prefixes.  We got a failure report because our little ``func(3)`` call did not return ``5``.  The report is formatted using the :ref:`standard traceback reporting`.
+py.test found the ``test_answer`` function by following :ref:`standard test discovery rules <test discovery>`, basically detecting the ``test_`` prefixes.  We got a failure report because our little ``func(3)`` call did not return ``5``.
 
 .. note::
 
-    You can simply use the ``assert`` statement for coding expectations because
-    intermediate values will be presented to you. This is arguably easier than
-    learning all the `the JUnit legacy methods`_.
+    You can simply use the ``assert`` statement for asserting test
+    expectations.  pytest's :ref:`assert introspection` will intelligently
+    report intermediate values of the assert expression freeing
+    you from the need to learn the many names of `JUnit legacy methods`_.
 
-    However, there remains one caveat to using simple asserts: your
-    assertion expression should better be side-effect free.  Because
-    after an assertion failed py.test will re-evaluate the expression
-    in order to present intermediate values. You will get a nice warning
-    and you can easily fix it: compute the value ahead of the assert and
-    then do the assertion.  Or maybe just use the assert "explicit message"
-    syntax::
-
-       assert expr, "message" # show "message" if expr is not True
-
-.. _`the JUnit legacy methods`: http://docs.python.org/library/unittest.html#test-cases
+.. _`JUnit legacy methods`: http://docs.python.org/library/unittest.html#test-cases
 
 .. _`assert statement`: http://docs.python.org/reference/simple_stmts.html#the-assert-statement
 
-Asserting a certain exception is raised
+Asserting that a certain exception is raised
 --------------------------------------------------------------
 
-If you want to assert some code raises an exception you can
+If you want to assert that some code raises an exception you can
 use the ``raises`` helper::
 
     # content of test_sysexit.py
@@ -94,17 +94,18 @@ use the ``raises`` helper::
 Running it with, this time in "quiet" reporting mode::
 
     $ py.test -q test_sysexit.py
+    collecting ... collected 1 items
     .
-    1 passed in 0.00 seconds
+    1 passed in 0.01 seconds
 
 .. todo:: For further ways to assert exceptions see the `raises`
 
 Grouping multiple tests in a class
 --------------------------------------------------------------
 
-If you start to have more than a few tests it often makes sense
-to group tests logically, in classes and modules.  Let's put two
-tests in a class like this::
+Once you start to have more than a few tests it often makes sense
+to group tests logically, in classes and modules.  Let's write a class
+containing two tests::
 
     # content of test_class.py
     class TestClass:
@@ -121,19 +122,20 @@ There is no need to subclass anything.  We can simply
 run the module by passing its filename::
 
     $ py.test -q test_class.py
+    collecting ... collected 2 items
     .F
     ================================= FAILURES =================================
     ____________________________ TestClass.test_two ____________________________
-
-    self = <test_class.TestClass instance at 0x288fc20>
-
+    
+    self = <test_class.TestClass instance at 0x1013225a8>
+    
         def test_two(self):
             x = "hello"
     >       assert hasattr(x, 'check')
     E       assert hasattr('hello', 'check')
-
+    
     test_class.py:8: AssertionError
-    1 failed, 1 passed in 0.02 seconds
+    1 failed, 1 passed in 0.03 seconds
 
 The first test passed, the second failed. Again we can easily see
 the intermediate values used in the assertion, helping us to
@@ -157,21 +159,22 @@ py.test will lookup and call a factory to create the resource
 before performing the test function call.  Let's just run it::
 
     $ py.test -q test_tmpdir.py
+    collecting ... collected 1 items
     F
     ================================= FAILURES =================================
     _____________________________ test_needsfiles ______________________________
-
-    tmpdir = local('/tmp/pytest-122/test_needsfiles0')
-
+    
+    tmpdir = local('/Users/hpk/tmp/pytest-20/test_needsfiles0')
+    
         def test_needsfiles(tmpdir):
             print tmpdir
     >       assert 0
     E       assert 0
-
+    
     test_tmpdir.py:3: AssertionError
     ----------------------------- Captured stdout ------------------------------
-    /tmp/pytest-122/test_needsfiles0
-    1 failed in 0.05 seconds
+    /Users/hpk/tmp/pytest-20/test_needsfiles0
+    1 failed in 0.11 seconds
 
 Before the test runs, a unique-per-test-invocation temporary directory
 was created.  More info at :ref:`tmpdir handling`.
@@ -180,13 +183,13 @@ You can find out what kind of builtin :ref:`funcargs` exist by typing::
 
     py.test --funcargs   # shows builtin and custom function arguments
 
-where to go next
+Where to go next
 -------------------------------------
 
 Here are a few suggestions where to go next:
 
 * :ref:`cmdline` for command line invocation examples
-* :ref:`good practises` for virtualenv, test layout, genscript support
+* :ref:`good practises <goodpractises>` for virtualenv, test layout, genscript support
 * :ref:`apiref` for documentation and examples on using py.test
 * :ref:`plugins` managing and writing plugins
 
@@ -198,10 +201,15 @@ Known Installation issues
 easy_install or pip not found?
 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 
-Consult `distribute docs`_ to install the ``easy_install``
-tool on your machine.  You may also use the older
-`setuptools`_ project but it lacks bug fixes and does not
-work on Python3.  If you use Python2 you may also install pip_.
+.. _`install pip`: http://www.pip-installer.org/en/latest/index.html
+
+`Install pip`_ for a state of the art python package installer.
+
+Or consult `distribute docs`_ to install the ``easy_install``
+tool on your machine.  
+
+You may also use the older `setuptools`_ project but it lacks bug fixes
+and does not work on Python3.
 
 py.test not found on Windows despite installation?
 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
@@ -220,7 +228,7 @@ py.test not found on Windows despite installation?
 - **Jython2.5.1 on Windows XP**: `Jython does not create command line launchers`_
   so ``py.test`` will not work correctly.  You may install py.test on
   CPython and type ``py.test --genscript=mytest`` and then use
-  ``jython mytest`` to run py.test for your tests to run in Jython.
+  ``jython mytest`` to run py.test for your tests to run with Jython.
 
  :ref:`examples` for more complex examples
 

doc/goodpractises.txt

@@ -8,31 +8,29 @@ Good Integration Practises
 Work with virtual environments
 -----------------------------------------------------------
 
-We recommend to work with virtualenv_ environments and use easy_install_
+We recommend to use virtualenv_ environments and use easy_install_
 (or pip_) for installing your application dependencies as well as
-the ``pytest`` package itself.  This way you get a much more reproducible
+the ``pytest`` package itself.  This way you will get a much more reproducible
 environment.  A good tool to help you automate test runs against multiple
-dependency configurations or Python interpreters is `tox`_,
-independently created by the main py.test author.  The latter
-is also useful for integration with the continous integration
-server Hudson_.
+dependency configurations or Python interpreters is `tox`_.
 
