/*
** Translate UTF-8 to UTF-8.
**
** This has the effect of making sure that the string is well-formed
** UTF-8. Miscoded characters are removed.
**
** The translation is done in-place and aborted if the output
** overruns the input.
*/
int sqlite3Utf8To8(unsigned char *zIn){
unsigned char *zOut = zIn;
unsigned char *zStart = zIn;
u32 c;
while( zIn[0] && zOut<=zIn ){
c = sqlite3Utf8Read((const u8**)&zIn);
if( c!=0xfffd ){
WRITE_UTF8(zOut, c);
}
}
*zOut = 0;
return (int)(zOut - zStart);
}
/*
** Translate UTF-8 to UTF-8.
**
** This has the effect of making sure that the string is well-formed
** UTF-8. Miscoded characters are removed.
**
** The translation is done in-place (since it is impossible for the
** correct UTF-8 encoding to be longer than a malformed encoding).
*/
int sqlite3Utf8To8(unsigned char *zIn){
unsigned char *zOut = zIn;
unsigned char *zStart = zIn;
unsigned char *zTerm;
u32 c;
while( zIn[0] ){
c = sqlite3Utf8Read(zIn, zTerm, (const u8**)&zIn);
if( c!=0xfffd ){
WRITE_UTF8(zOut, c);
}
}
*zOut = 0;
return zOut - zStart;
}
/*
** Implementation of the like() SQL function. This function implements
** the build-in LIKE operator. The first argument to the function is the
** pattern and the second argument is the string. So, the SQL statements:
**
** A LIKE B
**
** is implemented as like(B,A).
**
** This same function (with a different compareInfo structure) computes
** the GLOB operator.
*/
static void likeFunc(
sqlite3_context *context,
int argc,
sqlite3_value **argv
){
const unsigned char *zA, *zB;
int escape = 0;
int nPat;
sqlite3 *db = sqlite3_context_db_handle(context);
zB = sqlite3_value_text(argv[0]);
zA = sqlite3_value_text(argv[1]);
/* Limit the length of the LIKE or GLOB pattern to avoid problems
** of deep recursion and N*N behavior in patternCompare().
*/
nPat = sqlite3_value_bytes(argv[0]);
testcase( nPat==db->aLimit[SQLITE_LIMIT_LIKE_PATTERN_LENGTH] );
testcase( nPat==db->aLimit[SQLITE_LIMIT_LIKE_PATTERN_LENGTH]+1 );
if( nPat > db->aLimit[SQLITE_LIMIT_LIKE_PATTERN_LENGTH] ){
sqlite3_result_error(context, "LIKE or GLOB pattern too complex", -1);
return;
}
assert( zB==sqlite3_value_text(argv[0]) ); /* Encoding did not change */
if( argc==3 ){
/* The escape character string must consist of a single UTF-8 character.
** Otherwise, return an error.
*/
const unsigned char *zEsc = sqlite3_value_text(argv[2]);
if( zEsc==0 ) return;
if( sqlite3Utf8CharLen((char*)zEsc, -1)!=1 ){
sqlite3_result_error(context,
"ESCAPE expression must be a single character", -1);
return;
}
escape = sqlite3Utf8Read(zEsc, &zEsc);
}
if( zA && zB ){
struct compareInfo *pInfo = sqlite3_user_data(context);
#ifdef SQLITE_TEST
sqlite3_like_count++;
#endif
sqlite3_result_int(context, patternCompare(zB, zA, pInfo, escape));
}
}
/*
** This routine is called from the TCL test function "translate_selftest".
** It checks that the primitives for serializing and deserializing
** characters in each encoding are inverses of each other.
*/
void sqlite3UtfSelfTest(void){
unsigned int i, t;
unsigned char zBuf[20];
unsigned char *z;
int n;
unsigned int c;
for(i=0; i<0x00110000; i++){
z = zBuf;
WRITE_UTF8(z, i);
n = (int)(z-zBuf);
assert( n>0 && n<=4 );
z[0] = 0;
z = zBuf;
c = sqlite3Utf8Read((const u8**)&z);
t = i;
if( i>=0xD800 && i<=0xDFFF ) t = 0xFFFD;
if( (i&0xFFFFFFFE)==0xFFFE ) t = 0xFFFD;
assert( c==t );
assert( (z-zBuf)==n );
}
for(i=0; i<0x00110000; i++){
if( i>=0xD800 && i<0xE000 ) continue;
z = zBuf;
WRITE_UTF16LE(z, i);
n = (int)(z-zBuf);
assert( n>0 && n<=4 );
z[0] = 0;
z = zBuf;
READ_UTF16LE(z, 1, c);
assert( c==i );
assert( (z-zBuf)==n );
}
for(i=0; i<0x00110000; i++){
if( i>=0xD800 && i<0xE000 ) continue;
z = zBuf;
WRITE_UTF16BE(z, i);
n = (int)(z-zBuf);
assert( n>0 && n<=4 );
z[0] = 0;
z = zBuf;
READ_UTF16BE(z, 1, c);
assert( c==i );
assert( (z-zBuf)==n );
