feat: add pcre2 lib to precheck geometry wkt

cmake/cmake.options

@@ -64,13 +64,6 @@ IF(${TD_WINDOWS})
             ON
         )
 
-    MESSAGE("build geos Win32")
-    option(
-            BUILD_GEOS
-            "If build geos on Windows"
-            ON
-        )
-
 ELSEIF (TD_DARWIN_64)
     IF(${BUILD_TEST})
         add_definitions(-DCOMPILER_SUPPORTS_CXX13)
@@ -79,7 +72,7 @@ ENDIF ()
 
 option(
         BUILD_GEOS
-        "If build geos on Windows"
+        "If build with geos"
         ON
     )
 
@@ -95,6 +88,12 @@ option(
     ON
     )
 
+option(
+    BUILD_PCRE2
+    "If build with pcre2"
+    ON
+)
+
 option(
     JEMALLOC_ENABLED
     "If build with jemalloc"

cmake/pcre2_CMakeLists.txt.in

@@ -0,0 +1,13 @@
+
+# pcre2
+ExternalProject_Add(pcre2
+        GIT_REPOSITORY https://github.com/PCRE2Project/pcre2.git
+        GIT_TAG pcre2-10.43
+        SOURCE_DIR "${TD_CONTRIB_DIR}/pcre2"
+        #BINARY_DIR "${TD_CONTRIB_DIR}/pcre2"
+        #BUILD_IN_SOURCE TRUE
+        CONFIGURE_COMMAND ""
+        BUILD_COMMAND ""
+        INSTALL_COMMAND ""
+        TEST_COMMAND ""
+        )

contrib/CMakeLists.txt

@@ -183,6 +183,11 @@ if(${BUILD_GEOS})
     cat("${TD_SUPPORT_DIR}/geos_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
 endif()
 
+#
+if(${BUILD_PCRE2})
+    cat("${TD_SUPPORT_DIR}/pcre2_CMakeLists.txt.in" ${CONTRIB_TMP_FILE})
+endif()
+
 # download dependencies
 configure_file(${CONTRIB_TMP_FILE} "${TD_CONTRIB_DIR}/deps-download/CMakeLists.txt")
 execute_process(COMMAND "${CMAKE_COMMAND}" -G "${CMAKE_GENERATOR}" .
@@ -601,6 +606,10 @@ if(${BUILD_GEOS})
     )
 endif(${BUILD_GEOS})
 
+if (${BUILD_PCRE2})
+    add_subdirectory(pcre2 EXCLUDE_FROM_ALL)
+endif(${BUILD_PCRE2})
+
 # ================================================================================================
 # Build test
 # ================================================================================================

include/util/tgeosctx.h

@@ -21,6 +21,7 @@ extern "C" {
 #endif
 
 #include <geos_c.h>
+#include <tpcre2.h>
 
 typedef struct SGeosContext {
   GEOSContextHandle_t handle;
@@ -31,6 +32,9 @@ typedef struct SGeosContext {
   GEOSWKBReader *WKBReader;
   GEOSWKBWriter *WKBWriter;
 
+  pcre2_code *WKTRegex;
+  pcre2_match_data *WKTMatchData;
+
   char errMsg[512];
 } SGeosContext;
 

include/util/tpcre2.h

@@ -0,0 +1,34 @@
+/*
+ * Copyright (c) 2019 TAOS Data, Inc. <jhtao@taosdata.com>
+ *
+ * This program is free software: you can use, redistribute, and/or modify
+ * it under the terms of the GNU Affero General Public License, version 3
+ * or later ("AGPL"), as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef _TD_ULIT_PCRE2_H_
+#define _TD_ULIT_PCRE2_H_
+
+#define PCRE2_CODE_UNIT_WIDTH 8
+#include "pcre2.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+int32_t doRegComp(pcre2_code** ppRegex, pcre2_match_data** ppMatchData, const char* pattern);
+int32_t doRegExec(const char* pString, pcre2_code* pRegex, pcre2_match_data* pMatchData);
+void    destroyRegexes(pcre2_code* pWktRegex, pcre2_match_data* pWktMatchData);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif  // _TD_UTIL_PAGEDBUF_H_

source/libs/geometry/src/geosWrapper.c

@@ -86,6 +86,62 @@ _exit:
   return code;
 }
 
+static int initWktRegex(pcre2_code **ppRegex, pcre2_match_data **ppMatchData) {
+  int   ret = 0;
+  char *wktPatternWithSpace = taosMemoryCalloc(4, 1024);
+  sprintf(
+      wktPatternWithSpace,
+      "^( *)point( *)z?m?( *)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *)\\)))|linestring( *)z?m?( "
+      "*)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))|polygon( *)z?m?( "
+      "*)((empty)|(\\(( *)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))(( *)(,)( "
+      "*)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))( *))*( "
+      "*)\\)))|multipoint( *)z?m?( *)((empty)|(\\(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))|multilinestring( "
+      "*)z?m?( *)((empty)|(\\(( *)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))(( *)(,)( "
+      "*)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))( *))*( "
+      "*)\\)))|multipolygon( *)z?m?( *)((empty)|(\\(( *)((empty)|(\\(( *)((empty)|(\\(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))(( *)(,)( "
+      "*)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))( *))*( *)\\)))(( *)(,)( "
+      "*)((empty)|(\\(( *)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))(( *)(,)( "
+      "*)((empty)|(\\(( *)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}(( *)(,)( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?(( "
+      "*)(([-+]?[0-9]+\\.?[0-9]*)|([-+]?[0-9]*\\.?[0-9]+))(e[-+]?[0-9]+)?){1,3}( *))*( *)\\)))( *))*( *)\\)))( *))*( "
+      "*)\\)))|(GEOCOLLECTION\\((?R)(( *)(,)( *)(?R))*( *)\\))( *)$");
+
+  ret = doRegComp(ppRegex, ppMatchData, wktPatternWithSpace);
+  taosMemoryFree(wktPatternWithSpace);
+  return ret;
+}
+
 int32_t initCtxGeomFromText() {
   int32_t code = TSDB_CODE_FAILED;
   SGeosContext* geosCtx = getThreadLocalGeosCtx();
@@ -113,6 +169,10 @@ int32_t initCtxGeomFromText() {
     }
   }
 
