+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
/*
**************************************************************************
* Copyright (C) 2002-2016 International Business Machines Corporation
status = U_REGEX_INVALID_CAPTURE_GROUP_NAME;
}
}
-
+
} else if (u_isdigit(nextChar)) {
// $n Scan for a capture group number
int32_t numCaptureGroups = fPattern->fGroupMap->size();
break;
}
(void)UTEXT_NEXT32(replacement);
- groupNum=groupNum*10 + nextDigitVal;
+ groupNum=groupNum*10 + nextDigitVal;
++numDigits;
}
} else {
if (findProgressInterrupt(startPos, status))
return FALSE;
}
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
case START_START:
// Matches are only possible at the start of the input string
return FALSE;
}
}
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
case START_STRING:
case START_CHAR:
return FALSE;
}
}
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
case START_LINE:
{
- UChar32 c;
+ UChar32 ch;
if (startPos == fAnchorStart) {
MatchAt(startPos, FALSE, status);
if (U_FAILURE(status)) {
return TRUE;
}
UTEXT_SETNATIVEINDEX(fInputText, startPos);
- c = UTEXT_NEXT32(fInputText);
+ ch = UTEXT_NEXT32(fInputText);
startPos = UTEXT_GETNATIVEINDEX(fInputText);
} else {
UTEXT_SETNATIVEINDEX(fInputText, startPos);
- c = UTEXT_PREVIOUS32(fInputText);
+ ch = UTEXT_PREVIOUS32(fInputText);
UTEXT_SETNATIVEINDEX(fInputText, startPos);
}
if (fPattern->fFlags & UREGEX_UNIX_LINES) {
for (;;) {
- if (c == 0x0a) {
+ if (ch == 0x0a) {
MatchAt(startPos, FALSE, status);
if (U_FAILURE(status)) {
return FALSE;
fHitEnd = TRUE;
return FALSE;
}
- c = UTEXT_NEXT32(fInputText);
+ ch = UTEXT_NEXT32(fInputText);
startPos = UTEXT_GETNATIVEINDEX(fInputText);
// Note that it's perfectly OK for a pattern to have a zero-length
// match at the end of a string, so we must make sure that the loop
}
} else {
for (;;) {
- if (isLineTerminator(c)) {
- if (c == 0x0d && startPos < fActiveLimit && UTEXT_CURRENT32(fInputText) == 0x0a) {
+ if (isLineTerminator(ch)) {
+ if (ch == 0x0d && startPos < fActiveLimit && UTEXT_CURRENT32(fInputText) == 0x0a) {
(void)UTEXT_NEXT32(fInputText);
startPos = UTEXT_GETNATIVEINDEX(fInputText);
}
fHitEnd = TRUE;
return FALSE;
}
- c = UTEXT_NEXT32(fInputText);
+ ch = UTEXT_NEXT32(fInputText);
startPos = UTEXT_GETNATIVEINDEX(fInputText);
// Note that it's perfectly OK for a pattern to have a zero-length
// match at the end of a string, so we must make sure that the loop
}
default:
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
}
- U_ASSERT(FALSE);
- return FALSE;
+ UPRV_UNREACHABLE;
}
if (findProgressInterrupt(startPos, status))
return FALSE;
}
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
case START_START:
// Matches are only possible at the start of the input string
return FALSE;
}
}
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
case START_STRING:
case START_CHAR:
return FALSE;
}
}
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
case START_LINE:
{
- UChar32 c;
+ UChar32 ch;
if (startPos == fAnchorStart) {
MatchChunkAt(startPos, FALSE, status);
if (U_FAILURE(status)) {
if (fMatch) {
return TRUE;
}
+ // In bug 31063104 which has a zero-length text buffer we get here with
+ // inputBuf=NULL, startPos=fActiveLimit=0 (and fMatch F) which violates the
+ // requirement for U16_FWD_1 (utf16.h) that startPos < fActiveLimit. Having
+ // inputBuf=NULL (chunkContexts NULL) is probably due to an error in the
+ // CFStringUText functions. Nevertheless, to be defensive, add test below.
