From 08119c361d1181b3e8f1abb429236e488a664753 Mon Sep 17 00:00:00 2001 From: Hans-Christoph Steiner Date: Tue, 13 Aug 2013 15:42:54 -0400 Subject: Imported Upstream version 2.2.1 --- ext/fts3/fts3_tokenize_vtab.c | 454 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 454 insertions(+) create mode 100644 ext/fts3/fts3_tokenize_vtab.c (limited to 'ext/fts3/fts3_tokenize_vtab.c') diff --git a/ext/fts3/fts3_tokenize_vtab.c b/ext/fts3/fts3_tokenize_vtab.c new file mode 100644 index 0000000..364852e --- /dev/null +++ b/ext/fts3/fts3_tokenize_vtab.c @@ -0,0 +1,454 @@ +/* +** 2013 Apr 22 +** +** The author disclaims copyright to this source code. In place of +** a legal notice, here is a blessing: +** +** May you do good and not evil. +** May you find forgiveness for yourself and forgive others. +** May you share freely, never taking more than you give. +** +****************************************************************************** +** +** This file contains code for the "fts3tokenize" virtual table module. +** An fts3tokenize virtual table is created as follows: +** +** CREATE VIRTUAL TABLE USING fts3tokenize( +** , , ... +** ); +** +** The table created has the following schema: +** +** CREATE TABLE (input, token, start, end, position) +** +** When queried, the query must include a WHERE clause of type: +** +** input = +** +** The virtual table module tokenizes this , using the FTS3 +** tokenizer specified by the arguments to the CREATE VIRTUAL TABLE +** statement and returns one row for each token in the result. With +** fields set as follows: +** +** input: Always set to a copy of +** token: A token from the input. +** start: Byte offset of the token within the input . +** end: Byte offset of the byte immediately following the end of the +** token within the input string. +** pos: Token offset of token within input. +** +*/ +#include "fts3Int.h" +#if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS3) + +#include +#include + +typedef struct Fts3tokTable Fts3tokTable; +typedef struct Fts3tokCursor Fts3tokCursor; + +/* +** Virtual table structure. +*/ +struct Fts3tokTable { + sqlite3_vtab base; /* Base class used by SQLite core */ + const sqlite3_tokenizer_module *pMod; + sqlite3_tokenizer *pTok; +}; + +/* +** Virtual table cursor structure. +*/ +struct Fts3tokCursor { + sqlite3_vtab_cursor base; /* Base class used by SQLite core */ + char *zInput; /* Input string */ + sqlite3_tokenizer_cursor *pCsr; /* Cursor to iterate through zInput */ + int iRowid; /* Current 'rowid' value */ + const char *zToken; /* Current 'token' value */ + int nToken; /* Size of zToken in bytes */ + int iStart; /* Current 'start' value */ + int iEnd; /* Current 'end' value */ + int iPos; /* Current 'pos' value */ +}; + +/* +** Query FTS for the tokenizer implementation named zName. +*/ +static int fts3tokQueryTokenizer( + Fts3Hash *pHash, + const char *zName, + const sqlite3_tokenizer_module **pp, + char **pzErr +){ + sqlite3_tokenizer_module *p; + int nName = (int)strlen(zName); + + p = (sqlite3_tokenizer_module *)sqlite3Fts3HashFind(pHash, zName, nName+1); + if( !p ){ + *pzErr = sqlite3_mprintf("unknown tokenizer: %s", zName); + return SQLITE_ERROR; + } + + *pp = p; + return SQLITE_OK; +} + +/* +** The second argument, argv[], is an array of pointers to nul-terminated +** strings. This function makes a copy of the array and strings into a +** single block of memory. It then dequotes any of the strings that appear +** to be quoted. +** +** If successful, output parameter *pazDequote is set to point at the +** array of dequoted strings and SQLITE_OK is