+  if (geosCtx->WKTRegex == NULL) {
+    if (initWktRegex(&geosCtx->WKTRegex, &geosCtx->WKTMatchData) != 0) return code;
+  }
+
   return TSDB_CODE_SUCCESS;
 }
 
@@ -125,6 +185,11 @@ int32_t doGeomFromText(const char *inputWKT, unsigned char **outputGeom, size_t
   GEOSGeometry *geom = NULL;
   unsigned char *wkb = NULL;
 
+  if (doRegExec(inputWKT, geosCtx->WKTRegex, geosCtx->WKTMatchData) != 0) {
+     code = TSDB_CODE_FUNC_FUNTION_PARA_VALUE;
+     goto _exit;
+  }
+
   geom = GEOSWKTReader_read_r(geosCtx->handle, geosCtx->WKTReader, inputWKT);
   if (geom == NULL) {
     code = TSDB_CODE_FUNC_FUNTION_PARA_VALUE;

source/util/CMakeLists.txt

@@ -21,27 +21,27 @@ target_include_directories(
     PRIVATE "${TD_SOURCE_DIR}/utils/TSZ/sz/inc"
     PRIVATE "${TD_SOURCE_DIR}/utils/TSZ/zstd/"
     PRIVATE "${TD_SOURCE_DIR}/contrib/lzma2/"
+    PRIVATE "${TD_SOURCE_DIR}/contrib/pcre2/"
 )
 
 target_link_directories(
     util
     PUBLIC "${TD_SOURCE_DIR}/contrib/lzma2"
+    PUBLIC "${TD_SOURCE_DIR}/contrib/pcre2"
 )
 
-
 if (TD_LINUX)
     target_link_libraries(
         util
         PUBLIC os common
-        PUBLIC lz4_static fast-lzma2 
+        PUBLIC lz4_static fast-lzma2 pcre2-8
         PUBLIC api cjson geos_c TSZ
-    
     )
 else()
     target_link_libraries(
     util
     PUBLIC os common
-    PUBLIC lz4_static 
+    PUBLIC lz4_static pcre2-8
     PUBLIC api cjson geos_c TSZ
     )
 endif()

source/util/src/tgeosctx.c

@@ -43,6 +43,10 @@ void destroyThreadLocalGeosCtx() {
     tlGeosCtx.WKBWriter = NULL;
   }
 
+  if (tlGeosCtx.WKTRegex) {
+    destroyRegexes(tlGeosCtx.WKTRegex, tlGeosCtx.WKTMatchData);
+  }
+
   if(tlGeosCtx.handle) {
     GEOS_finish_r(tlGeosCtx.handle);
     tlGeosCtx.handle = NULL;

source/util/src/tpcr2.c

@@ -0,0 +1,35 @@
+#include "tpcre2.h"
+
+int32_t doRegComp(pcre2_code** ppRegex, pcre2_match_data** ppMatchData, const char* pattern) {
+  uint32_t   options = PCRE2_CASELESS;
+  int        errorcode;
+  PCRE2_SIZE erroroffset;
+
+  *ppRegex = pcre2_compile((PCRE2_SPTR8)pattern, PCRE2_ZERO_TERMINATED, options, &errorcode, &erroroffset, NULL);
+  if (*ppRegex == NULL) {
+    PCRE2_UCHAR buffer[256];
+    pcre2_get_error_message(errorcode, buffer, sizeof(buffer));
+    return 1;
+  }
+
+  *ppMatchData = pcre2_match_data_create_from_pattern(*ppRegex, NULL);
+
+  return 0;
+}
+
+int32_t doRegExec(const char* pString, pcre2_code* pRegex, pcre2_match_data* pMatchData) {
+  int32_t ret = 0;
+  ret = pcre2_match(pRegex, (PCRE2_SPTR)pString, PCRE2_ZERO_TERMINATED, 0, 0, pMatchData, NULL);
+  if (ret < 0) {
+    PCRE2_UCHAR buffer[256];
+    pcre2_get_error_message(ret, buffer, sizeof(buffer));
+    return 1;
+  }
+
+  return (ret > 0) ? 0 : 1;
+}
+
+void destroyRegexes(pcre2_code* pWktRegex, pcre2_match_data* pWktMatchData) {
+  pcre2_code_free(pWktRegex);
+  pcre2_match_data_free(pWktMatchData);
+}