+ if (startPos >= testLen) {
+ fHitEnd = TRUE;
+ return FALSE;
+ }
U16_FWD_1(inputBuf, startPos, fActiveLimit);
}
if (fPattern->fFlags & UREGEX_UNIX_LINES) {
for (;;) {
- c = inputBuf[startPos-1];
- if (c == 0x0a) {
+ ch = inputBuf[startPos-1];
+ if (ch == 0x0a) {
MatchChunkAt(startPos, FALSE, status);
if (U_FAILURE(status)) {
return FALSE;
}
} else {
for (;;) {
- c = inputBuf[startPos-1];
- if (isLineTerminator(c)) {
- if (c == 0x0d && startPos < fActiveLimit && inputBuf[startPos] == 0x0a) {
+ ch = inputBuf[startPos-1];
+ if (isLineTerminator(ch)) {
+ if (ch == 0x0d && startPos < fActiveLimit && inputBuf[startPos] == 0x0a) {
startPos++;
}
MatchChunkAt(startPos, FALSE, status);
}
default:
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
}
- U_ASSERT(FALSE);
- return FALSE;
+ UPRV_UNREACHABLE;
}
break;
}
i++;
- dest[i] = utext_extract_replace(fInputText, dest[i],
+ dest[i] = utext_extract_replace(fInputText, dest[i],
start64(groupNum, status), end64(groupNum, status), &status);
}
if (dest[i] == NULL) {
dest[i] = utext_openUChars(NULL, NULL, 0, &status);
} else {
- static UChar emptyString[] = {(UChar)0};
+ static const UChar emptyString[] = {(UChar)0};
utext_replace(dest[i], 0, utext_nativeLength(dest[i]), emptyString, 0, &status);
}
}
int64_t *pat = fPattern->fCompiledPat->getBuffer();
const UChar *litText = fPattern->fLiteralText.getBuffer();
- UVector *sets = fPattern->fSets;
+ UVector *fSets = fPattern->fSets;
fFrameSize = fPattern->fFrameSize;
REStackFrame *fp = resetStack();
// There is input left. Pick up one char and test it for set membership.
UChar32 c = UTEXT_NEXT32(fInputText);
- U_ASSERT(opValue > 0 && opValue < sets->size());
+ U_ASSERT(opValue > 0 && opValue < fSets->size());
if (c<256) {
Regex8BitSet *s8 = &fPattern->fSets8[opValue];
if (s8->contains(c)) {
break;
}
} else {
- UnicodeSet *s = (UnicodeSet *)sets->elementAt(opValue);
+ UnicodeSet *s = (UnicodeSet *)fSets->elementAt(opValue);
if (s->contains(c)) {
// The character is in the set. A Match.
fp->fInputIdx = UTEXT_GETNATIVEINDEX(fInputText);
}
}
fp = StateSave(fp, fp->fPatIdx, status);
+ } else {
+ // Increment time-out counter. (StateSave() does it if count >= minCount)
+ fTickCounter--;
+ if (fTickCounter <= 0) {
+ IncrementTime(status); // Re-initializes fTickCounter
+ }
}
+
fp->fPatIdx = opValue + 4; // Loop back.
}
break;
// We haven't met the minimum number of matches yet.
// Loop back for another one.
fp->fPatIdx = opValue + 4; // Loop back.
+ // Increment time-out counter. (StateSave() does it if count >= minCount)
+ fTickCounter--;
+ if (fTickCounter <= 0) {
+ IncrementTime(status); // Re-initializes fTickCounter
+ }
} else {
// We do have the minimum number of matches.
if (newFP == (int64_t *)fp) {
break;
}
- int32_t i;
- for (i=0; i<fFrameSize; i++) {
- newFP[i] = ((int64_t *)fp)[i];
+ int32_t j;
+ for (j=0; j<fFrameSize; j++) {
+ newFP[j] = ((int64_t *)fp)[j];
}
fp = (REStackFrame *)newFP;
fStack->setSize(newStackSize);
// This makes the capture groups from within the look-ahead
// expression available.
int64_t *newFP = fStack->getBuffer() + newStackSize - fFrameSize;
- int32_t i;
- for (i=0; i<fFrameSize; i++) {
- newFP[i] = ((int64_t *)fp)[i];
+ int32_t j;
+ for (j=0; j<fFrameSize; j++) {
+ newFP[j] = ((int64_t *)fp)[j];
}
fp = (REStackFrame *)newFP;
fStack->setSize(newStackSize);
// This op scans through all matching input.
// The following LOOP_C op emulates stack unwinding if the following pattern fails.
