diff options
Diffstat (limited to 'lib/libsqlite3/ext/fts3')
-rw-r--r-- | lib/libsqlite3/ext/fts3/fts3.c | 65 | ||||
-rw-r--r-- | lib/libsqlite3/ext/fts3/fts3Int.h | 8 | ||||
-rw-r--r-- | lib/libsqlite3/ext/fts3/fts3_expr.c | 205 | ||||
-rw-r--r-- | lib/libsqlite3/ext/fts3/fts3_unicode.c | 16 | ||||
-rw-r--r-- | lib/libsqlite3/ext/fts3/fts3_unicode2.c | 10 | ||||
-rw-r--r-- | lib/libsqlite3/ext/fts3/fts3_write.c | 359 | ||||
-rw-r--r-- | lib/libsqlite3/ext/fts3/tool/fts3view.c | 15 | ||||
-rw-r--r-- | lib/libsqlite3/ext/fts3/unicode/mkunicode.tcl | 10 |
8 files changed, 484 insertions, 204 deletions
diff --git a/lib/libsqlite3/ext/fts3/fts3.c b/lib/libsqlite3/ext/fts3/fts3.c index 44b7f431dfe..4f4b6674309 100644 --- a/lib/libsqlite3/ext/fts3/fts3.c +++ b/lib/libsqlite3/ext/fts3/fts3.c @@ -1333,7 +1333,7 @@ static int fts3InitVtab( p->bHasStat = isFts4; p->bFts4 = isFts4; p->bDescIdx = bDescIdx; - p->bAutoincrmerge = 0xff; /* 0xff means setting unknown */ + p->nAutoincrmerge = 0xff; /* 0xff means setting unknown */ p->zContentTbl = zContent; p->zLanguageid = zLanguageid; zContent = 0; @@ -1376,7 +1376,9 @@ static int fts3InitVtab( int n = (int)strlen(p->azColumn[iCol]); for(i=0; i<nNotindexed; i++){ char *zNot = azNotindexed[i]; - if( zNot && 0==sqlite3_strnicmp(p->azColumn[iCol], zNot, n) ){ + if( zNot && n==(int)strlen(zNot) + && 0==sqlite3_strnicmp(p->azColumn[iCol], zNot, n) + ){ p->abNotindexed[iCol] = 1; sqlite3_free(zNot); azNotindexed[i] = 0; @@ -1410,10 +1412,7 @@ static int fts3InitVtab( ** addition of a %_stat table so that it can use incremental merge. */ if( !isFts4 && !isCreate ){ - int rc2 = SQLITE_OK; - fts3DbExec(&rc2, db, "SELECT 1 FROM %Q.'%q_stat' WHERE id=2", - p->zDb, p->zName); - if( rc2==SQLITE_OK ) p->bHasStat = 1; + p->bHasStat = 2; } /* Figure out the page-size for the database. This is required in order to @@ -3305,7 +3304,10 @@ static int fts3SyncMethod(sqlite3_vtab *pVtab){ Fts3Table *p = (Fts3Table*)pVtab; int rc = sqlite3Fts3PendingTermsFlush(p); - if( rc==SQLITE_OK && p->bAutoincrmerge==1 && p->nLeafAdd>(nMinMerge/16) ){ + if( rc==SQLITE_OK + && p->nLeafAdd>(nMinMerge/16) + && p->nAutoincrmerge && p->nAutoincrmerge!=0xff + ){ int mxLevel = 0; /* Maximum relative level value in db */ int A; /* Incr-merge parameter A */ @@ -3313,14 +3315,41 @@ static int fts3SyncMethod(sqlite3_vtab *pVtab){ assert( rc==SQLITE_OK || mxLevel==0 ); A = p->nLeafAdd * mxLevel; A += (A/2); - if( A>(int)nMinMerge ) rc = sqlite3Fts3Incrmerge(p, A, 8); + if( A>(int)nMinMerge ) rc = sqlite3Fts3Incrmerge(p, A, p->nAutoincrmerge); } sqlite3Fts3SegmentsClose(p); return rc; } /* -** Implementation of xBegin() method. This is a no-op. +** If it is currently unknown whether or not the FTS table has an %_stat +** table (if p->bHasStat==2), attempt to determine this (set p->bHasStat +** to 0 or 1). Return SQLITE_OK if successful, or an SQLite error code +** if an error occurs. +*/ +static int fts3SetHasStat(Fts3Table *p){ + int rc = SQLITE_OK; + if( p->bHasStat==2 ){ + const char *zFmt ="SELECT 1 FROM %Q.sqlite_master WHERE tbl_name='%q_stat'"; + char *zSql = sqlite3_mprintf(zFmt, p->zDb, p->zName); + if( zSql ){ + sqlite3_stmt *pStmt = 0; + rc = sqlite3_prepare_v2(p->db, zSql, -1, &pStmt, 0); + if( rc==SQLITE_OK ){ + int bHasStat = (sqlite3_step(pStmt)==SQLITE_ROW); + rc = sqlite3_finalize(pStmt); + if( rc==SQLITE_OK ) p->bHasStat = bHasStat; + } + sqlite3_free(zSql); + }else{ + rc = SQLITE_NOMEM; + } + } + return rc; +} + +/* +** Implementation of xBegin() method. */ static int fts3BeginMethod(sqlite3_vtab *pVtab){ Fts3Table *p = (Fts3Table*)pVtab; @@ -3331,7 +3360,7 @@ static int fts3BeginMethod(sqlite3_vtab *pVtab){ TESTONLY( p->inTransaction = 1 ); TESTONLY( p->mxSavepoint = -1; ); p->nLeafAdd = 0; - return SQLITE_OK; + return fts3SetHasStat(p); } /* @@ -3580,6 +3609,10 @@ static int fts3RenameMethod( sqlite3 *db = p->db; /* Database connection */ int rc; /* Return Code */ + /* At this point it must be known if the %_stat table exists or not. + ** So bHasStat may not be 2. */ + rc = fts3SetHasStat(p); + /* As it happens, the pending terms table is always empty here. This is ** because an "ALTER TABLE RENAME TABLE" statement inside a transaction ** always opens a savepoint transaction. And the xSavepoint() method @@ -3587,7 +3620,9 @@ static int fts3RenameMethod( ** PendingTermsFlush() in in case that changes. */ assert( p->nPendingData==0 ); - rc = sqlite3Fts3PendingTermsFlush(p); + if( rc==SQLITE_OK ){ + rc = sqlite3Fts3PendingTermsFlush(p); + } if( p->zContentTbl==0 ){ fts3DbExec(&rc, db, @@ -3715,7 +3750,7 @@ static void hashDestroy(void *p){ */ void sqlite3Fts3SimpleTokenizerModule(sqlite3_tokenizer_module const**ppModule); void sqlite3Fts3PorterTokenizerModule(sqlite3_tokenizer_module const**ppModule); -#ifdef SQLITE_ENABLE_FTS4_UNICODE61 +#ifndef SQLITE_DISABLE_FTS3_UNICODE void sqlite3Fts3UnicodeTokenizer(sqlite3_tokenizer_module const**ppModule); #endif #ifdef SQLITE_ENABLE_ICU @@ -3733,7 +3768,7 @@ int sqlite3Fts3Init(sqlite3 *db){ Fts3Hash *pHash = 0; const sqlite3_tokenizer_module *pSimple = 0; const sqlite3_tokenizer_module *pPorter = 0; -#ifdef SQLITE_ENABLE_FTS4_UNICODE61 +#ifndef SQLITE_DISABLE_FTS3_UNICODE const sqlite3_tokenizer_module *pUnicode = 0; #endif @@ -3742,7 +3777,7 @@ int sqlite3Fts3Init(sqlite3 *db){ sqlite3Fts3IcuTokenizerModule(&pIcu); #endif -#ifdef SQLITE_ENABLE_FTS4_UNICODE61 +#ifndef SQLITE_DISABLE_FTS3_UNICODE sqlite3Fts3UnicodeTokenizer(&pUnicode); #endif @@ -3770,7 +3805,7 @@ int sqlite3Fts3Init(sqlite3 *db){ if( sqlite3Fts3HashInsert(pHash, "simple", 7, (void *)pSimple) || sqlite3Fts3HashInsert(pHash, "porter", 7, (void *)pPorter) -#ifdef SQLITE_ENABLE_FTS4_UNICODE61 +#ifndef SQLITE_DISABLE_FTS3_UNICODE || sqlite3Fts3HashInsert(pHash, "unicode61", 10, (void *)pUnicode) #endif #ifdef SQLITE_ENABLE_ICU diff --git a/lib/libsqlite3/ext/fts3/fts3Int.h b/lib/libsqlite3/ext/fts3/fts3Int.h index 7f717368885..b2827b7352f 100644 --- a/lib/libsqlite3/ext/fts3/fts3Int.h +++ b/lib/libsqlite3/ext/fts3/fts3Int.h @@ -210,20 +210,20 @@ struct Fts3Table { sqlite3_tokenizer *pTokenizer; /* tokenizer for inserts and queries */ char *zContentTbl; /* content=xxx option, or NULL */ char *zLanguageid; /* languageid=xxx option, or NULL */ - u8 bAutoincrmerge; /* True if automerge=1 */ + int nAutoincrmerge; /* Value configured by 'automerge' */ u32 nLeafAdd; /* Number of leaf blocks added this trans */ /* Precompiled statements used by the implementation. Each of these ** statements is run and reset within a single virtual table API call. */ - sqlite3_stmt *aStmt[37]; + sqlite3_stmt *aStmt[40]; char *zReadExprlist; char *zWriteExprlist; int nNodeSize; /* Soft limit for node size */ u8 bFts4; /* True for FTS4, false for FTS3 */ - u8 bHasStat; /* True if %_stat table exists */ + u8 bHasStat; /* True if %_stat table exists (2==unknown) */ u8 bHasDocsize; /* True if %_docsize table exists */ u8 bDescIdx; /* True if doclists are in reverse order */ u8 bIgnoreSavepoint; /* True to ignore xSavepoint invocations */ @@ -585,7 +585,7 @@ int sqlite3Fts3MsrIncrRestart(Fts3MultiSegReader *pCsr); int sqlite3Fts3InitTok(sqlite3*, Fts3Hash *); /* fts3_unicode2.c (functions generated by parsing unicode text files) */ -#ifdef SQLITE_ENABLE_FTS4_UNICODE61 +#ifndef SQLITE_DISABLE_FTS3_UNICODE int sqlite3FtsUnicodeFold(int, int); int sqlite3FtsUnicodeIsalnum(int); int sqlite3FtsUnicodeIsdiacritic(int); diff --git a/lib/libsqlite3/ext/fts3/fts3_expr.c b/lib/libsqlite3/ext/fts3/fts3_expr.c index 9c71f26ba1d..f5d28cbfcc7 100644 --- a/lib/libsqlite3/ext/fts3/fts3_expr.c +++ b/lib/libsqlite3/ext/fts3/fts3_expr.c @@ -185,40 +185,23 @@ static int getNextToken( int rc; sqlite3_tokenizer_cursor *pCursor; Fts3Expr *pRet = 0; - int nConsumed = 0; + int i = 0; - rc = sqlite3Fts3OpenTokenizer(pTokenizer, pParse->iLangid, z, n, &pCursor); + /* Set variable i to the maximum number of bytes of input to tokenize. */ + for(i=0; i<n; i++){ + if( sqlite3_fts3_enable_parentheses && (z[i]=='(' || z[i]==')') ) break; + if( z[i]=='*' || z[i]=='"' ) break; + } + + *pnConsumed = i; + rc = sqlite3Fts3OpenTokenizer(pTokenizer, pParse->iLangid, z, i, &pCursor); if( rc==SQLITE_OK ){ const char *zToken; int nToken = 0, iStart = 0, iEnd = 0, iPosition = 0; int nByte; /* total space to allocate */ rc = pModule->xNext(pCursor, &zToken, &nToken, &iStart, &iEnd, &iPosition); - - if( (rc==SQLITE_OK || rc==SQLITE_DONE) && sqlite3_fts3_enable_parentheses ){ - int i; - if( rc==SQLITE_DONE ) iStart = n; - for(i=0; i<iStart; i++){ - if( z[i]=='(' ){ - pParse->nNest++; - rc = fts3ExprParse(pParse, &z[i+1], n-i-1, &pRet, &nConsumed); - if( rc==SQLITE_OK && !pRet ){ - rc = SQLITE_DONE; - } - nConsumed = (int)(i + 1 + nConsumed); - break; - } - - if( z[i]==')' ){ - rc = SQLITE_DONE; - pParse->nNest--; - nConsumed = i+1; - break; - } - } - } - - if( nConsumed==0 && rc==SQLITE_OK ){ + if( rc==SQLITE_OK ){ nByte = sizeof(Fts3Expr) + sizeof(Fts3Phrase) + nToken; pRet = (Fts3Expr *)fts3MallocZero(nByte); if( !pRet ){ @@ -252,13 +235,14 @@ static int getNextToken( } } - nConsumed = iEnd; + *pnConsumed = iEnd; + }else if( i && rc==SQLITE_DONE ){ + rc = SQLITE_OK; } pModule->xClose(pCursor); } - *pnConsumed = nConsumed; *ppExpr = pRet; return rc; } @@ -508,6 +492,21 @@ static int getNextNode( return getNextString(pParse, &zInput[1], ii-1, ppExpr); } + if( sqlite3_fts3_enable_parentheses ){ + if( *zInput=='(' ){ + int nConsumed = 0; + pParse->nNest++; + rc = fts3ExprParse(pParse, zInput+1, nInput-1, ppExpr, &nConsumed); + if( rc==SQLITE_OK && !*ppExpr ){ rc = SQLITE_DONE; } + *pnConsumed = (int)(zInput - z) + 1 + nConsumed; + return rc; + }else if( *zInput==')' ){ + pParse->nNest--; + *pnConsumed = (int)((zInput - z) + 1); + *ppExpr = 0; + return SQLITE_DONE; + } + } /* If control flows to this point, this must be a regular token, or ** the end of the input. Read a regular token using the sqlite3_tokenizer @@ -626,96 +625,100 @@ static int fts3ExprParse( while( rc==SQLITE_OK ){ Fts3Expr *p = 0; int nByte = 0; + rc = getNextNode(pParse, zIn, nIn, &p, &nByte); + assert( nByte>0 || (rc!