Fix some problems in fts3 found by address-sanitizer.

FossilOrigin-Name: 16a8e84fa7f67a467f824bdd7f72cbd6a6e95dab8cc7aa1e0e751720b98f3e31

ext/fts3/fts3.c

@@ -349,8 +349,9 @@ int sqlite3Fts3PutVarint(char *p, sqlite_int64 v){
 ** Return the number of bytes read, or 0 on error.
 ** The value is stored in *v.
 */
-int sqlite3Fts3GetVarint(const char *p, sqlite_int64 *v){
-  const char *pStart = p;
+int sqlite3Fts3GetVarint(const char *pBuf, sqlite_int64 *v){
+  const unsigned char *p = (const unsigned char*)pBuf;
+  const unsigned char *pStart = p;
   u32 a;
   u64 b;
   int shift;

ext/fts3/fts3_unicode.c

@@ -136,16 +136,16 @@ static int unicodeAddExceptions(
 ){
   const unsigned char *z = (const unsigned char *)zIn;
   const unsigned char *zTerm = &z[nIn];
-  int iCode;
+  unsigned int iCode;
   int nEntry = 0;
 
   assert( bAlnum==0 || bAlnum==1 );
 
   while( z<zTerm ){
     READ_UTF8(z, zTerm, iCode);
-    assert( (sqlite3FtsUnicodeIsalnum(iCode) & 0xFFFFFFFE)==0 );
-    if( sqlite3FtsUnicodeIsalnum(iCode)!=bAlnum 
-     && sqlite3FtsUnicodeIsdiacritic(iCode)==0 
+    assert( (sqlite3FtsUnicodeIsalnum((int)iCode) & 0xFFFFFFFE)==0 );
+    if( sqlite3FtsUnicodeIsalnum((int)iCode)!=bAlnum 
+     && sqlite3FtsUnicodeIsdiacritic((int)iCode)==0 
     ){
       nEntry++;
     }
@@ -162,13 +162,13 @@ static int unicodeAddExceptions(
     z = (const unsigned char *)zIn;
     while( z<zTerm ){
       READ_UTF8(z, zTerm, iCode);
-      if( sqlite3FtsUnicodeIsalnum(iCode)!=bAlnum 
-       && sqlite3FtsUnicodeIsdiacritic(iCode)==0
+      if( sqlite3FtsUnicodeIsalnum((int)iCode)!=bAlnum 
+       && sqlite3FtsUnicodeIsdiacritic((int)iCode)==0
       ){
         int i, j;
-        for(i=0; i<nNew && aNew[i]<iCode; i++);
+        for(i=0; i<nNew && aNew[i]<(int)iCode; i++);
         for(j=nNew; j>i; j--) aNew[j] = aNew[j-1];
-        aNew[i] = iCode;
+        aNew[i] = (int)iCode;
         nNew++;
       }
     }
@@ -318,7 +318,7 @@ static int unicodeNext(
 ){
   unicode_cursor *pCsr = (unicode_cursor *)pC;
   unicode_tokenizer *p = ((unicode_tokenizer *)pCsr->base.pTokenizer);
-  int iCode = 0;
+  unsigned int iCode = 0;
   char *zOut;
   const unsigned char *z = &pCsr->aInput[pCsr->iOff];
   const unsigned char *zStart = z;
@@ -330,7 +330,7 @@ static int unicodeNext(
   ** the input.  */
   while( z<zTerm ){
     READ_UTF8(z, zTerm, iCode);
-    if( unicodeIsAlnum(p, iCode) ) break;
+    if( unicodeIsAlnum(p, (int)iCode) ) break;
     zStart = z;
   }
   if( zStart>=zTerm ) return SQLITE_DONE;
@@ -350,7 +350,7 @@ static int unicodeNext(
 
     /* Write the folded case of the last character read to the output */
     zEnd = z;
-    iOut = sqlite3FtsUnicodeFold(iCode, p->bRemoveDiacritic);
+    iOut = sqlite3FtsUnicodeFold((int)iCode, p->bRemoveDiacritic);
     if( iOut ){
       WRITE_UTF8(zOut, iOut);
     }
@@ -358,8 +358,8 @@ static int unicodeNext(
     /* If the cursor is not at EOF, read the next character */
     if( z>=zTerm ) break;
     READ_UTF8(z, zTerm, iCode);
-  }while( unicodeIsAlnum(p, iCode) 
-       || sqlite3FtsUnicodeIsdiacritic(iCode)
+  }while( unicodeIsAlnum(p, (int)iCode) 
+       || sqlite3FtsUnicodeIsdiacritic((int)iCode)
   );
 