 .. _`virtualenv`: http://pypi.python.org/pypi/virtualenv
 .. _`buildout`: http://www.buildout.org/
 .. _pip: http://pypi.python.org/pypi/pip
 
-Use tox and Continous Integration servers
+Use tox and Continuous Integration servers
 -------------------------------------------------
 
-If you are (often) releasing code to the public you
+If you frequently release code to the public you
 may want to look into `tox`_, the virtualenv test automation
 tool and its `pytest support <http://codespeak.net/tox/example/pytest.html>`_.
-The basic idea is to generate a JUnitXML file through the ``--junitxml=PATH`` option and have a continous integration server like Hudson_ pick it up.
+The basic idea is to generate a JUnitXML file through the ``--junitxml=PATH`` option and have a continuous integration server like Jenkins_ pick it up
+and generate reports.
 
 .. _standalone:
 .. _`genscript method`:
 
-Create a py.test standalone Script
+Create a py.test standalone script
 -------------------------------------------
 
 If you are a maintainer or application developer and want others
@@ -55,7 +53,7 @@ You can tell people to download the script and then e.g.  run it like this::
 Integrating with distutils / ``python setup.py test``
 --------------------------------------------------------
 
-You can easily integrate test runs into your distutils or
+You can integrate test runs into your distutils or
 setuptools based project.  Use the `genscript method`_
 to generate a standalone py.test script::
 
@@ -90,7 +88,7 @@ If you now type::
 this will execute your tests using ``runtest.py``. As this is a
 standalone version of ``py.test`` no prior installation whatsoever is
 required for calling the test command. You can also pass additional
-arguments to the subprocess-calls like your test directory or other
+arguments to the subprocess-calls such as your test directory or other
 options.
 
 .. _`test discovery`:
@@ -101,16 +99,16 @@ Conventions for Python test discovery
 
 ``py.test`` implements the following standard test discovery:
 
-* collection starts from initial command line arguments
+* collection starts from the initial command line arguments
   which may be directories, filenames or test ids.
 * recurse into directories, unless they match :confval:`norecursedirs`
 * ``test_*.py`` or ``*_test.py`` files, imported by their `package name`_.
 * ``Test`` prefixed test classes (without an ``__init__`` method)
 * ``test_`` prefixed test functions or methods are test items
 
-For changing and customization example, see :doc:`example/pythoncollection`.
+For examples of how to customize your test discovery :doc:`example/pythoncollection`.
 
-py.test additionally discovers tests using the standard
+Within Python modules, py.test also discovers tests using the standard
 :ref:`unittest.TestCase <unittest.TestCase>` subclassing technique.
 
 Choosing a test layout / import rules
@@ -140,7 +138,10 @@ py.test supports common test layouts:
         test_app.py
         ...
 
-You can always run your tests by pointing to it::
+In both cases you usually need to make sure that ``mypkg`` is importable,
+for example by using the setuptools ``python setup.py develop`` method.
+
+You can run your tests by pointing to it::
 
     py.test tests/test_app.py       # for external test dirs
     py.test mypkg/test/test_app.py  # for inlined test dirs
@@ -152,18 +153,27 @@ You can always run your tests by pointing to it::
 
 .. note::
 
-    Test modules are imported under their fully qualified name as follows:
+    If py.test finds a "a/b/test_module.py" test file while
+    recursing into the filesystem it determines the import name
+    as follows:
 
-    * find ``basedir`` -- this is the first "upward" directory not
-      containing an ``__init__.py``
+    * find ``basedir`` -- this is the first "upward" (towards the root)
+      directory not containing an ``__init__.py``.  If both the ``a``
+      and ``b`` directories contain an ``__init__.py`` the basedir will
+      be the parent dir of ``a``.
 
-    * perform ``sys.path.insert(0, basedir)`` to make the fully
-      qualified test module path importable.
+    * perform ``sys.path.insert(0, basedir)`` to make the test module
+      importable under the fully qualified import name.
 
-    * ``import path.to.test_module`` where the path is determined
-      by converting path separators into "." files.  This means
+    * ``import a.b.test_module`` where the path is determined
+      by converting path separators ``/`` into "." characters.  This means
       you must follow the convention of having directory and file
-      names map to the import names.
+      names map directly to the import names.
+
+    The reason for this somewhat evolved importing technique is
+    that in larger projects multiple test modules might import
+    from each other and thus deriving a canonical import name helps
+    to avoid surprises such as a test modules getting imported twice.
 
 
 .. include:: links.inc

doc/index.txt

@@ -1,36 +1,39 @@
 
 
-Welcome to ``py.test``!
+Welcome to pytest!
 =============================================
 
+- **a mature full-featured testing tool**
 
-- **a mature fully featured testing tool**
-
- - runs on Posix/Windows, Python 2.4-3.2, PyPy and Jython
- - continously `tested on many Python interpreters <http://hudson.testrun.org/view/pytest/job/pytest/>`_
- - used in :ref:`many projects <projects>`, ranging from 10 to 10000 tests
- - has :ref:`comprehensive documentation <toc>`
- - comes with :ref:`tested examples <examples>`
+ - runs on Posix/Windows, Python 2.4-3.2, PyPy and Jython-2.5.1
+ - :ref:`comprehensive online <toc>` and `PDF documentation <pytest.pdf>`_
+ - continuously `tested on many Python interpreters <http://hudson.testrun.org/view/pytest/job/pytest/>`_
+ - used in :ref:`many projects and organisations <projects>`, in test
+   suites ranging from 10 to 10s of thousands of tests
+ - comes with many :ref:`tested examples <examples>`
  - supports :ref:`good integration practises <goodpractises>`
 
 - **provides no-boilerplate testing**
 
- - makes it :ref:`easy to get started <getstarted>`, refined :ref:`usage options <usage>`
+ - makes it :ref:`easy to get started <getstarted>`,
+ - refined :ref:`usage options <usage>`
  - :ref:`assert with the assert statement`
  - helpful :ref:`traceback and failing assertion reporting <tbreportdemo>`
- - allows `print debugging <printdebugging>`_ and `generic output capturing <captures>`_
- - supports :pep:`8` compliant coding style in tests
+ - allows :ref:`print debugging <printdebugging>` and :ref:`the
+   capturing of standard output during test execution <captures>`
+ - supports :pep:`8` compliant coding styles in tests
 
 - **supports functional testing and complex test setups**
 
+ - (new in 2.2) :ref:`durations`
+ - (much improved in 2.2) :ref:`marking and test selection <mark>`
+ - (improved in 2.2) :ref:`parametrized test functions <parametrized test functions>`
  - advanced :ref:`skip and xfail`
- - generic :ref:`marking and test selection <mark>`
+ - unique :ref:`dependency injection through funcargs <funcargs>`
  - can :ref:`distribute tests to multiple CPUs <xdistcpu>` through :ref:`xdist plugin <xdist>`
- - can :ref:`continously re-run failing tests <looponfailing>`
+ - can :ref:`continuously re-run failing tests <looponfailing>`
  - many :ref:`builtin helpers <pytest helpers>`
  - flexible :ref:`Python test discovery`
- - unique :ref:`dependency injection through funcargs <funcargs>`
- - :ref:`parametrized test functions <parametrized test functions>`
 
 - **integrates many common testing methods**
 
@@ -38,8 +41,8 @@ Welcome to ``py.test``!
    tests, including running testcases made for Django and trial
  - supports extended :ref:`xUnit style setup <xunitsetup>`
  - supports domain-specific :ref:`non-python tests`
- - supports generating testing coverage reports
- - `Javasript unit- and functional testing`_
+ - supports the generation of testing coverage reports
+ - `Javascript unit- and functional testing`_
 
 - **extensive plugin and customization system**
 
@@ -47,13 +50,8 @@ Welcome to ``py.test``!
  - customizations can be per-directory, per-project or per PyPI released plugins
  - it is easy to add command line options or do other kind of add-ons and customizations.
 