}
}
/*
** Compare two UTF-8 strings for equality where the first string can
** potentially be a "glob" expression. Return true (1) if they
** are the same and false (0) if they are different.
**
** Globbing rules:
**
** '*' Matches any sequence of zero or more characters.
**
** '?' Matches exactly one character.
**
** [...] Matches one character from the enclosed list of
** characters.
**
** [^...] Matches one character not in the enclosed list.
**
** With the [...] and [^...] matching, a ']' character can be included
** in the list by making it the first character after '[' or '^'. A
** range of characters can be specified using '-'. Example:
** "[a-z]" matches any single lower-case letter. To match a '-', make
** it the last character in the list.
**
** This routine is usually quick, but can be N**2 in the worst case.
**
** Hints: to match '*' or '?', put them in "[]". Like this:
**
** abc[*]xyz Matches "abc*xyz" only
*/
static int patternCompare(
const u8 *zPattern, /* The glob pattern */
const u8 *zString, /* The string to compare against the glob */
const struct compareInfo *pInfo, /* Information about how to do the compare */
const int esc /* The escape character */
){
int c, c2;
int invert;
int seen;
u8 matchOne = pInfo->matchOne;
u8 matchAll = pInfo->matchAll;
u8 matchSet = pInfo->matchSet;
u8 noCase = pInfo->noCase;
int prevEscape = 0; /* True if the previous character was 'escape' */
while( (c = sqlite3Utf8Read(zPattern,&zPattern))!=0 ){
if( !prevEscape && c==matchAll ){
while( (c=sqlite3Utf8Read(zPattern,&zPattern)) == matchAll
|| c == matchOne ){
if( c==matchOne && sqlite3Utf8Read(zString, &zString)==0 ){
return 0;
}
}
if( c==0 ){
return 1;
}else if( c==esc ){
c = sqlite3Utf8Read(zPattern, &zPattern);
if( c==0 ){
return 0;
}
}else if( c==matchSet ){
assert( esc==0 ); /* This is GLOB, not LIKE */
assert( matchSet<0x80 ); /* '[' is a single-byte character */
while( *zString && patternCompare(&zPattern[-1],zString,pInfo,esc)==0 ){
SQLITE_SKIP_UTF8(zString);
}
return *zString!=0;
}
while( (c2 = sqlite3Utf8Read(zString,&zString))!=0 ){
if( noCase ){
GlogUpperToLower(c2);
GlogUpperToLower(c);
while( c2 != 0 && c2 != c ){
c2 = sqlite3Utf8Read(zString, &zString);
GlogUpperToLower(c2);
}
}else{
while( c2 != 0 && c2 != c ){
c2 = sqlite3Utf8Read(zString, &zString);
}
}
if( c2==0 ) return 0;
if( patternCompare(zPattern,zString,pInfo,esc) ) return 1;
}
return 0;
}else if( !prevEscape && c==matchOne ){
if( sqlite3Utf8Read(zString, &zString)==0 ){
return 0;
}
}else if( c==matchSet ){
int prior_c = 0;
assert( esc==0 ); /* This only occurs for GLOB, not LIKE */
seen = 0;
invert = 0;
c = sqlite3Utf8Read(zString, &zString);
if( c==0 ) return 0;
c2 = sqlite3Utf8Read(zPattern, &zPattern);
if( c2=='^' ){
invert = 1;
c2 = sqlite3Utf8Read(zPattern, &zPattern);
}
if( c2==']' ){
if( c==']' ) seen = 1;
c2 = sqlite3Utf8Read(zPattern, &zPattern);
}
while( c2 && c2!=']' ){
if( c2=='-' && zPattern[0]!=']' && zPattern[0]!=0 && prior_c>0 ){
c2 = sqlite3Utf8Read(zPattern, &zPattern);
if( c>=prior_c && c<=c2 ) seen = 1;
prior_c = 0;
}else{
if( c==c2 ){
seen = 1;
}
prior_c = c2;
}
c2 = sqlite3Utf8Read(zPattern, &zPattern);
}
if( c2==0 || (seen ^ invert)==0 ){
return 0;
}
}else if( esc==c && !prevEscape ){
prevEscape = 1;
}else{
c2 = sqlite3Utf8Read(zString, &zString);
if( noCase ){
GlogUpperToLower(c);
GlogUpperToLower(c2);
}
if( c!=c2 ){
return 0;
}
prevEscape = 0;
}
}
return *zString==0;
}