=SQLITE_OK && p==0) ); if( rc==SQLITE_OK ){ - int isPhrase; - - if( !sqlite3_fts3_enable_parentheses - && p->eType==FTSQUERY_PHRASE && pParse->isNot - ){ - /* Create an implicit NOT operator. */ - Fts3Expr *pNot = fts3MallocZero(sizeof(Fts3Expr)); - if( !pNot ){ - sqlite3Fts3ExprFree(p); - rc = SQLITE_NOMEM; - goto exprparse_out; - } - pNot->eType = FTSQUERY_NOT; - pNot->pRight = p; - p->pParent = pNot; - if( pNotBranch ){ - pNot->pLeft = pNotBranch; - pNotBranch->pParent = pNot; - } - pNotBranch = pNot; - p = pPrev; - }else{ - int eType = p->eType; - isPhrase = (eType==FTSQUERY_PHRASE || p->pLeft); - - /* The isRequirePhrase variable is set to true if a phrase or - ** an expression contained in parenthesis is required. If a - ** binary operator (AND, OR, NOT or NEAR) is encounted when - ** isRequirePhrase is set, this is a syntax error. - */ - if( !isPhrase && isRequirePhrase ){ - sqlite3Fts3ExprFree(p); - rc = SQLITE_ERROR; - goto exprparse_out; - } - - if( isPhrase && !isRequirePhrase ){ - /* Insert an implicit AND operator. */ - Fts3Expr *pAnd; - assert( pRet && pPrev ); - pAnd = fts3MallocZero(sizeof(Fts3Expr)); - if( !pAnd ){ + if( p ){ + int isPhrase; + + if( !sqlite3_fts3_enable_parentheses + && p->eType==FTSQUERY_PHRASE && pParse->isNot + ){ + /* Create an implicit NOT operator. */ + Fts3Expr *pNot = fts3MallocZero(sizeof(Fts3Expr)); + if( !pNot ){ sqlite3Fts3ExprFree(p); rc = SQLITE_NOMEM; goto exprparse_out; } - pAnd->eType = FTSQUERY_AND; - insertBinaryOperator(&pRet, pPrev, pAnd); - pPrev = pAnd; - } + pNot->eType = FTSQUERY_NOT; + pNot->pRight = p; + p->pParent = pNot; + if( pNotBranch ){ + pNot->pLeft = pNotBranch; + pNotBranch->pParent = pNot; + } + pNotBranch = pNot; + p = pPrev; + }else{ + int eType = p->eType; + isPhrase = (eType==FTSQUERY_PHRASE || p->pLeft); + + /* The isRequirePhrase variable is set to true if a phrase or + ** an expression contained in parenthesis is required. If a + ** binary operator (AND, OR, NOT or NEAR) is encounted when + ** isRequirePhrase is set, this is a syntax error. + */ + if( !isPhrase && isRequirePhrase ){ + sqlite3Fts3ExprFree(p); + rc = SQLITE_ERROR; + goto exprparse_out; + } - /* This test catches attempts to make either operand of a NEAR - ** operator something other than a phrase. For example, either of - ** the following: - ** - ** (bracketed expression) NEAR phrase - ** phrase NEAR (bracketed expression) - ** - ** Return an error in either case. - */ - if( pPrev && ( + if( isPhrase && !isRequirePhrase ){ + /* Insert an implicit AND operator. */ + Fts3Expr *pAnd; + assert( pRet && pPrev ); + pAnd = fts3MallocZero(sizeof(Fts3Expr)); + if( !pAnd ){ + sqlite3Fts3ExprFree(p); + rc = SQLITE_NOMEM; + goto exprparse_out; + } + pAnd->eType = FTSQUERY_AND; + insertBinaryOperator(&pRet, pPrev, pAnd); + pPrev = pAnd; + } + + /* This test catches attempts to make either operand of a NEAR + ** operator something other than a phrase. For example, either of + ** the following: + ** + ** (bracketed expression) NEAR phrase + ** phrase NEAR (bracketed expression) + ** + ** Return an error in either case. + */ + if( pPrev && ( (eType==FTSQUERY_NEAR && !isPhrase && pPrev->eType!=FTSQUERY_PHRASE) || (eType!=FTSQUERY_PHRASE && isPhrase && pPrev->eType==FTSQUERY_NEAR) - )){ - sqlite3Fts3ExprFree(p); - rc = SQLITE_ERROR; - goto exprparse_out; - } - - if( isPhrase ){ - if( pRet ){ - assert( pPrev && pPrev->pLeft && pPrev->pRight==0 ); - pPrev->pRight = p; - p->pParent = pPrev; + )){ + sqlite3Fts3ExprFree(p); + rc = SQLITE_ERROR; + goto exprparse_out; + } + + if( isPhrase ){ + if( pRet ){ + assert( pPrev && pPrev->pLeft && pPrev->pRight==0 ); + pPrev->pRight = p; + p->pParent = pPrev; + }else{ + pRet = p; + } }else{ - pRet = p; + insertBinaryOperator(&pRet, pPrev, p); } - }else{ - insertBinaryOperator(&pRet, pPrev, p); + isRequirePhrase = !isPhrase; } - isRequirePhrase = !isPhrase; + pPrev = p; } assert( nByte>0 ); } assert( rc!=SQLITE_OK || (nByte>0 && nByte<=nIn) ); nIn -= nByte; zIn += nByte; - pPrev = p; } if( rc==SQLITE_DONE && pRet && isRequirePhrase ){ diff --git a/lib/libsqlite3/ext/fts3/fts3_unicode.c b/lib/libsqlite3/ext/fts3/fts3_unicode.c index 188358eade4..94fc27b5b4e 100644 --- a/lib/libsqlite3/ext/fts3/fts3_unicode.c +++ b/lib/libsqlite3/ext/fts3/fts3_unicode.c @@ -13,7 +13,7 @@ ** Implementation of the "unicode" full-text-search tokenizer. */ -#ifdef SQLITE_ENABLE_FTS4_UNICODE61 +#ifndef SQLITE_DISABLE_FTS3_UNICODE #include "fts3Int.h" #if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS3) @@ -231,7 +231,7 @@ static int unicodeCreate( for(i=0; rc==SQLITE_OK && i<nArg; i++){ const char *z = azArg[i]; - int n = strlen(z); + int n = (int)strlen(z); if( n==19 && memcmp("remove_diacritics=1", z, 19)==0 ){ pNew->bRemoveDiacritic = 1; @@ -318,7 +318,7 @@ static int unicodeNext( ){ unicode_cursor *pCsr = (unicode_cursor *)pC; unicode_tokenizer *p = ((unicode_tokenizer *)pCsr->base.pTokenizer); - int iCode; + int iCode = 0; char *zOut; const unsigned char *z = &pCsr->aInput[pCsr->iOff]; const unsigned char *zStart = z; @@ -363,11 +363,11 @@ static int unicodeNext( ); /* Set the output variables and return. */ - pCsr->iOff = (z - pCsr->aInput); + pCsr->iOff = (int)(z - pCsr->aInput); *paToken = pCsr->zToken; - *pnToken = zOut - pCsr->zToken; - *piStart = (zStart - pCsr->aInput); - *piEnd = (zEnd - pCsr->aInput); + *pnToken = (int)(zOut - pCsr->zToken); + *piStart = (int)(zStart - pCsr->aInput); + *piEnd = (int)(zEnd - pCsr->aInput); *piPos = pCsr->iToken++; return SQLITE_OK; } @@ -390,4 +390,4 @@ void sqlite3Fts3UnicodeTokenizer(sqlite3_tokenizer_module const **ppModule){ } #endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS3) */ -#endif /* ifndef SQLITE_ENABLE_FTS4_UNICODE61 */ +#endif /* ifndef SQLITE_DISABLE_FTS3_UNICODE */ diff --git a/lib/libsqlite3/ext/fts3/fts3_unicode2.c b/lib/libsqlite3/ext/fts3/fts3_unicode2.c index fba688ff9ce..20b7a25dbfd 100644 --- a/lib/libsqlite3/ext/fts3/fts3_unicode2.c +++ b/lib/libsqlite3/ext/fts3/fts3_unicode2.c @@ -15,7 +15,7 @@ ** DO NOT EDIT THIS MACHINE GENERATED FILE. */ -#if defined(SQLITE_ENABLE_FTS4_UNICODE61) +#ifndef SQLITE_DISABLE_FTS3_UNICODE #if defined(SQLITE_ENABLE_FTS3) || defined(SQLITE_ENABLE_FTS4) #include <assert.h> @@ -39,7 +39,7 @@ int sqlite3FtsUnicodeIsalnum(int c){ ** C. It is not possible to represent a range larger than 1023 codepoints ** using this format. */ - const static unsigned int aEntry[] = { + static const unsigned int aEntry[] = { 0x00000030, 0x0000E807, 0x00016C06, 0x0001EC2F, 0x0002AC07, 0x0002D001, 0x0002D803, 0x0002EC01, 0x0002FC01, 0x00035C01, 0x0003DC01, 0x000B0804, 0x000B480E, 0x000B9407, 0x000BB401, @@ -131,7 +131,7 @@ int sqlite3FtsUnicodeIsalnum(int c){ return ( (aAscii[c >> 5] & (1 << (c & 0x001F)))==0 ); }else if( c<(1<<22) ){ unsigned int key = (((unsigned int)c)<<10) | 0x000003FF; - int iRes; + int iRes = 0; int iHi = sizeof(aEntry)/sizeof(aEntry[0]) - 1; int iLo = 0; while( iHi>=iLo ){ @@ -202,7 +202,7 @@ static int remove_diacritic(int c){ } assert( key>=aDia[iRes] ); return ((c > (aDia[iRes]>>3) + (aDia[iRes]&0x07)) ? c : (int)aChar[iRes]); -}; +} /* @@ -362,4 +362,4 @@ int sqlite3FtsUnicodeFold(int c, int bRemoveDiacritic){ return ret; } #endif /* defined(SQLITE_ENABLE_FTS3) || defined(SQLITE_ENABLE_FTS4) */ -#endif /* !defined(SQLITE_ENABLE_FTS4_UNICODE61) */ +#endif /* !defined(SQLITE_DISABLE_FTS3_UNICODE) */ diff --git a/lib/libsqlite3/ext/fts3/fts3_write.c b/lib/libsqlite3/ext/fts3/fts3_write.c index 3564cf8206d..0da08c62d8b 100644 --- a/lib/libsqlite3/ext/fts3/fts3_write.c +++ b/lib/libsqlite3/ext/fts3/fts3_write.c @@ -193,6 +193,7 @@ struct SegmentWriter { int nSize; /* Size of allocation at aData */ int nData; /* Bytes of data in aData */ char *aData; /* Pointer to block from malloc() */ + i64 nLeafData; /* Number of bytes of leaf data written */ }; /* @@ -268,6 +269,10 @@ struct SegmentNode { #define SQL_SELECT_INDEXES 35 #define SQL_SELECT_MXLEVEL 36 +#define SQL_SELECT_LEVEL_RANGE2 37 +#define SQL_UPDATE_LEVEL_IDX 38 +#define SQL_UPDATE_LEVEL 39 + /* ** This function is used to obtain an SQLite prepared statement handle ** for the statement identified by the second argument. If successful, @@ -369,7 +374,18 @@ static int fts3SqlStmt( /* SQL_SELECT_MXLEVEL ** Return the largest relative level in the FTS index or indexes. */ -/* 36 */ "SELECT max( level %% 1024 ) FROM %Q.'%q_segdir'" +/* 36 */ "SELECT max( level %% 1024 ) FROM %Q.'%q_segdir'", + + /* Return segments in order from oldest to newest.*/ +/* 37 */ "SELECT level, idx, end_block " + "FROM %Q.'%q_segdir' WHERE level BETWEEN ? AND ? " + "ORDER BY level DESC, idx ASC", + + /* Update statements used while promoting segments */ +/* 38 */ "UPDATE OR FAIL %Q.'%q_segdir' SET level=-1,idx=? " + "WHERE level=? AND idx=?", +/* 39 */ "UPDATE OR FAIL %Q.'%q_segdir' SET level=? WHERE level=-1" + }; int rc = SQLITE_OK; sqlite3_stmt *pStmt; @@ -1910,6 +1926,7 @@ static int fts3WriteSegdir( sqlite3_int64 iStartBlock, /* Value for "start_block" field */ sqlite3_int64 iLeafEndBlock, /* Value for "leaves_end_block" field */ sqlite3_int64 iEndBlock, /* Value for "end_block" field */ + sqlite3_int64 nLeafData, /* Bytes of leaf data in segment */ char *zRoot, /* Blob value for "root" field */ int nRoot /* Number of bytes in buffer zRoot */ ){ @@ -1920,7 +1937,13 @@ static int fts3WriteSegdir( sqlite3_bind_int(pStmt, 2, iIdx); sqlite3_bind_int64(pStmt, 3, iStartBlock); sqlite3_bind_int64(pStmt, 4, iLeafEndBlock); - sqlite3_bind_int64(pStmt, 5, iEndBlock); + if( nLeafData==0 ){ + sqlite3_bind_int64(pStmt, 5, iEndBlock); + }else{ + char *zEnd = sqlite3_mprintf("%lld %lld", iEndBlock, nLeafData); + if( !zEnd ) return SQLITE_NOMEM; + sqlite3_bind_text(pStmt, 5, zEnd, -1, sqlite3_free); + } sqlite3_bind_blob(pStmt, 6, zRoot, nRoot, SQLITE_STATIC); sqlite3_step(pStmt); rc = sqlite3_reset(pStmt); @@ -2246,6 +2269,9 @@ static int fts3SegWriterAdd( nDoclist; /* Doclist data */ } + /* Increase the total number of bytes written to account for the new entry. */ + pWriter->nLeafData += nReq; + /* If the buffer currently allocated is too small for this entry, realloc ** the buffer to make it large enough. */ @@ -2317,13 +2343,13 @@ static int fts3SegWriterFlush( pWriter->iFirst, pWriter->iFree, &iLast, &zRoot, &nRoot); } if( rc==SQLITE_OK ){ - rc = fts3WriteSegdir( - p, iLevel, iIdx, pWriter->iFirst, iLastLeaf, iLast, zRoot, nRoot); + rc = fts3WriteSegdir(p, iLevel, iIdx, + pWriter->iFirst, iLastLeaf, iLast, pWriter->nLeafData, zRoot, nRoot); } }else{ /* The entire tree fits on the root node. Write it to