-.. _`Javasript unit- and functional testing`: http://pypi.python.org/pypi/oejskit
+.. _`Javascript unit- and functional testing`: http://pypi.python.org/pypi/oejskit
 
 .. _`easy`: http://bruynooghe.blogspot.com/2009/12/skipping-slow-test-by-default-in-pytest.html
 
-.. toctree::
-   :hidden:
-
-   contact.txt
-   contents.txt
 

doc/links.inc

@@ -10,11 +10,11 @@
 .. _pytest: http://pypi.python.org/pypi/pytest
 .. _mercurial: http://mercurial.selenic.com/wiki/
 .. _`setuptools`: http://pypi.python.org/pypi/setuptools
-
 .. _`easy_install`:
 .. _`distribute docs`:
 .. _`distribute`: http://pypi.python.org/pypi/distribute
 .. _`pip`: http://pypi.python.org/pypi/pip
 .. _`virtualenv`: http://pypi.python.org/pypi/virtualenv
 .. _hudson: http://hudson-ci.org/
+.. _jenkins: http://jenkins-ci.org/
 .. _tox: http://codespeak.net/tox

doc/mark.txt

@@ -1,122 +1,25 @@
 
 .. _mark:
 
-mark test functions with attributes
+Marking test functions with attributes
 =================================================================
 
 .. currentmodule:: _pytest.mark
 
-By using the ``pytest.mark`` helper you can instantiate
-decorators that will set named meta data on test functions.
+By using the ``pytest.mark`` helper you can easily set
+metadata on your test functions. There are
+some builtin markers, for example:
 
-Marking a single function
-----------------------------------------------------
+* :ref:`skipif <skipif>` - skip a test function if a certain condition is met
+* :ref:`xfail <xfail>` - produce an "expected failure" outcome if a certain
+  condition is met
+* :ref:`parametrize <parametrizemark>` to perform multiple calls
+  to the same test function.
 
-You can "mark" a test function with meta data like this::
+It's easy to create custom markers or to apply markers
+to whole test classes or modules. See :ref:`mark examples` for examples
+which also serve as documentation.
 
-    import pytest
-    @pytest.mark.webtest
-    def test_send_http():
-        ...
-
-This will set the function attribute ``webtest`` to a :py:class:`MarkInfo`
-instance.  You can also specify parametrized meta data like this::
-
-    # content of test_mark.py
-
-    import pytest
-    @pytest.mark.webtest(firefox=30)
-    def test_receive():
-        pass
-
-    @pytest.mark.webtest("functional", firefox=30)
-    def test_run_and_look():
-        pass
-
-and access it from other places like this::
-
-    test_receive.webtest.kwargs['firefox'] == 30
-    test_run_and_look.webtest.args[0] == "functional"
-
-.. _`scoped-marking`:
-
-Marking whole classes or modules
-----------------------------------------------------
-
-If you are programming with Python2.6 you may use ``pytest.mark`` decorators
-with classes to apply markers to all its test methods::
-
-    # content of test_mark_classlevel.py
-    import pytest
-    @pytest.mark.webtest
-    class TestClass:
-        def test_startup(self):
-            pass
-        def test_startup_and_more(self):
-            pass
-
-This is equivalent to directly applying the decorator to the
-two test functions.
-
-To remain compatible with Python2.5 you can also set a
-``pytestmark`` attribute on a TestClass like this::
-
-    import pytest
-
-    class TestClass:
-        pytestmark = pytest.mark.webtest
-
-or if you need to use multiple markers you can use a list::
-
-    import pytest
-
-    class TestClass:
-        pytestmark = [pytest.mark.webtest, pytest.mark.slowtest]
-
-You can also set a module level marker::
-
-    import pytest
-    pytestmark = pytest.mark.webtest
-
-in which case it will be applied to all functions and
-methods defined in the module.
-
-Using ``-k TEXT`` to select tests
-----------------------------------------------------
-
-You can use the ``-k`` command line option to select tests::
-
-    $ py.test -k webtest  # running with the above defined examples yields
-    =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: /tmp/doc-exec-74
-    
-    test_mark.py ..
-    test_mark_classlevel.py ..
-    
-    ========================= 4 passed in 0.01 seconds =========================
-
-And you can also run all tests except the ones that match the keyword::
-
-    $ py.test -k-webtest
-    =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: /tmp/doc-exec-74
-    
-    ===================== 4 tests deselected by '-webtest' =====================
-    ======================= 4 deselected in 0.01 seconds =======================
-
-Or to only select the class::
-
-    $ py.test -kTestClass
-    =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: /tmp/doc-exec-74
-    
-    test_mark_classlevel.py ..
-    
-    ==================== 2 tests deselected by 'TestClass' =====================
-    ================== 2 passed, 2 deselected in 0.01 seconds ==================
 
 API reference for mark related objects
 ------------------------------------------------

doc/monkeypatch.txt

@@ -1,5 +1,5 @@
 
-monkeypatching/mocking modules and environments
+Monkeypatching/mocking modules and environments
 ================================================================
 
 .. currentmodule:: _pytest.monkeypatch
@@ -9,46 +9,39 @@ on global settings or which invokes code which cannot be easily
 tested such as network access.  The ``monkeypatch`` function argument
 helps you to safely set/delete an attribute, dictionary item or
 environment variable or to modify ``sys.path`` for importing.
-See the `monkeypatch blog post`_ one some introduction material
-and motivation.
+See the `monkeypatch blog post`_ for some introduction material
+and a discussion of its motivation.
 
 .. _`monkeypatch blog post`: http://tetamap.wordpress.com/2009/03/03/monkeypatching-in-unit-tests-done-right/
 
-
-Simple example: patching ``os.path.expanduser``
+Simple example: monkeypatching functions
 ---------------------------------------------------
 
-If you e.g. want to pretend that ``os.expanduser`` returns a certain
+If you want to pretend that ``os.expanduser`` returns a certain
 directory, you can use the :py:meth:`monkeypatch.setattr` method to
 patch this function before calling into a function which uses it::
 
+    # content of test_module.py
     import os.path
     def getssh(): # pseudo application code
-        return os.path.join(os.expanduser("~admin"), '.ssh')
+        return os.path.join(os.path.expanduser("~admin"), '.ssh')
 
     def test_mytest(monkeypatch):
         def mockreturn(path):
             return '/abc'
         monkeypatch.setattr(os.path, 'expanduser', mockreturn)
         x = getssh()
-        assert x == '/abc'
+        assert x == '/abc/.ssh'
 
-After the test function finishes the ``os.path.expanduser`` modification
-will be undone.
-
-.. background check:
-   $ py.test
-   =========================== test session starts ============================
-   platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-   test path 1: /tmp/doc-exec-75
-   
-   =============================  in 0.00 seconds =============================
+Here our test function monkeypatches ``os.path.expanduser`` and
+then calls into an function that calls it.  After the test function 
+finishes the ``os.path.expanduser`` modification will be undone.
 