returned. The caller is +** responsible for eventually calling sqlite3_free() to free the array +** in this case. Or, if an error occurs, an SQLite error code is returned. +** The final value of *pazDequote is undefined in this case. +*/ +static int fts3tokDequoteArray( + int argc, /* Number of elements in argv[] */ + const char * const *argv, /* Input array */ + char ***pazDequote /* Output array */ +){ + int rc = SQLITE_OK; /* Return code */ + if( argc==0 ){ + *pazDequote = 0; + }else{ + int i; + int nByte = 0; + char **azDequote; + + for(i=0; ixCreate((nDequote>1 ? nDequote-1 : 0), azArg, &pTok); + } + + if( rc==SQLITE_OK ){ + pTab = (Fts3tokTable *)sqlite3_malloc(sizeof(Fts3tokTable)); + if( pTab==0 ){ + rc = SQLITE_NOMEM; + } + } + + if( rc==SQLITE_OK ){ + memset(pTab, 0, sizeof(Fts3tokTable)); + pTab->pMod = pMod; + pTab->pTok = pTok; + *ppVtab = &pTab->base; + }else{ + if( pTok ){ + pMod->xDestroy(pTok); + } + } + + sqlite3_free(azDequote); + return rc; +} + +/* +** This function does the work for both the xDisconnect and xDestroy methods. +** These tables have no persistent representation of their own, so xDisconnect +** and xDestroy are identical operations. +*/ +static int fts3tokDisconnectMethod(sqlite3_vtab *pVtab){ + Fts3tokTable *pTab = (Fts3tokTable *)pVtab; + + pTab->pMod->xDestroy(pTab->pTok); + sqlite3_free(pTab); + return SQLITE_OK; +} + +/* +** xBestIndex - Analyze a WHERE and ORDER BY clause. +*/ +static int fts3tokBestIndexMethod( + sqlite3_vtab *pVTab, + sqlite3_index_info *pInfo +){ + int i; + UNUSED_PARAMETER(pVTab); + + for(i=0; inConstraint; i++){ + if( pInfo->aConstraint[i].usable + && pInfo->aConstraint[i].iColumn==0 + && pInfo->aConstraint[i].op==SQLITE_INDEX_CONSTRAINT_EQ + ){ + pInfo->idxNum = 1; + pInfo->aConstraintUsage[i].argvIndex = 1; + pInfo->aConstraintUsage[i].omit = 1; + pInfo->estimatedCost = 1; + return SQLITE_OK; + } + } + + pInfo->idxNum = 0; + assert( pInfo->estimatedCost>1000000.0 ); + + return SQLITE_OK; +} + +/* +** xOpen - Open a cursor. +*/ +static int fts3tokOpenMethod(sqlite3_vtab *pVTab, sqlite3_vtab_cursor **ppCsr){ + Fts3tokCursor *pCsr; + UNUSED_PARAMETER(pVTab); + + pCsr = (Fts3tokCursor *)sqlite3_malloc(sizeof(Fts3tokCursor)); + if( pCsr==0 ){ + return SQLITE_NOMEM; + } + memset(pCsr, 0, sizeof(Fts3tokCursor)); + + *ppCsr = (sqlite3_vtab_cursor *)pCsr; + return SQLITE_OK; +} + +/* +** Reset the tokenizer cursor passed as the only argument. As if it had +** just been returned by fts3tokOpenMethod(). +*/ +static void fts3tokResetCursor(Fts3tokCursor *pCsr){ + if( pCsr->pCsr ){ + Fts3tokTable *pTab = (Fts3tokTable *)(pCsr->base.pVtab); + pTab->pMod->xClose(pCsr->pCsr); + pCsr->pCsr = 0; + } + sqlite3_free(pCsr->zInput); + pCsr->zInput = 0; + pCsr->zToken = 0; + pCsr->nToken = 0; + pCsr->iStart = 0; + pCsr->iEnd = 0; + pCsr->iPos = 0; + pCsr->iRowid = 0; +} + +/* +** xClose - Close a cursor. +*/ +static int fts3tokCloseMethod(sqlite3_vtab_cursor *pCursor){ + Fts3tokCursor *pCsr = (Fts3tokCursor *)pCursor; + + fts3tokResetCursor(pCsr); + sqlite3_free(pCsr); + return SQLITE_OK; +} + +/* +** xNext - Advance the cursor to the next row, if any. +*/ +static int fts3tokNextMethod(sqlite3_vtab_cursor *pCursor){ + Fts3tokCursor *pCsr = (Fts3tokCursor *)pCursor; + Fts3tokTable *pTab = (Fts3tokTable *)(pCursor->pVtab); + int rc; /* Return code */ + + pCsr->iRowid++; + rc = pTab->pMod->xNext(pCsr->pCsr, + &pCsr->zToken, &pCsr->nToken, + &pCsr->iStart, &pCsr->iEnd, &pCsr->iPos + ); + + if( rc!