   /* Set the output variables and return. */

ext/fts3/fts3_unicode2.c

@@ -127,9 +127,9 @@ int sqlite3FtsUnicodeIsalnum(int c){
     0xFFFFFFFF, 0xFC00FFFF, 0xF8000001, 0xF8000001,
   };
 
-  if( c<128 ){
-    return ( (aAscii[c >> 5] & (1 << (c & 0x001F)))==0 );
-  }else if( c<(1<<22) ){
+  if( (unsigned int)c<128 ){
+    return ( (aAscii[c >> 5] & ((unsigned int)1 << (c & 0x001F)))==0 );
+  }else if( (unsigned int)c<(1<<22) ){
     unsigned int key = (((unsigned int)c)<<10) | 0x000003FF;
     int iRes = 0;
     int iHi = sizeof(aEntry)/sizeof(aEntry[0]) - 1;
@@ -322,16 +322,17 @@ int sqlite3FtsUnicodeFold(int c, int bRemoveDiacritic){
 
   int ret = c;
 
-  assert( c>=0 );
   assert( sizeof(unsigned short)==2 && sizeof(unsigned char)==1 );
 
   if( c<128 ){
     if( c>='A' && c<='Z' ) ret = c + ('a' - 'A');
   }else if( c<65536 ){
+    const struct TableEntry *p;
     int iHi = sizeof(aEntry)/sizeof(aEntry[0]) - 1;
     int iLo = 0;
     int iRes = -1;
 
+    assert( c>aEntry[0].iCode );
     while( iHi>=iLo ){
       int iTest = (iHi + iLo) / 2;
       int cmp = (c - aEntry[iTest].iCode);
@@ -342,15 +343,13 @@ int sqlite3FtsUnicodeFold(int c, int bRemoveDiacritic){
         iHi = iTest-1;
       }
     }
-    assert( iRes<0 || c>=aEntry[iRes].iCode );
 
-    if( iRes>=0 ){
-      const struct TableEntry *p = &aEntry[iRes];
+    assert( iRes>=0 && c>=aEntry[iRes].iCode );
+    p = &aEntry[iRes];
     if( c<(p->iCode + p->nRange) && 0==(0x01 & p->flags & (p->iCode ^ c)) ){
       ret = (c + (aiOff[p->flags>>1])) & 0x0000FFFF;
       assert( ret>0 );
     }
-    }
 
     if( bRemoveDiacritic ) ret = remove_diacritic(ret);
   }

ext/fts3/fts3_write.c

@@ -4956,11 +4956,14 @@ int sqlite3Fts3Incrmerge(Fts3Table *p, int nMerge, int nMin){
 ** Convert the text beginning at *pz into an integer and return
 ** its value.  Advance *pz to point to the first character past
 ** the integer.
+**
+** This function used for parameters to merge= and incrmerge=
+** commands. 
 */
 static int fts3Getint(const char **pz){
   const char *z = *pz;
   int i = 0;
-  while( (*z)>='0' && (*z)<='9' ) i = 10*i + *(z++) - '0';
+  while( (*z)>='0' && (*z)<='9' && i<214748363 ) i = 10*i + *(z++) - '0';
   *pz = z;
   return i;
 }

ext/fts3/unicode/mkunicode.tcl

@@ -227,7 +227,7 @@ proc print_isalnum {zFunc lRange} {
   an_print_ascii_bitmap $lRange
   puts {
   if( (unsigned int)c<128 ){
-    return ( (aAscii[c >> 5] & (1 << (c & 0x001F)))==0 );
+    return ( (aAscii[c >> 5] & ((unsigned int)1 << (c & 0x001F)))==0 );
   }else if( (unsigned int)c<(1<<22) ){
     unsigned int key = (((unsigned int)c)<<10) | 0x000003FF;
     int iRes = 0;