 Method reference of the monkeypatch function argument
 -----------------------------------------------------
 
 .. autoclass:: monkeypatch
-    :members: setattr, delattr, setitem, delitem, setenv, delenv, syspath_prepend, undo
+    :members: setattr, delattr, setitem, delitem, setenv, delenv, syspath_prepend, chdir, undo
 
 ``monkeypatch.setattr/delattr/delitem/delenv()`` all
 by default raise an Exception if the target does not exist.

doc/naming20.txt

@@ -16,5 +16,5 @@ these renaming rules::
     py.test.cmdline.main -> pytest.main
 
 The old ``py.test.*`` ways to access functionality remain
-valid but you are encouraged to do global renames according
+valid but you are encouraged to do global renaming according
 to the above rules in your test code.

doc/nose.txt

@@ -1,4 +1,4 @@
-Running test written for nose
+Running tests written for nose
 =======================================
 
 .. include:: links.inc

doc/overview.txt

@@ -5,6 +5,7 @@ Getting started basics
 .. toctree::
    :maxdepth: 2
 
+   index.txt
    getting-started.txt
    usage.txt
    goodpractises.txt

doc/plugins.txt

@@ -1,7 +1,7 @@
-Writing, managing and understanding plugins
-=============================================
+.. _plugins:
 
-.. _`local plugin`:
+Working with plugins and conftest files
+=============================================
 
 py.test implements all aspects of configuration, collection, running and reporting by calling `well specified hooks`_.  Virtually any Python module can be registered as a plugin.  It can implement any number of hook functions (usually two or three) which all have a ``pytest_`` prefix, making hook functions easy to distinguish and find.  There are three basic locations types:
 
@@ -11,14 +11,18 @@ py.test implements all aspects of configuration, collection, running and reporti
 
 .. _`pytest/plugin`: http://bitbucket.org/hpk42/pytest/src/tip/pytest/plugin/
 .. _`conftest.py plugins`:
+.. _`conftest.py`:
+.. _`localplugin`:
+.. _`conftest`:
 
 conftest.py: local per-directory plugins
 --------------------------------------------------------------
 
 local ``conftest.py`` plugins contain directory-specific hook
 implementations.  Session and test running activities will
-invoke all hooks defined in "higher up" ``conftest.py`` files.
-Example: Assume the following layout and content of files::
+invoke all hooks defined in ``conftest.py`` files closer to the
+root of the filesystem.  Example: Assume the following layout
+and content of files::
 
     a/conftest.py:
         def pytest_runtest_setup(item):
@@ -38,21 +42,17 @@ Here is how you might run it::
      py.test test_flat.py   # will not show "setting up"
      py.test a/test_sub.py  # will show "setting up"
 
-A note on ordering: ``py.test`` loads all ``conftest.py`` files upwards
-from the command line file arguments.  It usually performs look up
-right-to-left, i.e.  the hooks in "closer" conftest files will be called
-earlier than further away ones.
-
 .. Note::
     If you have ``conftest.py`` files which do not reside in a
     python package directory (i.e. one containing an ``__init__.py``) then
-    "import conftest" can be ambigous because there might be other
+    "import conftest" can be ambiguous because there might be other
     ``conftest.py`` files as well on your PYTHONPATH or ``sys.path``.
     It is thus good practise for projects to either put ``conftest.py``
     under a package scope or to never import anything from a
     conftest.py file.
 
 .. _`external plugins`:
+.. _`extplugins`:
 
 Installing External Plugins / Searching
 ------------------------------------------------------
@@ -64,9 +64,26 @@ tool, for example::
     pip uninstall pytest-NAME
 
 If a plugin is installed,  py.test automatically finds and integrates it,
-there is no need to activate it.  If you don't need a plugin anymore simply
-de-install it.  You can find a list of available plugins through a
-`pytest- pypi.python.org search`_.
+there is no need to activate it.  Here is a list of known plugins:
+
+* `pytest-capturelog <http://pypi.python.org/pypi/pytest-capturelog>`_:
+  to capture and assert about messages from the logging module
+
+* `pytest-xdist <http://pypi.python.org/pypi/pytest-xdist>`_:
+  to distribute tests to CPUs and remote hosts, looponfailing mode,
+  see also :ref:`xdist`
+
+* `pytest-cov <http://pypi.python.org/pypi/pytest-cov>`_:
+  coverage reporting, compatible with distributed testing
+
+* `pytest-pep8 <http://pypi.python.org/pypi/pytest-pep8>`_:
+  a ``--pep8`` option to enable PEP8 compliance checking.
+
+* `oejskit <http://pypi.python.org/pypi/oejskit>`_:
+  a plugin to run javascript unittests in life browsers
+  (**version 0.8.9 not compatible with pytest-2.0**)
+
+You may discover more plugins through a `pytest- pypi.python.org search`_.
 
 .. _`available installable plugins`:
 .. _`pytest- pypi.python.org search`: http://pypi.python.org/pypi?%3Aaction=search&term=pytest-&submit=search
@@ -93,12 +110,12 @@ Making your plugin installable by others
 -----------------------------------------------
 
 If you want to make your plugin externally available, you
-may define a so called entry point for your distribution so
+may define a so-called entry point for your distribution so
 that ``py.test`` finds your plugin module.  Entry points are
 a feature that is provided by `setuptools`_ or `Distribute`_.
-The concrete entry point is ``pytest11``.  To make your plugin
-available you can insert the following lines in your
-setuptools/distribute-based setup-invocation:
+py.test looks up the ``pytest11`` entrypoint to discover its
+plugins and you can thus make your plugin available by definig
+it in your setuptools/distribute-based setup-invocation:
 
 .. sourcecode:: python
 
@@ -118,8 +135,8 @@ setuptools/distribute-based setup-invocation:
     )
 
 If a package is installed this way, py.test will load
-``myproject.pluginmodule`` and accordingly call functions
-if they match the `well specified hooks`_.
+``myproject.pluginmodule`` as a plugin which can define
+`well specified hooks`_.
 
 Plugin discovery order at tool startup
 --------------------------------------------
@@ -170,12 +187,42 @@ the plugin manager like this:
 If you want to look at the names of existing plugins, use
 the ``--traceconfig`` option.
 
+.. _`findpluginname`:
+
+Finding out which plugins are active
+----------------------------------------------------------------------------
+
+If you want to find out which plugins are active in your
+environment you can type::
+
+    py.test --traceconfig
+
+and will get an extended test header which shows activated plugins
+and their names. It will also print local plugins aka
+:ref:`conftest.py <conftest>` files when they are loaded.
+
+.. _`cmdunregister`:
+
+Deactivating / unregistering a plugin by name
+----------------------------------------------------------------------------
+
+You can prevent plugins from loading or unregister them::
+
+    py.test -p no:NAME
+
+This means that any subsequent try to activate/load the named
+plugin will it already existing.  See :ref:`findpluginname` for
+how to obtain the name of a plugin.
 