the segdir table. */ - rc = fts3WriteSegdir( - p, iLevel, iIdx, 0, 0, 0, pWriter->aData, pWriter->nData); + rc = fts3WriteSegdir(p, iLevel, iIdx, + 0, 0, 0, pWriter->nLeafData, pWriter->aData, pWriter->nData); } p->nLeafAdd++; return rc; @@ -2408,6 +2434,37 @@ static int fts3SegmentMaxLevel( } /* +** iAbsLevel is an absolute level that may be assumed to exist within +** the database. This function checks if it is the largest level number +** within its index. Assuming no error occurs, *pbMax is set to 1 if +** iAbsLevel is indeed the largest level, or 0 otherwise, and SQLITE_OK +** is returned. If an error occurs, an error code is returned and the +** final value of *pbMax is undefined. +*/ +static int fts3SegmentIsMaxLevel(Fts3Table *p, i64 iAbsLevel, int *pbMax){ + + /* Set pStmt to the compiled version of: + ** + ** SELECT max(level) FROM %Q.'%q_segdir' WHERE level BETWEEN ? AND ? + ** + ** (1024 is actually the value of macro FTS3_SEGDIR_PREFIXLEVEL_STR). + */ + sqlite3_stmt *pStmt; + int rc = fts3SqlStmt(p, SQL_SELECT_SEGDIR_MAX_LEVEL, &pStmt, 0); + if( rc!=SQLITE_OK ) return rc; + sqlite3_bind_int64(pStmt, 1, iAbsLevel+1); + sqlite3_bind_int64(pStmt, 2, + ((iAbsLevel/FTS3_SEGDIR_MAXLEVEL)+1) * FTS3_SEGDIR_MAXLEVEL + ); + + *pbMax = 0; + if( SQLITE_ROW==sqlite3_step(pStmt) ){ + *pbMax = sqlite3_column_type(pStmt, 0)==SQLITE_NULL; + } + return sqlite3_reset(pStmt); +} + +/* ** Delete all entries in the %_segments table associated with the segment ** opened with seg-reader pSeg. This function does not affect the contents ** of the %_segdir table. @@ -2943,6 +3000,140 @@ void sqlite3Fts3SegReaderFinish( } /* +** Decode the "end_block" field, selected by column iCol of the SELECT +** statement passed as the first argument. +** +** The "end_block" field may contain either an integer, or a text field +** containing the text representation of two non-negative integers separated +** by one or more space (0x20) characters. In the first case, set *piEndBlock +** to the integer value and *pnByte to zero before returning. In the second, +** set *piEndBlock to the first value and *pnByte to the second. +*/ +static void fts3ReadEndBlockField( + sqlite3_stmt *pStmt, + int iCol, + i64 *piEndBlock, + i64 *pnByte +){ + const unsigned char *zText = sqlite3_column_text(pStmt, iCol); + if( zText ){ + int i; + int iMul = 1; + i64 iVal = 0; + for(i=0; zText[i]>='0' && zText[i]<='9'; i++){ + iVal = iVal*10 + (zText[i] - '0'); + } + *piEndBlock = iVal; + while( zText[i]==' ' ) i++; + iVal = 0; + if( zText[i]=='-' ){ + i++; + iMul = -1; + } + for(/* no-op */; zText[i]>='0' && zText[i]<='9'; i++){ + iVal = iVal*10 + (zText[i] - '0'); + } + *pnByte = (iVal * (i64)iMul); + } +} + + +/* +** A segment of size nByte bytes has just been written to absolute level +** iAbsLevel. Promote any segments that should be promoted as a result. +*/ +static int fts3PromoteSegments( + Fts3Table *p, /* FTS table handle */ + sqlite3_int64 iAbsLevel, /* Absolute level just updated */ + sqlite3_int64 nByte /* Size of new segment at iAbsLevel */ +){ + int rc = SQLITE_OK; + sqlite3_stmt *pRange; + + rc = fts3SqlStmt(p, SQL_SELECT_LEVEL_RANGE2, &pRange, 0); + + if( rc==SQLITE_OK ){ + int bOk = 0; + i64 iLast = (iAbsLevel/FTS3_SEGDIR_MAXLEVEL + 1) * FTS3_SEGDIR_MAXLEVEL - 1; + i64 nLimit = (nByte*3)/2; + + /* Loop through all entries in the %_segdir table corresponding to + ** segments in this index on levels greater than iAbsLevel. If there is + ** at least one such segment, and it is possible to determine that all + ** such segments are smaller than nLimit bytes in size, they will be + ** promoted to level iAbsLevel. */ + sqlite3_bind_int64(pRange, 1, iAbsLevel+1); + sqlite3_bind_int64(pRange, 2, iLast); + while( SQLITE_ROW==sqlite3_step(pRange) ){ + i64 nSize = 0, dummy; + fts3ReadEndBlockField(pRange, 2, &dummy, &nSize); + if( nSize<=0 || nSize>nLimit ){ + /* If nSize==0, then the %_segdir.end_block field does not not + ** contain a size value. This happens if it was written by an + ** old version of FTS. In this case it is not possible to determine + ** the size of the segment, and so segment promotion does not + ** take place. */ + bOk = 0; + break; + } + bOk = 1; + } + rc = sqlite3_reset(pRange); + + if( bOk ){ + int iIdx = 0; + sqlite3_stmt *pUpdate1; + sqlite3_stmt *pUpdate2; + + if( rc==SQLITE_OK ){ + rc = fts3SqlStmt(p, SQL_UPDATE_LEVEL_IDX, &pUpdate1, 0); + } + if( rc==SQLITE_OK ){ + rc = fts3SqlStmt(p, SQL_UPDATE_LEVEL, &pUpdate2, 0); + } + + if( rc==SQLITE_OK ){ + + /* Loop through all %_segdir entries for segments in this index with + ** levels equal to or greater than iAbsLevel. As each entry is visited, + ** updated it to set (level = -1) and (idx = N), where N is 0 for the + ** oldest segment in the range, 1 for the next oldest, and so on. + ** + ** In other words, move all segments being promoted to level -1, + ** setting the "idx" fields as appropriate to keep them in the same + ** order. The contents of level -1 (which is never used, except + ** transiently here), will be moved back to level iAbsLevel below. */ + sqlite3_bind_int64(pRange, 1, iAbsLevel); + while( SQLITE_ROW==sqlite3_step(pRange) ){ + sqlite3_bind_int(pUpdate1, 1, iIdx++); + sqlite3_bind_int(pUpdate1, 2, sqlite3_column_int(pRange, 0)); + sqlite3_bind_int(pUpdate1, 3, sqlite3_column_int(pRange, 1)); + sqlite3_step(pUpdate1); + rc = sqlite3_reset(pUpdate1); + if( rc!