add fuzzy search

source/libs/index/inc/indexFstDfa.h

@@ -0,0 +1,72 @@
+/*
+ * 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 __INDEX_FST_DFA_H__
+#define __INDEX_FST_DFA_H__
+
+#include "indexFstRegex.h"
+#include "indexFstSparse.h"
+#include "tarray.h"
+#include "thash.h"
+
+#ifdef __cplusplus
+
+extern "C" {
+#endif
+
+typedef struct FstDfa FstDfa;
+
+typedef struct {
+  SArray * insts;
+  uint32_t next[256];
+  bool     isMatch;
+} State;
+
+/*
+ * dfa builder related func
+ **/
+typedef struct FstDfaBuilder {
+  FstDfa *  dfa;
+  SHashObj *cache;
+} FstDfaBuilder;
+
+FstDfaBuilder *dfaBuilderCreate(SArray *insts);
+
+FstDfa *dfaBuilderBuild(FstDfaBuilder *builder);
+
+bool dfaBuilderRunState(FstDfaBuilder *builder, FstSparseSet *cur, FstSparseSet *next, uint32_t state, uint8_t bytes,
+                        uint32_t *result);
+
+bool dfaBuilderCachedState(FstDfaBuilder *builder, FstSparseSet *set, uint32_t *result);
+
+/*
+ * dfa related func
+ **/
+typedef struct FstDfa {
+  SArray *insts;
+  SArray *states;
+} FstDfa;
+
+FstDfa *dfaCreate(SArray *insts, SArray *states);
+bool    dfaIsMatch(FstDfa *dfa, uint32_t si);
+bool    dfaAccept(FstDfa *dfa, uint32_t si, uint8_t byte, uint32_t *result);
+void    dfaAdd(FstDfa *dfa, FstSparseSet *set, uint32_t ip);
+bool    dfaRun(FstDfa *dfa, FstSparseSet *from, FstSparseSet *to, uint8_t byte);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

source/libs/index/inc/indexFstRegex.h

@@ -0,0 +1,74 @@
+/*
+ * 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_INDEX_FST_REGEX_H_
+#define _TD_INDEX_FST_REGEX_H_
+
+//#include "indexFstDfa.h"
+#include "taos.h"
+#include "tarray.h"
+#include "tchecksum.h"
+#include "thash.h"
+#include "tlog.h"
+#include "tutil.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+typedef enum { MATCH, JUMP, SPLIT, RANGE } InstType;
+
+typedef struct MatchValue {
+} MatchValue;
+typedef struct JumpValue {
+  uint32_t step;
+} JumpValue;
+
+typedef struct SplitValue {
+  uint32_t len1;
+  uint32_t len2;
+} SplitValue;
+
+typedef struct RangeValue {
+  uint8_t start;
+  uint8_t end;
+} RangeValue;
+
+typedef struct {
+  InstType ty;
+  union {
+    MatchValue mv;
+    JumpValue  jv;
+    SplitValue sv;
+    RangeValue rv;
+  };
+} Inst;
+
+typedef struct {
+  char *orig;
+  void *dfa;
+} FstRegex;
+
+FstRegex *regexCreate(const char *str);
+
+void regexSetup(FstRegex *regex, uint32_t size, const char *str);
+
+// uint32_t regexStart()
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

source/libs/index/inc/indexFstSparse.h

@@ -13,8 +13,8 @@
  * along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#ifndef _TD_INDEX_SPARSE_H_
+#ifndef _TD_INDEX_FST_SPARSE_H_
-#define _TD_INDEX_SPARSE_H_
+#define _TD_INDEX_FST_SPARSE_H_
 