 .. _`builtin plugins`:
 
 py.test default plugin reference
 ====================================
 
+
+You can find the source code for the following plugins
+in the `pytest repository <http://bitbucket.org/hpk42/pytest/>`_.
+
 .. autosummary::
 
     _pytest.assertion
@@ -195,7 +242,7 @@ py.test default plugin reference
     _pytest.recwarn
     _pytest.resultlog
     _pytest.runner
-    _pytest.session
+    _pytest.main
     _pytest.skipping
     _pytest.terminal
     _pytest.tmpdir
@@ -206,22 +253,23 @@ py.test default plugin reference
 py.test hook reference
 ====================================
 
-hook specification and validation
+Hook specification and validation
 -----------------------------------------
 
 py.test calls hook functions to implement initialization, running,
 test execution and reporting.  When py.test loads a plugin it validates
-that all hook functions conform to their respective hook specification.
+that each hook function conforms to its respective hook specification.
 Each hook function name and its argument names need to match a hook
-specification exactly but it is allowed for a hook function to accept
-*less* parameters than specified.  If you mistype argument names or the
-hook name itself you get useful errors.
+specification.  However, a hook function may accept *fewer* parameters
+by simply not specifying them.  If you mistype argument names or the
+hook name itself you get an error showing the available arguments.
 
-initialisation, command line and configuration hooks
+Initialization, command line and configuration hooks
 --------------------------------------------------------------------
 
 .. currentmodule:: _pytest.hookspec
 
+.. autofunction:: pytest_cmdline_preparse
 .. autofunction:: pytest_cmdline_parse
 .. autofunction:: pytest_namespace
 .. autofunction:: pytest_addoption
@@ -229,7 +277,7 @@ initialisation, command line and configuration hooks
 .. autofunction:: pytest_configure
 .. autofunction:: pytest_unconfigure
 
-generic "runtest" hooks
+Generic "runtest" hooks
 ------------------------------
 
 All all runtest related hooks receive a :py:class:`pytest.Item` object.
@@ -242,13 +290,14 @@ All all runtest related hooks receive a :py:class:`pytest.Item` object.
 
 For deeper understanding you may look at the default implementation of
 these hooks in :py:mod:`_pytest.runner` and maybe also
-in :py:mod:`_pytest.pdb` which intercepts creation
-of reports in order to drop to interactive debugging.
+in :py:mod:`_pytest.pdb` which interacts with :py:mod:`_pytest.capture`
+and its input/output capturing in order to immediately drop
+into interactive debugging when a test failure occurs.
 
 The :py:mod:`_pytest.terminal` reported specifically uses
 the reporting hook to print information about a test run.
 
-collection hooks
+Collection hooks
 ------------------------------
 
 py.test calls the following hooks for collecting files and directories:
@@ -263,7 +312,7 @@ you can use the following hook:
 .. autofunction:: pytest_pycollect_makeitem
 
 
-reporting hooks
+Reporting hooks
 ------------------------------
 
 Session related reporting hooks:
@@ -278,7 +327,6 @@ test execution:
 
 .. autofunction: pytest_runtest_logreport
 
-
 Reference of important objects involved in hooks
 ===========================================================
 
@@ -288,14 +336,14 @@ Reference of important objects involved in hooks
 .. autoclass:: _pytest.config.Parser
     :members:
 
-.. autoclass:: _pytest.session.Node(name, parent)
+.. autoclass:: _pytest.main.Node(name, parent)
     :members:
 
 ..
-    .. autoclass:: _pytest.session.File(fspath, parent)
+    .. autoclass:: _pytest.main.File(fspath, parent)
         :members:
 
-    .. autoclass:: _pytest.session.Item(name, parent)
+    .. autoclass:: _pytest.main.Item(name, parent)
         :members:
 
     .. autoclass:: _pytest.python.Module(name, parent)
@@ -313,4 +361,3 @@ Reference of important objects involved in hooks
 .. autoclass:: _pytest.runner.TestReport
     :members:
 
-

doc/projects.txt

@@ -3,21 +3,28 @@
 Project examples
 ==========================
 
-Here are some examples of projects using py.test:
+Here are some examples of projects using py.test (please send notes via :ref:`contact`):
 
-* `PyPy <http://pypy.org>`_, Python with a JIT compiler, running over `16000 tests <http://test.pypy.org>`_
+* `PyPy <http://pypy.org>`_, Python with a JIT compiler, running over 
+  `16000 tests <http://buildbot.pypy.org/summary?branch=%3Ctrunk%3E>`_
 * the `MoinMoin <http://moinmo.in>`_ Wiki Engine
 * `tox <http://codespeak.net/tox>`_, virtualenv/Hudson integration tool
 * `PIDA <http://pida.co.uk>`_ framework for integrated development
+* `PyPM <http://code.activestate.com/pypm/>`_ ActiveState's package manager
 * `Fom <http://packages.python.org/Fom/>`_ a fluid object mapper for FluidDB
+* `applib <https://github.com/ActiveState/applib>`_ cross-platform utilities
 * `six <http://pypi.python.org/pypi/six/>`_ Python 2 and 3 compatibility utilities
 * `pediapress <http://code.pediapress.com/wiki/wiki>`_ MediaWiki articles
 * `mwlib <http://pypi.python.org/pypi/mwlib>`_ mediawiki parser and utility library
 * `The Translate Toolkit <http://translate.sourceforge.net/wiki/toolkit/index>`_ for localization and conversion
 * `execnet <http://codespeak.net/execnet>`_ rapid multi-Python deployment
+* `pylib <http://pylib.org>`_ cross-platform path, IO, dynamic code library
+* `Pacha <http://pacha.cafepais.com/>`_ configuration management in five minutes
 * `bbfreeze <http://pypi.python.org/pypi/bbfreeze>`_ create standalone executables from Python scripts
 * `pdb++ <http://bitbucket.org/antocuni/pdb>`_ a fancier version of PDB
 * `py-s3fuse <http://code.google.com/p/py-s3fuse/>`_ Amazon S3 FUSE based filesystem
+* `waskr <http://pacha.cafepais.com/>`_ WSGI Stats Middleware
+* `guachi <http://code.google.com/p/guachi/>`_ global persistent configs for Python modules
 * `Circuits <http://pypi.python.org/pypi/circuits>`_ lightweight Event Driven Framework
 * `pygtk-helpers <http://bitbucket.org/aafshar/pygtkhelpers-main/>`_ easy interaction with PyGTK
 * `QuantumCore <http://quantumcore.org/>`_ statusmessage and repoze openid plugin
@@ -31,15 +38,17 @@ Here are some examples of projects using py.test:
 * `bu <http://packages.python.org/bu/>`_ a microscopic build system
 * `katcp <https://bitbucket.org/hodgestar/katcp>`_ Telescope communication protocol over Twisted
 * `kss plugin timer <http://pypi.python.org/pypi/kss.plugin.timer>`_
-* many more ... (please send notes via the :ref:`contact`)
 