manifest

@@ -1,5 +1,5 @@
-C Avoid\sthe\spossibility\sof\ssigned\sinteger\soverflow\swith\soversized\sprecisions\nin\s%d\sconversions\sin\sthe\sprintf()\simplementation.
-D 2017-03-20T16:34:18.983
+C Fix\ssome\sproblems\sin\sfts3\sfound\sby\saddress-sanitizer.
+D 2017-03-20T18:53:32.346
 F Makefile.in 1cc758ce3374a32425e4d130c2fe7b026b20de5b8843243de75f087c0a2661fb
 F Makefile.linux-gcc 7bc79876b875010e8c8f9502eb935ca92aa3c434
 F Makefile.msc 1faf9f06aadc9284c212dea7bbc7c0dea7e8337f0287c81001eff500912c790a
@@ -70,7 +70,7 @@ F ext/fts3/README.content fdc666a70d5257a64fee209f97cf89e0e6e32b51
 F ext/fts3/README.syntax a19711dc5458c20734b8e485e75fb1981ec2427a
 F ext/fts3/README.tokenizers e0a8b81383ea60d0334d274fadf305ea14a8c314
 F ext/fts3/README.txt 8c18f41574404623b76917b9da66fcb0ab38328d
-F ext/fts3/fts3.c 95c7041ea75d82d2d9a4cd058904ba889751f5b8
+F ext/fts3/fts3.c df964dbcc7bb97b8871c53866eb6c4e9c05c0d8af42038fa6439fb411260dd2e
 F ext/fts3/fts3.h 3a10a0af180d502cecc50df77b1b22df142817fe
 F ext/fts3/fts3Int.h eb2502000148e80913b965db3e59f29251266d0a
 F ext/fts3/fts3_aux.c 9edc3655fcb287f0467d0a4b886a01c6185fe9f1
@@ -86,15 +86,15 @@ F ext/fts3/fts3_tokenize_vtab.c a27593ab19657166f6fa5ec073b678cc29a75860
 F ext/fts3/fts3_tokenizer.c a22bf311a71f3efa9d7012d8cc48fc9b0f3dace7
 F ext/fts3/fts3_tokenizer.h 64c6ef6c5272c51ebe60fc607a896e84288fcbc3
 F ext/fts3/fts3_tokenizer1.c 5c98225a53705e5ee34824087478cf477bdb7004
-F ext/fts3/fts3_unicode.c a93f5edc0aff44ef8b06d7cb55b52026541ca145
-F ext/fts3/fts3_unicode2.c c3d01968d497bd7001e7dc774ba75b372738c057
-F ext/fts3/fts3_write.c c3863f23b6b4623c8b9d5cf31c12ce4469f78ca9
+F ext/fts3/fts3_unicode.c 525a3bd9a7564603c5c061b7de55403a565307758a94600e8a2f6b00d1c40d9d
+F ext/fts3/fts3_unicode2.c cc04fc672bfd42b1e650398cb0bf71f64f9aae032cfe75bbcfe75b9cf966029c
+F ext/fts3/fts3_write.c a51d48d646974ee2fb4b17fcd5da0416a5759a32dcacc2cce2ba00d5a767848e
 F ext/fts3/fts3speed.tcl b54caf6a18d38174f1a6e84219950d85e98bb1e9
 F ext/fts3/mkfts3amal.tcl 252ecb7fe6467854f2aa237bf2c390b74e71f100
 F ext/fts3/tool/fts3view.c 202801a2056995b763864d60c2dee744d46f1677
 F ext/fts3/unicode/CaseFolding.txt 8c678ca52ecc95e16bc7afc2dbf6fc9ffa05db8c
 F ext/fts3/unicode/UnicodeData.txt cd07314edb62d49fde34debdaf92fa2aa69011e7
-F ext/fts3/unicode/mkunicode.tcl 2debed3f582d77b3fdd0b8830880250021571fd8
+F ext/fts3/unicode/mkunicode.tcl ab0543a3b2399092ea2dd75df1bef333405b0d7f6b8c4951a0fbb60e780cb69f
 F ext/fts3/unicode/parseunicode.tcl da577d1384810fb4e2b209bf3313074353193e95
 F ext/fts5/extract_api_docs.tcl a36e54ec777172ddd3f9a88daf593b00848368e0
 F ext/fts5/fts5.h 62f3e33ceeb9a428db139f9c012186b371da1cc7
@@ -1566,7 +1566,7 @@ F vsixtest/vsixtest.tcl 6a9a6ab600c25a91a7acc6293828957a386a8a93
 F vsixtest/vsixtest.vcxproj.data 2ed517e100c66dc455b492e1a33350c1b20fbcdc
 F vsixtest/vsixtest.vcxproj.filters 37e51ffedcdb064aad6ff33b6148725226cd608e
 F vsixtest/vsixtest_TemporaryKey.pfx e5b1b036facdb453873e7084e1cae9102ccc67a0
-P 285005a9bcb210bb2a9aa9fed6a19d4b78641a6e7622d469bd0d2a365b2c0735
-R d28630b090e29d249202016ad6a6d827
-U drh
-Z dd88d1a97157edc79245fb3d73eca0b6
+P ef3a7c877a7549b351aafd983cfa96c863eb2641b6218bdd5cb563f659f879d8
+R 646761e50bb7679c40f23d2ddea8c902
+U dan
+Z f4376c790f10e78888f80aca2d0f09f6

manifest.uuid

@@ -1 +1 @@
-ef3a7c877a7549b351aafd983cfa96c863eb2641b6218bdd5cb563f659f879d8
+16a8e84fa7f67a467f824bdd7f72cbd6a6e95dab8cc7aa1e0e751720b98f3e31