=SQLITE_OK ){ + fts3tokResetCursor(pCsr); + if( rc==SQLITE_DONE ) rc = SQLITE_OK; + } + + return rc; +} + +/* +** xFilter - Initialize a cursor to point at the start of its data. +*/ +static int fts3tokFilterMethod( + sqlite3_vtab_cursor *pCursor, /* The cursor used for this query */ + int idxNum, /* Strategy index */ + const char *idxStr, /* Unused */ + int nVal, /* Number of elements in apVal */ + sqlite3_value **apVal /* Arguments for the indexing scheme */ +){ + int rc = SQLITE_ERROR; + Fts3tokCursor *pCsr = (Fts3tokCursor *)pCursor; + Fts3tokTable *pTab = (Fts3tokTable *)(pCursor->pVtab); + UNUSED_PARAMETER(idxStr); + UNUSED_PARAMETER(nVal); + + fts3tokResetCursor(pCsr); + if( idxNum==1 ){ + const char *zByte = (const char *)sqlite3_value_text(apVal[0]); + int nByte = sqlite3_value_bytes(apVal[0]); + pCsr->zInput = sqlite3_malloc(nByte+1); + if( pCsr->zInput==0 ){ + rc = SQLITE_NOMEM; + }else{ + memcpy(pCsr->zInput, zByte, nByte); + pCsr->zInput[nByte] = 0; + rc = pTab->pMod->xOpen(pTab->pTok, pCsr->zInput, nByte, &pCsr->pCsr); + if( rc==SQLITE_OK ){ + pCsr->pCsr->pTokenizer = pTab->pTok; + } + } + } + + if( rc!=SQLITE_OK ) return rc; + return fts3tokNextMethod(pCursor); +} + +/* +** xEof - Return true if the cursor is at EOF, or false otherwise. +*/ +static int fts3tokEofMethod(sqlite3_vtab_cursor *pCursor){ + Fts3tokCursor *pCsr = (Fts3tokCursor *)pCursor; + return (pCsr->zToken==0); +} + +/* +** xColumn - Return a column value. +*/ +static int fts3tokColumnMethod( + sqlite3_vtab_cursor *pCursor, /* Cursor to retrieve value from */ + sqlite3_context *pCtx, /* Context for sqlite3_result_xxx() calls */ + int iCol /* Index of column to read value from */ +){ + Fts3tokCursor *pCsr = (Fts3tokCursor *)pCursor; + + /* CREATE TABLE x(input, token, start, end, position) */ + switch( iCol ){ + case 0: + sqlite3_result_text(pCtx, pCsr->zInput, -1, SQLITE_TRANSIENT); + break; + case 1: + sqlite3_result_text(pCtx, pCsr->zToken, pCsr->nToken, SQLITE_TRANSIENT); + break; + case 2: + sqlite3_result_int(pCtx, pCsr->iStart); + break; + case 3: + sqlite3_result_int(pCtx, pCsr->iEnd); + break; + default: + assert( iCol==4 ); + sqlite3_result_int(pCtx, pCsr->iPos); + break; + } + return SQLITE_OK; +} + +/* +** xRowid - Return the current rowid for the cursor. +*/ +static int fts3tokRowidMethod( + sqlite3_vtab_cursor *pCursor, /* Cursor to retrieve value from */ + sqlite_int64 *pRowid /* OUT: Rowid value */ +){ + Fts3tokCursor *pCsr = (Fts3tokCursor *)pCursor; + *pRowid = (sqlite3_int64)pCsr->iRowid; + return SQLITE_OK; +} + +/* +** Register the fts3tok module with database connection db. Return SQLITE_OK +** if successful or an error code if sqlite3_create_module() fails. +*/ +int sqlite3Fts3InitTok(sqlite3 *db, Fts3Hash *pHash){ + static const sqlite3_module fts3tok_module = { + 0, /* iVersion */ + fts3tokConnectMethod, /* xCreate */ + fts3tokConnectMethod, /* xConnect */ + fts3tokBestIndexMethod, /* xBestIndex */ + fts3tokDisconnectMethod, /* xDisconnect */ + fts3tokDisconnectMethod, /* xDestroy */ + fts3tokOpenMethod, /* xOpen */ + fts3tokCloseMethod, /* xClose */ + fts3tokFilterMethod, /* xFilter */ + fts3tokNextMethod, /* xNext */ + fts3tokEofMethod, /* xEof */ + fts3tokColumnMethod, /* xColumn */ + fts3tokRowidMethod, /* xRowid */ + 0, /* xUpdate */ + 0, /* xBegin */ + 0, /* xSync */ + 0, /* xCommit */ + 0, /* xRollback */ + 0, /* xFindFunction */ + 0, /* xRename */ + 0, /* xSavepoint */ + 0, /* xRelease */ + 0 /* xRollbackTo */ + }; + int rc; /* Return code */ + + rc = sqlite3_create_module(db, "fts3tokenize", &fts3tok_module, (void*)pHash); + return rc; +} + +#endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS3) */ -- cgit v1.2.3