 Some organisations using py.test
 -----------------------------------
 
-* `Square Kilometre Array <http://ska.ac.za/>`_
+* `Square Kilometre Array, Cape Town <http://ska.ac.za/>`_
+* `Some Mozilla QA people <http://www.theautomatedtester.co.uk/blog/2011/pytest_and_xdist_plugin.html>`_ use pytest to distribute their Selenium tests
 * `Tandberg <http://www.tandberg.com/>`_
-* `Stups department of Heinrich Heine University <http://www.stups.uni-duesseldorf.de/projects.php>`_
-* `Open End <http://openend.se>`_
-* `Laboraratory of Bioinformatics <http://genesilico.pl/>`_
-* `merlinux <http://merlinux.eu>`_
+* `Shootq <http://web.shootq.com/>`_
+* `Stups department of Heinrich Heine University Duesseldorf <http://www.stups.uni-duesseldorf.de/projects.php>`_
+* `cellzome <http://www.cellzome.com/>`_
+* `Open End, Gothenborg <http://www.openend.se>`_
+* `Laboraratory of Bioinformatics, Warsaw <http://genesilico.pl/>`_
+* `merlinux, Germany <http://merlinux.eu>`_
 * many more ... (please be so kind to send a note via :ref:`contact`)

doc/pytest.ini

@@ -1,2 +1,2 @@
 [pytest]
-# just defined to prevent the root level tox.ini to kick in
+# just defined to prevent the root level tox.ini from kicking in

doc/recwarn.txt

@@ -1,8 +1,8 @@
 
-asserting deprecation and other warnings.
+Asserting deprecation and other warnings
 =====================================================
 
-recwarn function argument
+The recwarn function argument
 ------------------------------------
 
 You can use the ``recwarn`` funcarg to assert that code triggers
@@ -24,7 +24,7 @@ The ``recwarn`` function argument provides these methods:
 * ``pop(category=None)``: return last warning matching the category.
 * ``clear()``: clear list of warnings
 
-ensuring a function triggers a deprecation warning
+Ensuring a function triggers a deprecation warning
 -------------------------------------------------------
 
 You can also call a global helper for checking
@@ -35,4 +35,3 @@ warning::
 
     def test_global():
         pytest.deprecated_call(myfunction, 17)
-

doc/skipping.txt

@@ -1,49 +1,60 @@
-
 .. _`skip and xfail`:
 
-skip and xfail mechanisms
+Skip and xfail: dealing with tests that can not succeed
 =====================================================================
 
-You can skip or "xfail" test functions, either by marking functions
-through a decorator or by calling the ``pytest.skip|xfail`` helpers.
-A *skip* means that you expect your test to pass unless a certain configuration or condition (e.g. wrong Python interpreter, missing dependency) prevents it to run.  And *xfail* means that you expect your test to fail because there is an
-implementation problem.  py.test counts and lists *xfailing* tests separately
-and you can provide info such as a bug number or a URL to provide a
-human readable problem context.
+If you have test functions that cannot be run on certain platforms
+or that you expect to fail you can mark them accordingly or you
+may call helper functions during execution of setup or test functions.
 
-Usually detailed information about skipped/xfailed tests is not shown
-to avoid cluttering the output.  You can use the ``-r`` option to
-see details corresponding to the "short" letters shown in the
-test progress::
+A *skip* means that you expect your test to pass unless a certain
+configuration or condition (e.g. wrong Python interpreter, missing
+dependency) prevents it to run.  And *xfail* means that your test
+can run but you expect it to fail because there is an implementation problem.
 
-    py.test -rxs  # show extra info on skips and xfail tests
+py.test counts and lists *skip* and *xfail* tests separately. However,
+detailed information about skipped/xfailed tests is not shown by default
+to avoid cluttering the output.  You can use the ``-r`` option to see
+details corresponding to the "short" letters shown in the test
+progress::
+
+    py.test -rxs  # show extra info on skips and xfails
 
 (See :ref:`how to change command line options defaults`)
 
 .. _skipif:
 
-Skipping a single function
+Marking a test function to be skipped
 -------------------------------------------
 
-Here is an example for marking a test function to be skipped
+Here is an example of marking a test function to be skipped
 when run on a Python3 interpreter::
 
+    import sys
     @pytest.mark.skipif("sys.version_info >= (3,0)")
     def test_function():
         ...
 
 During test function setup the skipif condition is
-evaluated by calling ``eval(expr, namespace)``.  The namespace
-contains the  ``sys`` and ``os`` modules and the test
-``config`` object.  The latter allows you to skip based
-on a test configuration value e.g. like this::
+evaluated by calling ``eval('sys.version_info >= (3,0)', namespace)``.
+(*New in version 2.0.2*) The namespace contains all the module globals of the test function so that
+you can for example check for versions of a module you are using::
 
-    @pytest.mark.skipif("not config.getvalue('db')")
-    def test_function(...):
+    import mymodule
+
+    @pytest.mark.skipif("mymodule.__version__ < '1.2'")
+    def test_function():
         ...
+  
+The test function will not be run ("skipped") if
+``mymodule`` is below the specified version.  The reason
+for specifying the condition as a string is mainly that
+py.test can report a summary of skip conditions.
+For information on the construction of the ``namespace``
+see `evaluation of skipif/xfail conditions`_.
 
-Create a shortcut for your conditional skip decorator
-at module level like this::
+You can of course create a shortcut for your conditional skip
+decorator at module level like this::
 
     win32only = pytest.mark.skipif("sys.platform != 'win32'")
 
@@ -51,13 +62,12 @@ at module level like this::
     def test_function():
         ...
 
-
-skip test functions of a class
+Skip all test functions of a class
 --------------------------------------
 
-As with all function :ref:`marking` you can do it at
+As with all function :ref:`marking <mark>` you can skip test functions at the
 `whole class- or module level`_.  Here is an example
-for skipping all methods of a test class based on platform::
+for skipping all methods of a test class based on the platform::
 
     class TestPosixCalls:
         pytestmark = pytest.mark.skipif("sys.platform == 'win32'")
@@ -65,9 +75,10 @@ for skipping all methods of a test class based on platform::
         def test_function(self):
             "will not be setup or run under 'win32' platform"
 
-The ``pytestmark`` decorator will be applied to each test function.
-If your code targets python2.6 or above you can equivalently use
-the skipif decorator on classes::
+The ``pytestmark`` special name tells py.test to apply it to each test
+function in the class.  If your code targets python2.6 or above you can
+more naturally use the skipif decorator (and any other marker) on
+classes::
 
     @pytest.mark.skipif("sys.platform == 'win32'")
     class TestPosixCalls:
@@ -75,15 +86,13 @@ the skipif decorator on classes::
         def test_function(self):
             "will not be setup or run under 'win32' platform"
 
-It is fine in general to apply multiple "skipif" decorators
-on a single function - this means that if any of the conditions
-apply the function will be skipped.
+Using multiple "skipif" decorators on a single function is generally fine - it means that if any of the conditions apply the function execution will be skipped.
 
 .. _`whole class- or module level`: mark.html#scoped-marking
 
 .. _xfail:
 
-mark a test function as expected to fail
+Mark a test function as expected to fail
 -------------------------------------------------------
 
 You can use the ``xfail`` marker to indicate that you
@@ -104,16 +113,16 @@ By specifying on the commandline::
 you can force the running and reporting of an ``xfail`` marked test
 as if it weren't marked at all.
 