 #include "tarray.h"
 

source/libs/index/src/indexFstDfa.c

@@ -0,0 +1,96 @@
+/*
+ * 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/>.
+ */
+
+#include "indexFstDfa.h"
+#include "thash.h"
+
+static int dfaInstsEqual(const void *a, const void *b, size_t size) {
+  SArray *ar = (SArray *)a;
+  SArray *br = (SArray *)b;
+  size_t  al = ar != NULL ? taosArrayGetSize(ar) : 0;
+  size_t  bl = br != NULL ? taosArrayGetSize(br) : 0;
+  if (al != bl) {
+    return -1;
+  }
+  for (int i = 0; i < al; i++) {
+    uint32_t v1 = *(uint32_t *)taosArrayGet(ar, i);
+    uint32_t v2 = *(uint32_t *)taosArrayGet(br, i);
+    if (v1 != v2) {
+      return -1;
+    }
+  }
+  return 0;
+}
+FstDfaBuilder *dfaBuilderCreate(SArray *insts) {
+  FstDfaBuilder *builder = taosMemoryCalloc(1, sizeof(FstDfaBuilder));
+  if (builder == NULL) {
+    return NULL;
+  }
+
+  SArray *states = taosArrayInit(4, sizeof(State));
+
+  builder->dfa = dfaCreate(insts, states);
+  builder->cache = taosHashInit(
+      4, taosGetDefaultHashFunction(POINTER_BYTES == sizeof(int64_t) ? TSDB_DATA_TYPE_BIGINT : TSDB_DATA_TYPE_INT),
+      false, HASH_NO_LOCK);
+  taosHashSetEqualFp(builder->cache, dfaInstsEqual);
+  return builder;
+}
+
+FstDfa *dfaBuilderBuild(FstDfaBuilder *builder) {
+  uint32_t      sz = taosArrayGetSize(builder->dfa->insts);
+  FstSparseSet *cur = sparSetCreate(sz);
+  FstSparseSet *nxt = sparSetCreate(sz);
+
+  dfaAdd(builder->dfa, cur, 0);
+}
+
+bool dfaBuilderRunState(FstDfaBuilder *builder, FstSparseSet *cur, FstSparseSet *next, uint32_t state, uint8_t bytes,
+                        uint32_t *result) {
+  // impl run state
+  return true;
+}
+
+bool dfaBuilderCachedState(FstDfaBuilder *builder, FstSparseSet *set, uint32_t *result) {
+  // impl cache state
+  return true;
+}
+
+FstDfa *dfaCreate(SArray *insts, SArray *states) {
+  FstDfa *dfa = taosMemoryCalloc(1, sizeof(FstDfa));
+  if (dfa == NULL) {
+    return NULL;
+  }
+
+  dfa->insts = insts;
+  dfa->states = states;
+  return dfa;
+}
+bool dfaIsMatch(FstDfa *dfa, uint32_t si) {
+  // impl match
+  return true;
+}
+bool dfaAccept(FstDfa *dfa, uint32_t si, uint8_t byte, uint32_t *result) {
+  // impl accept
+  return true;
+}
+void dfaAdd(FstDfa *dfa, FstSparseSet *set, uint32_t ip) {
+  // impl add
+  return;
+}
+bool dfaRun(FstDfa *dfa, FstSparseSet *from, FstSparseSet *to, uint8_t byte) {
+  // impl run
+  return true;
+}

source/libs/index/src/indexFstRegex.c

@@ -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/>.
+ */
+
+#include "indexFstRegex.h"
+#include "indexFstSparse.h"
+
+FstRegex *regexCreate(const char *str) {
+  FstRegex *regex = taosMemoryCalloc(1, sizeof(FstRegex));
+  if (regex == NULL) {
+    return NULL;
+  }
+  int32_t sz = (int32_t)strlen(str);
+  char *  orig = taosMemoryCalloc(1, sz);
+  memcpy(orig, str, sz);
+
+  regex->orig = orig;
+}
+
+void regexSetup(FstRegex *regex, uint32_t size, const char *str) {
+  // return
+  // return;
+}

source/libs/index/src/indexSparse.c

@@ -13,7 +13,7 @@
  * along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "indexSparse.h"
+#include "indexFstSparse.h"
 
 FstSparseSet *sparSetCreate(int32_t sz) {
   FstSparseSet *ss = taosMemoryCalloc(1, sizeof(FstSparseSet));