=SQLITE_OK ){ + sqlite3_reset(pRange); + break; + } + } + } + if( rc==SQLITE_OK ){ + rc = sqlite3_reset(pRange); + } + + /* Move level -1 to level iAbsLevel */ + if( rc==SQLITE_OK ){ + sqlite3_bind_int64(pUpdate2, 1, iAbsLevel); + sqlite3_step(pUpdate2); + rc = sqlite3_reset(pUpdate2); + } + } + } + + + return rc; +} + +/* ** Merge all level iLevel segments in the database into a single ** iLevel+1 segment. Or, if iLevel<0, merge all segments into a ** single segment with a level equal to the numerically largest level @@ -2966,6 +3157,7 @@ static int fts3SegmentMerge( Fts3SegFilter filter; /* Segment term filter condition */ Fts3MultiSegReader csr; /* Cursor to iterate through level(s) */ int bIgnoreEmpty = 0; /* True to ignore empty segments */ + i64 iMaxLevel = 0; /* Max level number for this index/langid */ assert( iLevel==FTS3_SEGCURSOR_ALL || iLevel==FTS3_SEGCURSOR_PENDING @@ -2977,6 +3169,11 @@ static int fts3SegmentMerge( rc = sqlite3Fts3SegReaderCursor(p, iLangid, iIndex, iLevel, 0, 0, 1, 0, &csr); if( rc!=SQLITE_OK || csr.nSegment==0 ) goto finished; + if( iLevel!=FTS3_SEGCURSOR_PENDING ){ + rc = fts3SegmentMaxLevel(p, iLangid, iIndex, &iMaxLevel); + if( rc!=SQLITE_OK ) goto finished; + } + if( iLevel==FTS3_SEGCURSOR_ALL ){ /* This call is to merge all segments in the database to a single ** segment. The level of the new segment is equal to the numerically @@ -2986,21 +3183,21 @@ static int fts3SegmentMerge( rc = SQLITE_DONE; goto finished; } - rc = fts3SegmentMaxLevel(p, iLangid, iIndex, &iNewLevel); + iNewLevel = iMaxLevel; bIgnoreEmpty = 1; - }else if( iLevel==FTS3_SEGCURSOR_PENDING ){ - iNewLevel = getAbsoluteLevel(p, iLangid, iIndex, 0); - rc = fts3AllocateSegdirIdx(p, iLangid, iIndex, 0, &iIdx); }else{ /* This call is to merge all segments at level iLevel. find the next ** available segment index at level iLevel+1. The call to ** fts3AllocateSegdirIdx() will merge the segments at level iLevel+1 to ** a single iLevel+2 segment if necessary. */ - rc = fts3AllocateSegdirIdx(p, iLangid, iIndex, iLevel+1, &iIdx); + assert( FTS3_SEGCURSOR_PENDING==-1 ); iNewLevel = getAbsoluteLevel(p, iLangid, iIndex, iLevel+1); + rc = fts3AllocateSegdirIdx(p, iLangid, iIndex, iLevel+1, &iIdx); + bIgnoreEmpty = (iLevel!=FTS3_SEGCURSOR_PENDING) && (iNewLevel>iMaxLevel); } if( rc!=SQLITE_OK ) goto finished; + assert( csr.nSegment>0 ); assert( iNewLevel>=getAbsoluteLevel(p, iLangid, iIndex, 0) ); assert( iNewLevel<getAbsoluteLevel(p, iLangid, iIndex,FTS3_SEGDIR_MAXLEVEL) ); @@ -3017,7 +3214,7 @@ static int fts3SegmentMerge( csr.zTerm, csr.nTerm, csr.aDoclist, csr.nDoclist); } if( rc!=SQLITE_OK ) goto finished; - assert( pWriter ); + assert( pWriter || bIgnoreEmpty ); if( iLevel!=FTS3_SEGCURSOR_PENDING ){ rc = fts3DeleteSegdir( @@ -3025,7 +3222,14 @@ static int fts3SegmentMerge( ); if( rc!=SQLITE_OK ) goto finished; } - rc = fts3SegWriterFlush(p, pWriter, iNewLevel, iIdx); + if( pWriter ){ + rc = fts3SegWriterFlush(p, pWriter, iNewLevel, iIdx); + if( rc==SQLITE_OK ){ + if( iLevel==FTS3_SEGCURSOR_PENDING || iNewLevel<iMaxLevel ){ + rc = fts3PromoteSegments(p, iNewLevel, pWriter->nLeafData); + } + } + } finished: fts3SegWriterFree(pWriter); @@ -3035,7 +3239,7 @@ static int fts3SegmentMerge( /* -** Flush the contents of pendingTerms to level 0 segments. +** Flush the contents of pendingTerms to level 0 segments. */ int sqlite3Fts3PendingTermsFlush(Fts3Table *p){ int rc = SQLITE_OK; @@ -3051,14 +3255,19 @@ int sqlite3Fts3PendingTermsFlush(Fts3Table *p){ ** estimate the number of leaf blocks of content to be written */ if( rc==SQLITE_OK && p->bHasStat - && p->bAutoincrmerge==0xff && p->nLeafAdd>0 + && p->nAutoincrmerge==0xff && p->nLeafAdd>0 ){ sqlite3_stmt *pStmt = 0; rc = fts3SqlStmt(p, SQL_SELECT_STAT, &pStmt, 0); if( rc==SQLITE_OK ){ sqlite3_bind_int(pStmt, 1, FTS_STAT_AUTOINCRMERGE); rc = sqlite3_step(pStmt); - p->bAutoincrmerge = (rc==SQLITE_ROW && sqlite3_column_int(pStmt, 0)); + if( rc==SQLITE_ROW ){ + p->nAutoincrmerge = sqlite3_column_int(pStmt, 0); + if( p->nAutoincrmerge==1 ) p->nAutoincrmerge = 8; + }else if( rc==SQLITE_DONE ){ + p->nAutoincrmerge = 0; + } rc = sqlite3_reset(pStmt); } } @@ -3426,6 +3635,8 @@ struct IncrmergeWriter { int iIdx; /* Index of *output* segment in iAbsLevel+1 */ sqlite3_int64 iStart; /* Block number of first allocated block */ sqlite3_int64 iEnd; /* Block number of last allocated block */ + sqlite3_int64 nLeafData; /* Bytes of leaf page data so far */ + u8 bNoLeafData; /* If true, store 0 for segment size */ NodeWriter aNodeWriter[FTS_MAX_APPENDABLE_HEIGHT]; }; @@ -3764,8 +3975,8 @@ static int fts3IncrmergeAppend( nSpace += sqlite3Fts3VarintLen(nDoclist) + nDoclist; } + pWriter->nLeafData += nSpace; blobGrowBuffer(&pLeaf->block, pLeaf->block.n + nSpace, &rc); - if( rc==SQLITE_OK ){ if( pLeaf->block.n==0 ){ pLeaf->block.n = 1; @@ -3864,6 +4075,7 @@ static void fts3IncrmergeRelease( pWriter->iStart, /* start_block */ pWriter->aNodeWriter[0].iBlock, /* leaves_end_block */ pWriter->iEnd, /* end_block */ + (pWriter->bNoLeafData==0 ? pWriter->nLeafData : 0), /* end_block */ pRoot->block.a, pRoot->block.n /* root */ ); } @@ -3965,7 +4177,11 @@ static int fts3IncrmergeLoad( if( sqlite3_step(pSelect)==SQLITE_ROW ){ iStart = sqlite3_column_int64(pSelect, 1); iLeafEnd = sqlite3_column_int64(pSelect, 2); - iEnd = sqlite3_column_int64(pSelect, 3); + fts3ReadEndBlockField(pSelect, 3, &iEnd, &pWriter->nLeafData); + if( pWriter->nLeafData<0 ){ + pWriter->nLeafData = pWriter->nLeafData * -1; + } + pWriter->bNoLeafData = (pWriter->nLeafData==0); nRoot = sqlite3_column_bytes(pSelect, 4); aRoot = sqlite3_column_blob(pSelect, 4); }else{ @@ -4566,11 +4782,11 @@ static int fts3IncrmergeHintPop(Blob *pHint, i64 *piAbsLevel, int *pnInput){ /* ** Attempt an incremental merge that writes nMerge leaf blocks. ** -** Incremental merges happen nMin segments at a time. The two -** segments to be merged are the nMin oldest segments (the ones with -** the smallest indexes) in the highest level that contains at least -** nMin segments. Multiple merges might occur in an attempt to write the -** quota of nMerge leaf blocks. +** Incremental merges happen nMin segments at a time. The segments +** to be merged are the nMin oldest segments (the ones with the smallest +** values for the _segdir.idx field) in the highest level that contains +** at least nMin segments. Multiple merges might occur in an attempt to +** write the quota of nMerge leaf blocks. */ int sqlite3Fts3Incrmerge(Fts3Table *p, int nMerge, int nMin){ int rc; /* Return code */ @@ -4595,6 +4811,7 @@ int sqlite3Fts3Incrmerge(Fts3Table *p, int nMerge, int nMin){ const i64 nMod = FTS3_SEGDIR_MAXLEVEL * p->nIndex; sqlite3_stmt *pFindLevel = 0; /* SQL used to determine iAbsLevel */ int bUseHint = 0; /* True if attempting to append */ + int iIdx = 0; /* Largest idx in level (iAbsLevel+1) */ /* Search the %_segdir table for the absolute level with the smallest ** relative level number that contains at least nMin segments, if any. @@ -4648,6 +4865,19 @@ int sqlite3Fts3Incrmerge(Fts3Table *p, int nMerge, int nMin){ ** to start work on some other level. */ memset(pWriter, 0, nAlloc); pFilter->flags = FTS3_SEGMENT_REQUIRE_POS; + + if( rc==SQLITE_OK ){ + rc = fts3IncrmergeOutputIdx(p, iAbsLevel, &iIdx); + assert( bUseHint==1 || bUseHint==0 ); + if( iIdx==0 || (bUseHint && iIdx==1) ){ + int bIgnore = 0; + rc = fts3SegmentIsMaxLevel(p, iAbsLevel+1, &bIgnore); + if( bIgnore ){ + pFilter->flags |= FTS3_SEGMENT_IGNORE_EMPTY; + } + } + } + if( rc==SQLITE_OK ){ rc = fts3IncrmergeCsr(p, iAbsLevel, nSeg, pCsr); } @@ -4655,16 +4885,12 @@ int sqlite3Fts3Incrmerge(Fts3Table *p, int nMerge, int nMin){ && SQLITE_OK==(rc = sqlite3Fts3SegReaderStart(p, pCsr, pFilter)) && SQLITE_ROW==(rc = sqlite3Fts3SegReaderStep(p, pCsr)) ){ - int iIdx = 0; /* Largest idx in level (iAbsLevel+1) */ - rc = fts3IncrmergeOutputIdx(p, iAbsLevel, &iIdx); - if( rc==SQLITE_OK ){ - if( bUseHint && iIdx>0 ){ - const char *zKey = pCsr->zTerm; - int nKey = pCsr->nTerm; - rc = fts3IncrmergeLoad(p, iAbsLevel, iIdx-1, zKey, nKey, pWriter); - }else{ - rc = fts3IncrmergeWriter(p, iAbsLevel, iIdx, pCsr, pWriter); - } + if( bUseHint && iIdx>0 ){ + const char *zKey = pCsr->zTerm; + int nKey = pCsr->nTerm; + rc = fts3IncrmergeLoad(p, iAbsLevel, iIdx-1, zKey, nKey, pWriter); + }else{ + rc = fts3IncrmergeWriter(p, iAbsLevel, iIdx, pCsr, pWriter); } if( rc==SQLITE_OK && pWriter->nLeafEst ){ @@ -4686,7 +4912,13 @@ int sqlite3Fts3Incrmerge(Fts3Table *p, int nMerge, int nMin){ } } + if( nSeg!=0 ){ + pWriter->nLeafData = pWriter->nLeafData * -1; + } fts3IncrmergeRelease(p, pWriter, &rc); + if( nSeg==0 && pWriter->bNoLeafData==0 ){ + fts3PromoteSegments(p, iAbsLevel+1, pWriter->nLeafData); + } } sqlite3Fts3SegReaderFinish(pCsr); @@ -4773,7 +5005,10 @@ static int fts3DoAutoincrmerge( ){ int rc = SQLITE_OK; sqlite3_stmt *pStmt = 0; - p->bAutoincrmerge = fts3Getint(&zParam)!=0; + p->nAutoincrmerge = fts3Getint(&zParam); + if( p->nAutoincrmerge==1 || p->nAutoincrmerge>FTS3_MERGE_COUNT ){ + p->nAutoincrmerge = 8; + } if( !p->bHasStat ){ assert( p->bFts4==0 ); sqlite3Fts3CreateStatTable(&rc, p); @@ -4782,7 +5017,7 @@ static int fts3DoAutoincrmerge( rc = fts3SqlStmt(p, SQL_REPLACE_STAT, &pStmt, 0); if( rc ) return rc; sqlite3_bind_int(pStmt, 1, FTS_STAT_AUTOINCRMERGE); - sqlite3_bind_int(pStmt, 2, p->bAutoincrmerge); + sqlite3_bind_int(pStmt, 2, p->nAutoincrmerge); sqlite3_step(pStmt); rc = sqlite3_reset(pStmt); return rc; @@ -4939,34 +5174,36 @@ static int fts3IntegrityCheck(Fts3Table *p, int *pbOk){ int iCol; for(iCol=0; rc==SQLITE_OK && iCol<p->nColumn; iCol++){ - const char *zText = (const char *)sqlite3_column_text(pStmt, iCol+1); - int nText = sqlite3_column_bytes(pStmt, iCol+1); - sqlite3_tokenizer_cursor *pT = 0; - - rc = sqlite3Fts3OpenTokenizer(p->pTokenizer, iLang, zText, nText, &pT); - while( rc==SQLITE_OK ){ - char const *zToken; /* Buffer containing token */ - int nToken = 0; /* Number of bytes in token */ - int iDum1 = 0, iDum2 = 0; /* Dummy variables */ - int iPos = 0; /* Position of token in zText */ - - rc = pModule->xNext(pT, &zToken, &nToken, &iDum1, &iDum2, &iPos); - if( rc==SQLITE_OK ){ - int i; - cksum2 = cksum2 ^ fts3ChecksumEntry( - zToken, nToken, iLang, 0, iDocid, iCol, iPos - ); - for(i=1; i<p->nIndex; i++){ - if( p->aIndex[i].nPrefix<=nToken ){ - cksum2 = cksum2 ^ fts3ChecksumEntry( - zToken, p->aIndex[i].nPrefix, iLang, i, iDocid, iCol, iPos - ); + if( p->abNotindexed[iCol]==0 ){ + const char *zText = (const char *)sqlite3_column_text(pStmt, iCol+1); + int nText = sqlite3_column_bytes(pStmt, iCol+1); + sqlite3_tokenizer_cursor *pT = 0; + + rc = sqlite3Fts3OpenTokenizer(p->pTokenizer, iLang, zText, nText,&pT); + while( rc==SQLITE_OK ){ + char const *zToken; /* Buffer containing token */ + int nToken = 0; /* Number of bytes in token */ + int iDum1 = 0, iDum2 = 0; /* Dummy variables */ + int iPos = 0; /* Position of token in zText */ + + rc = pModule->xNext(pT, &zToken, &nToken, &iDum1, &iDum2, &iPos); + if( rc==SQLITE_OK ){ + int i; + cksum2 = cksum2 ^ fts3ChecksumEntry( + zToken, nToken, iLang, 0, iDocid, iCol, iPos + ); + for(i=1; i<p->nIndex; i++){ + if( p->aIndex[i].nPrefix<=nToken ){ + cksum2 = cksum2 ^ fts3ChecksumEntry( + zToken, p->aIndex[i].nPrefix, iLang, i, iDocid, iCol, iPos + ); + } } } } + if( pT ) pModule->xClose(pT); + if( rc==SQLITE_DONE ) rc = SQLITE_OK; } - if( pT ) pModule->xClose(pT); - if( rc==SQLITE_DONE ) rc = SQLITE_OK; } } @@ -5271,6 +5508,10 @@ int sqlite3Fts3UpdateMethod( int nChng = 0; /* Net change in number of documents */ int bInsertDone = 0; + /* At this point it must be known if the %_stat table exists or not. + ** So bHasStat may not be 2. */ + assert( p->bHasStat==0 || p->bHasStat==1 ); + assert( p->pSegments==0 ); assert( nArg==1 /* DELETE operations */ diff --git a/lib/libsqlite3/ext/fts3/tool/fts3view.c b/lib/libsqlite3/ext/fts3/tool/fts3view.c index 479ae9868d4..3dc1ba80fee 100644 --- a/lib/libsqlite3/ext/fts3/tool/fts3view.c +++ b/lib/libsqlite3/ext/fts3/tool/fts3view.c @@ -376,7 +376,7 @@ static void showSegmentStats(sqlite3 *db, const char *zTab){ sqlite3_finalize(pStmt); nLeaf = nSeg - nIdx; printf("Leaf segments larger than %5d bytes.... %9d %5.2f%%\n", - pgsz-45, n, n*100.0/nLeaf); + pgsz-45, n, nLeaf>0 ? n*100.0/nLeaf : 0.0); pStmt = prepare(db, "SELECT max(level%%1024) FROM '%q_segdir'", zTab); mxLevel = 0; @@ -554,7 +554,7 @@ static void decodeSegment( sqlite3_int64 n; sqlite3_int64 iDocsz; int iHeight; - int i = 0; + sqlite3_int64 i = 0; int cnt = 0; char zTerm[1000]; @@ -576,12 +576,12 @@ static void decodeSegment( fprintf(stderr, "term to long\n"); exit(1); } - memcpy(zTerm+iPrefix, aData+i, nTerm); + memcpy(zTerm+iPrefix, aData+i, (size_t)nTerm); zTerm[iPrefix+nTerm] = 0; i += nTerm; if( iHeight==0 ){ i += getVarint(aData+i, &iDocsz); - printf("term: %-25s doclist %7lld bytes offset %d\n", zTerm, iDocsz, i); + printf("term: %-25s doclist %7lld bytes offset %lld\n", zTerm, iDocsz, i); i += iDocsz; }else{ printf("term: %-25s child %lld\n", zTerm, ++iChild); @@ -749,18 +749,19 @@ static void decodeDoclist( */ static void showDoclist(sqlite3 *db, const char *zTab){ const unsigned char *aData; - sqlite3_int64 offset, nData; + sqlite3_int64 offset; + int nData; sqlite3_stmt *pStmt; offset = atoi64(azExtra[1]); - nData = atoi64(azExtra[2]); + nData = atoi(azExtra[2]); pStmt = prepareToGetSegment(db, zTab, azExtra[0]); if( sqlite3_step(pStmt)!=SQLITE_ROW ){ sqlite3_finalize(pStmt); return; } aData = sqlite3_column_blob(pStmt, 0); - printf("Doclist at %s offset %lld of size %lld bytes:\n", + printf("Doclist at %s offset %lld of size %d bytes:\n", azExtra[0], offset, nData); if( findOption("raw", 0, 0)!=0 ){ printBlob(aData+offset, nData); diff --git a/lib/libsqlite3/ext/fts3/unicode/mkunicode.tcl b/lib/libsqlite3/ext/fts3/unicode/mkunicode.tcl index 2da17c51a53..c3083ee3686 100644 --- a/lib/libsqlite3/ext/fts3/unicode/mkunicode.tcl +++ b/lib/libsqlite3/ext/fts3/unicode/mkunicode.tcl @@ -160,7 +160,7 @@ proc print_rd {map} { } assert( key>=aDia[iRes] ); return ((c > (aDia[iRes]>>3) + (aDia[iRes]&0x07)) ? c : (int)aChar[iRes]);} - puts "\};" + puts "\}" } proc print_isdiacritic {zFunc map} { @@ -298,7 +298,7 @@ proc an_print_range_array {lRange} { ** using this format. */ }] - puts -nonewline " const static unsigned int aEntry\[\] = \{" + puts -nonewline " static const unsigned int aEntry\[\] = \{" set i 0 foreach range $lRange { foreach {iFirst nRange} $range {} @@ -349,7 +349,7 @@ proc print_isalnum {zFunc lRange} { return ( (aAscii[c >> 5] & (1 << (c & 0x001F)))==0 ); }else if( c<(1<<22) ){ unsigned int key = (((unsigned int)c)<<10) | 0x000003FF; - int iRes; + int iRes = 0; int iHi = sizeof(aEntry)/sizeof(aEntry[0]) - 1; int iLo = 0; while( iHi>=iLo ){ @@ -732,7 +732,7 @@ proc print_fileheader {} { */ }] puts "" - puts "#if defined(SQLITE_ENABLE_FTS4_UNICODE61)" + puts "#ifndef SQLITE_DISABLE_FTS3_UNICODE" puts "#if defined(SQLITE_ENABLE_FTS3) || defined(SQLITE_ENABLE_FTS4)" puts "" puts "#include <assert.h>" @@ -808,4 +808,4 @@ if {$::generate_test_code} { } puts "#endif /* defined(SQLITE_ENABLE_FTS3) || defined(SQLITE_ENABLE_FTS4) */" -puts "#endif /* !defined(SQLITE_ENABLE_FTS4_UNICODE61) */" +puts "#endif /* !defined(SQLITE_DISABLE_FTS3_UNICODE) */" |