-Same as with skipif_ you can also selectively expect a failure
-depending on platform::
+As with skipif_ you can also mark your expectation of a failure
+on a particular platform::
 
     @pytest.mark.xfail("sys.version_info >= (3,0)")
     def test_function():
         ...
 
-You can also avoid running an "xfail" test at all or
+You can furthermore prevent the running of an "xfail" test or
 specify a reason such as a bug ID or similar.  Here is
-a simple test file with usages:
+a simple test file with the several usages:
 
 .. literalinclude:: example/xfail_demo.py
 
@@ -121,24 +130,51 @@ Running it with the report-on-xfail option gives this output::
 
     example $ py.test -rx xfail_demo.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev31
-    test path 1: xfail_demo.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 6 items
     
-    xfail_demo.py xxxxx
+    xfail_demo.py xxxxxx
     ========================= short test summary info ==========================
     XFAIL xfail_demo.py::test_hello
     XFAIL xfail_demo.py::test_hello2
-      reason: [NOTRUN]
+      reason: [NOTRUN] 
     XFAIL xfail_demo.py::test_hello3
       condition: hasattr(os, 'sep')
     XFAIL xfail_demo.py::test_hello4
       bug 110
     XFAIL xfail_demo.py::test_hello5
+      condition: pytest.__version__[0] != "17"
+    XFAIL xfail_demo.py::test_hello6
       reason: reason
     
-    ======================== 5 xfailed in 0.04 seconds =========================
+    ======================== 6 xfailed in 0.16 seconds =========================
 
-imperative xfail from within a test or setup function
+.. _`evaluation of skipif/xfail conditions`:
+
+Evaluation of skipif/xfail expressions
+----------------------------------------------------
+
+.. versionadded:: 2.0.2
+
+The evaluation of a condition string in ``pytest.mark.skipif(conditionstring)``
+or ``pytest.mark.xfail(conditionstring)`` takes place in a namespace
+dictionary which is constructed as follows:
+
+* the namespace is initialized by putting the ``sys`` and ``os`` modules
+  and the pytest ``config`` object into it.
+  
+* updated with the module globals of the test function for which the
+  expression is applied.
+
+The pytest ``config`` object allows you to skip based on a test configuration value
+which you might have added::
+
+    @pytest.mark.skipif("not config.getvalue('db')")
+    def test_function(...):
+        ...
+
+
+Imperative xfail from within a test or setup function
 ------------------------------------------------------
 
 If you cannot declare xfail-conditions at import time
@@ -147,10 +183,10 @@ within test or setup code.  Example::
 
     def test_function():
         if not valid_config():
-            pytest.xfail("unsuppored configuration")
+            pytest.xfail("unsupported configuration")
 
 
-skipping on a missing import dependency
+Skipping on a missing import dependency
 --------------------------------------------------
 
 You can use the following import helper at module level
@@ -159,21 +195,21 @@ or within a test or test setup function::
     docutils = pytest.importorskip("docutils")
 
 If ``docutils`` cannot be imported here, this will lead to a
-skip outcome of the test.  You can also skip dependeing if
-if a library does not come with a high enough version::
+skip outcome of the test.  You can also skip based on the
+version number of a library::
 
     docutils = pytest.importorskip("docutils", minversion="0.3")
 
 The version will be read from the specified module's ``__version__`` attribute.
 
-imperative skip from within a test or setup function
+Imperative skip from within a test or setup function
 ------------------------------------------------------
 
 If for some reason you cannot declare skip-conditions
-you can also imperatively produce a Skip-outcome from
+you can also imperatively produce a skip-outcome from
 within test or setup code.  Example::
 
     def test_function():
         if not valid_config():
-            pytest.skip("unsuppored configuration")
+            pytest.skip("unsupported configuration")
 

doc/talks.txt

@@ -4,35 +4,40 @@ Talks and Tutorials
 
 .. _`funcargs`: funcargs.html
 
-tutorial examples and blog postings
+Tutorial examples and blog postings
 ---------------------------------------------
 
 .. _`tutorial1 repository`: http://bitbucket.org/hpk42/pytest-tutorial1/
 .. _`pycon 2010 tutorial PDF`: http://bitbucket.org/hpk42/pytest-tutorial1/raw/tip/pytest-basic.pdf
 
-basic usage and funcargs:
+Basic usage and funcargs:
 
 - `pycon 2010 tutorial PDF`_ and `tutorial1 repository`_
 
-function arguments:
+Function arguments:
 
 - :ref:`mysetup`
 - `application setup in test functions with funcargs`_
 - `monkey patching done right`_ (blog post, consult `monkeypatch
   plugin`_ for actual 1.0 API)
 
-test parametrization:
+Test parametrization:
 
-- `generating parametrized tests with funcargs`_
+- `generating parametrized tests with funcargs`_ (uses deprecated ``addcall()`` API.
 - `test generators and cached setup`_
 - `parametrizing tests, generalized`_ (blog post)
 - `putting test-hooks into local or global plugins`_ (blog post)
 
-distributed testing:
+Assertion introspection:
 
-- `simultanously test your code on all platforms`_ (blog entry)
+- `(07/2011) Behind the scenes of py.test's new assertion rewriting
+  <http://pybites.blogspot.com/2011/07/behind-scenes-of-pytests-new-assertion.html>`_ 
 
-plugin specific examples:
+Distributed testing:
+
+- `simultaneously test your code on all platforms`_ (blog entry)
+
+Plugin specific examples:
 
 - `skipping slow tests by default in py.test`_ (blog entry)
 
@@ -42,14 +47,14 @@ plugin specific examples:
 .. _`many examples in the docs for plugins`: plugin/index.html
 .. _`monkeypatch plugin`: plugin/monkeypatch.html
 .. _`application setup in test functions with funcargs`: funcargs.html#appsetup
-.. _`simultanously test your code on all platforms`: http://tetamap.wordpress.com/2009/03/23/new-simultanously-test-your-code-on-all-platforms/
+.. _`simultaneously test your code on all platforms`: http://tetamap.wordpress.com/2009/03/23/new-simultanously-test-your-code-on-all-platforms/
 .. _`monkey patching done right`: http://tetamap.wordpress.com/2009/03/03/monkeypatching-in-unit-tests-done-right/
 .. _`putting test-hooks into local or global plugins`: http://tetamap.wordpress.com/2009/05/14/putting-test-hooks-into-local-and-global-plugins/
 .. _`parametrizing tests, generalized`: http://tetamap.wordpress.com/2009/05/13/parametrizing-python-tests-generalized/
 .. _`generating parametrized tests with funcargs`: funcargs.html#test-generators
 .. _`test generators and cached setup`: http://bruynooghe.blogspot.com/2010/06/pytest-test-generators-and-cached-setup.html
 
-conference talks and tutorials
+Conference talks and tutorials
 ----------------------------------------
 
 - `ep2009-rapidtesting.pdf`_ tutorial slides (July 2009):

doc/test/index.txt

@@ -17,7 +17,7 @@ customize_: configuration, customization, extensions
 
 changelog_: history of changes covering last releases
 
-**Continous Integration of py.test's own tests and plugins with Hudson**:
+**Continuous Integration of py.test's own tests and plugins with Hudson**:
 
     `http://hudson.testrun.org/view/pytest`_
 

doc/test/mission.txt

@@ -5,7 +5,7 @@ Mission
 py.test strives to make testing a fun and no-boilerplate effort.
 