{
- U_ASSERT(opValue > 0 && opValue < sets->size());
+ U_ASSERT(opValue > 0 && opValue < fSets->size());
Regex8BitSet *s8 = &fPattern->fSets8[opValue];
- UnicodeSet *s = (UnicodeSet *)sets->elementAt(opValue);
+ UnicodeSet *s = (UnicodeSet *)fSets->elementAt(opValue);
// Loop through input, until either the input is exhausted or
// we reach a character that is not a member of the set.
default:
// Trouble. The compiled pattern contains an entry with an
// unrecognized type tag.
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
}
if (U_FAILURE(status)) {
int64_t *pat = fPattern->fCompiledPat->getBuffer();
const UChar *litText = fPattern->fLiteralText.getBuffer();
- UVector *sets = fPattern->fSets;
+ UVector *fSets = fPattern->fSets;
const UChar *inputBuf = fInputText->chunkContents;
break;
}
- U_ASSERT(opValue > 0 && opValue < sets->size());
+ U_ASSERT(opValue > 0 && opValue < fSets->size());
// There is input left. Pick up one char and test it for set membership.
UChar32 c;
break;
}
} else {
- UnicodeSet *s = (UnicodeSet *)sets->elementAt(opValue);
+ UnicodeSet *s = (UnicodeSet *)fSets->elementAt(opValue);
if (s->contains(c)) {
// The character is in the set. A Match.
break;
}
}
fp = StateSave(fp, fp->fPatIdx, status);
+ } else {
+ // Increment time-out counter. (StateSave() does it if count >= minCount)
+ fTickCounter--;
+ if (fTickCounter <= 0) {
+ IncrementTime(status); // Re-initializes fTickCounter
+ }
}
fp->fPatIdx = opValue + 4; // Loop back.
}
// We haven't met the minimum number of matches yet.
// Loop back for another one.
fp->fPatIdx = opValue + 4; // Loop back.
+ fTickCounter--;
+ if (fTickCounter <= 0) {
+ IncrementTime(status); // Re-initializes fTickCounter
+ }
} else {
// We do have the minimum number of matches.
if (newFP == (int64_t *)fp) {
break;
}
- int32_t i;
- for (i=0; i<fFrameSize; i++) {
- newFP[i] = ((int64_t *)fp)[i];
+ int32_t j;
+ for (j=0; j<fFrameSize; j++) {
+ newFP[j] = ((int64_t *)fp)[j];
}
fp = (REStackFrame *)newFP;
fStack->setSize(newStackSize);
// This makes the capture groups from within the look-ahead
// expression available.
int64_t *newFP = fStack->getBuffer() + newStackSize - fFrameSize;
- int32_t i;
- for (i=0; i<fFrameSize; i++) {
- newFP[i] = ((int64_t *)fp)[i];
+ int32_t j;
+ for (j=0; j<fFrameSize; j++) {
+ newFP[j] = ((int64_t *)fp)[j];
}
fp = (REStackFrame *)newFP;
fStack->setSize(newStackSize);
if (lbStartIdx < 0) {
// First time through loop.
lbStartIdx = fp->fInputIdx - minML;
- if (lbStartIdx > 0) {
+ if (lbStartIdx > 0 && lbStartIdx < fInputLength) {
U16_SET_CP_START(inputBuf, 0, lbStartIdx);
}
} else {
if (lbStartIdx < 0) {
// First time through loop.
lbStartIdx = fp->fInputIdx - minML;
- if (lbStartIdx > 0) {
+ if (lbStartIdx > 0 && lbStartIdx < fInputLength) {
U16_SET_CP_START(inputBuf, 0, lbStartIdx);
}
} else {
// This op scans through all matching input.
// The following LOOP_C op emulates stack unwinding if the following pattern fails.
{
- U_ASSERT(opValue > 0 && opValue < sets->size());
+ U_ASSERT(opValue > 0 && opValue < fSets->size());
Regex8BitSet *s8 = &fPattern->fSets8[opValue];
- UnicodeSet *s = (UnicodeSet *)sets->elementAt(opValue);
+ UnicodeSet *s = (UnicodeSet *)fSets->elementAt(opValue);
// Loop through input, until either the input is exhausted or
// we reach a character that is not a member of the set.
default:
// Trouble. The compiled pattern contains an entry with an
// unrecognized type tag.
- U_ASSERT(FALSE);
+ UPRV_UNREACHABLE;
}
if (U_FAILURE(status)) {
U_NAMESPACE_END
#endif // !UCONFIG_NO_REGULAR_EXPRESSIONS
+
projects / apple / icu.git / blobdiff