 The tool is distributed as part of the `py` package which contains supporting APIs that
-are also useable independently.  The project independent ``py.test`` command line tool helps you to:
+are also usable independently.  The project independent ``py.test`` command line tool helps you to:
 
 * rapidly collect and run tests
 * run unit- or doctests, functional or integration tests

doc/test/plugin/cov.txt

@@ -35,10 +35,11 @@ However easy_install does not provide an uninstall facility.
 
 .. IMPORTANT::
 
-    Ensure that you manually delete the init_covmain.pth file in your site-packages directory.
+    Ensure that you manually delete the init_covmain.pth file in your
+    site-packages directory.
 
-    This file starts coverage collection of subprocesses if appropriate during site initialisation
-    at python startup.
+    This file starts coverage collection of subprocesses if appropriate during
+    site initialization at python startup.
 
 
 Usage
@@ -198,11 +199,11 @@ slave.
 
 For subprocess measurement environment variables must make it from the main process to the
 subprocess.  The python used by the subprocess must have pytest-cov installed.  The subprocess must
-do normal site initialisation so that the environment variables can be detected and coverage
+do normal site initialization so that the environment variables can be detected and coverage
 started.
 
 
-Acknowledgements
+Acknowledgments
 ----------------
 
 Holger Krekel for pytest with its distributed testing support.

doc/test/plugin/oejskit.txt

@@ -1,7 +1,7 @@
 pytest_oejskit plugin (EXTERNAL)
 ==========================================
 
-The `oejskit`_ offers a py.test plugin for running Javascript tests in life browers.   Running inside the browsers comes with some speed cost, on the other hand it means for example the code is tested against the real-word DOM implementations.
+The `oejskit`_ offers a py.test plugin for running Javascript tests in life browsers.   Running inside the browsers comes with some speed cost, on the other hand it means for example the code is tested against the real-word DOM implementations.
 The approach enables to write integration tests such that the JavaScript code is tested against server-side Python code mocked as necessary. Any server-side framework that can already be exposed through WSGI (or for which a subset of WSGI can be written to accommodate the jskit own needs) can play along.
 
 For more info and download please visit the `oejskit PyPI`_ page.

doc/test/plugin/xdist.txt

@@ -129,7 +129,7 @@ put options values in a ``conftest.py`` file like this::
     option_tx = ['ssh=myhost//python=python2.5', 'popen//python=python2.5']
     option_dist = True
 
-Any commandline ``--tx`` specifictions  will add to the list of
+Any commandline ``--tx`` specifications  will add to the list of
 available execution environments.
 
 Specifying "rsync" dirs in a conftest.py

doc/tmpdir.txt

@@ -1,10 +1,10 @@
 
 .. _`tmpdir handling`:
 
-temporary directories and files
+Temporary directories and files
 ================================================
 
-the 'tmpdir' test function argument
+The 'tmpdir' test function argument
 -----------------------------------
 
 You can use the ``tmpdir`` function argument which will
@@ -28,15 +28,15 @@ Running this would result in a passed test except for the last
 
     $ py.test test_tmpdir.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_tmpdir.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
     
     test_tmpdir.py F
     
     ================================= FAILURES =================================
     _____________________________ test_create_file _____________________________
     
-    tmpdir = local('/tmp/pytest-123/test_create_file0')
+    tmpdir = local('/Users/hpk/tmp/pytest-21/test_create_file0')
     
         def test_create_file(tmpdir):
             p = tmpdir.mkdir("sub").join("hello.txt")
@@ -47,16 +47,16 @@ Running this would result in a passed test except for the last
     E       assert 0
     
     test_tmpdir.py:7: AssertionError
-    ========================= 1 failed in 0.02 seconds =========================
+    ========================= 1 failed in 0.07 seconds =========================
 
 .. _`base temporary directory`:
 
-the default base temporary directory
+The default base temporary directory
 -----------------------------------------------
 
-Temporary directories are by default created as sub directories of
+Temporary directories are by default created as sub-directories of
 the system temporary directory.  The base name will be ``pytest-NUM`` where
-``NUM`` will be incremenated with each test run.  Moreover, entries older
+``NUM`` will be incremented with each test run.  Moreover, entries older
 than 3 temporary directories will be removed.
 
 You can override the default temporary directory setting like this::
@@ -64,8 +64,8 @@ You can override the default temporary directory setting like this::
     py.test --basetemp=mydir
 
 When distributing tests on the local machine, ``py.test`` takes care to
-configure a basetemp directory for the sub processes such that all
-temporary data lands below below a single per-test run basetemp directory.
+configure a basetemp directory for the sub processes such that all temporary
+data lands below a single per-test run basetemp directory.
 
 .. _`py.path.local`: http://pylib.org/path.html
 

doc/unittest.txt

@@ -1,14 +1,14 @@
 
 .. _`unittest.TestCase`:
 
-unittest.TestCase support
+Support for unittest.TestCase
 =====================================================================
 
 py.test has limited support for running Python `unittest.py style`_ tests.
 It will automatically collect ``unittest.TestCase`` subclasses
 and their ``test`` methods in test files.  It will invoke
 ``setUp/tearDown`` methods but also perform py.test's standard ways
-of treating tests like e.g. IO capturing::
+of treating tests such as IO capturing::
 
     # content of test_unittest.py
 
@@ -24,8 +24,8 @@ Running it yields::
 
     $ py.test test_unittest.py
     =========================== test session starts ============================
-    platform linux2 -- Python 2.6.5 -- pytest-2.0.0.dev30
-    test path 1: test_unittest.py
+    platform darwin -- Python 2.7.1 -- pytest-2.2.2
+    collecting ... collected 1 items
     
     test_unittest.py F
     
@@ -37,26 +37,12 @@ Running it yields::
         def test_method(self):
             x = 1
     >       self.assertEquals(x, 3)
+    E       AssertionError: 1 != 3
     
-    test_unittest.py:8: 
-    _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
-    
-    self = <test_unittest.MyTest testMethod=test_method>, first = 1, second = 3
-    msg = None
-    
-        def failUnlessEqual(self, first, second, msg=None):
-            """Fail if the two objects are unequal as determined by the '=='
-                   operator.
-                """
-            if not first == second:
-                raise self.failureException, \
-    >                 (msg or '%r != %r' % (first, second))
-    E                 AssertionError: 1 != 3
-    
-    /usr/lib/python2.6/unittest.py:350: AssertionError
+    test_unittest.py:8: AssertionError
     ----------------------------- Captured stdout ------------------------------
     hello
-    ========================= 1 failed in 0.02 seconds =========================
+    ========================= 1 failed in 0.15 seconds =========================
 
 .. _`unittest.py style`: http://docs.python.org/library/unittest.html