+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
//
// file: repattrn.cpp
//
/*
***************************************************************************
-* Copyright (C) 2002-2012 International Business Machines Corporation *
-* and others. All rights reserved. *
+* Copyright (C) 2002-2016 International Business Machines Corporation
+* and others. All rights reserved.
***************************************************************************
*/
#include "unicode/regex.h"
#include "unicode/uclean.h"
+#include "cmemory.h"
+#include "cstr.h"
#include "uassert.h"
+#include "uhash.h"
#include "uvector.h"
#include "uvectr32.h"
#include "uvectr64.h"
init();
// Copy simple fields
- if ( other.fPatternString == NULL ) {
+ fDeferredStatus = other.fDeferredStatus;
+
+ if (U_FAILURE(fDeferredStatus)) {
+ return *this;
+ }
+
+ if (other.fPatternString == NULL) {
fPatternString = NULL;
- fPattern = utext_clone(fPattern, other.fPattern, FALSE, TRUE, &fDeferredStatus);
+ fPattern = utext_clone(fPattern, other.fPattern, FALSE, TRUE, &fDeferredStatus);
} else {
fPatternString = new UnicodeString(*(other.fPatternString));
- UErrorCode status = U_ZERO_ERROR;
- fPattern = utext_openConstUnicodeString(NULL, fPatternString, &status);
- if (U_FAILURE(status)) {
+ if (fPatternString == NULL) {
fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
- return *this;
+ } else {
+ fPattern = utext_openConstUnicodeString(NULL, fPatternString, &fDeferredStatus);
}
}
+ if (U_FAILURE(fDeferredStatus)) {
+ return *this;
+ }
+
fFlags = other.fFlags;
fLiteralText = other.fLiteralText;
- fDeferredStatus = other.fDeferredStatus;
fMinMatchLen = other.fMinMatchLen;
fFrameSize = other.fFrameSize;
fDataSize = other.fDataSize;
- fMaxCaptureDigits = other.fMaxCaptureDigits;
fStaticSets = other.fStaticSets;
fStaticSets8 = other.fStaticSets8;
fSets8[i] = other.fSets8[i];
}
+ // Copy the named capture group hash map.
+ int32_t hashPos = UHASH_FIRST;
+ while (const UHashElement *hashEl = uhash_nextElement(other.fNamedCaptureMap, &hashPos)) {
+ if (U_FAILURE(fDeferredStatus)) {
+ break;
+ }
+ const UnicodeString *name = (const UnicodeString *)hashEl->key.pointer;
+ UnicodeString *key = new UnicodeString(*name);
+ int32_t val = hashEl->value.integer;
+ if (key == NULL) {
+ fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
+ } else {
+ uhash_puti(fNamedCaptureMap, key, val, &fDeferredStatus);
+ }
+ }
return *this;
}
fFrameSize = 0;
fDataSize = 0;
fGroupMap = NULL;
- fMaxCaptureDigits = 1;
fStaticSets = NULL;
fStaticSets8 = NULL;
fStartType = START_NO_INFO;
fInitialChar = 0;
fInitialChars8 = NULL;
fNeedsAltInput = FALSE;
+ fNamedCaptureMap = NULL;
fPattern = NULL; // will be set later
fPatternString = NULL; // may be set later
fSets = new UVector(fDeferredStatus);
fInitialChars = new UnicodeSet;
fInitialChars8 = new Regex8BitSet;
+ fNamedCaptureMap = uhash_open(uhash_hashUnicodeString, // Key hash function
+ uhash_compareUnicodeString, // Key comparator function
+ uhash_compareLong, // Value comparator function
+ &fDeferredStatus);
if (U_FAILURE(fDeferredStatus)) {
return;
}
if (fCompiledPat == NULL || fGroupMap == NULL || fSets == NULL ||
- fInitialChars == NULL || fInitialChars8 == NULL) {
+ fInitialChars == NULL || fInitialChars8 == NULL || fNamedCaptureMap == NULL) {
fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
return;
}
// Slot zero of the vector of sets is reserved. Fill it here.
fSets->addElement((int32_t)0, fDeferredStatus);
+
+ // fNamedCaptureMap owns its key strings, type (UnicodeString *)
+ uhash_setKeyDeleter(fNamedCaptureMap, uprv_deleteUObject);
}
delete fPatternString;
fPatternString = NULL;
}
+ uhash_close(fNamedCaptureMap);
+ fNamedCaptureMap = NULL;
}
if (U_FAILURE(status)) {
return NULL;
}
-
+
const uint32_t allFlags = UREGEX_CANON_EQ | UREGEX_CASE_INSENSITIVE | UREGEX_COMMENTS |
UREGEX_DOTALL | UREGEX_MULTILINE | UREGEX_UWORD |
UREGEX_ERROR_ON_UNKNOWN_ESCAPES | UREGEX_UNIX_LINES | UREGEX_LITERAL;
-
+
if ((flags & ~allFlags) != 0) {
status = U_REGEX_INVALID_FLAG;
return NULL;
}
-
+
if ((flags & UREGEX_CANON_EQ) != 0) {
status = U_REGEX_UNIMPLEMENTED;
return NULL;
}
-
+
RegexPattern *This = new RegexPattern;
if (This == NULL) {
status = U_MEMORY_ALLOCATION_ERROR;
return NULL;
}
This->fFlags = flags;
-
+
RegexCompile compiler(This, status);
compiler.compile(regex, pe, status);
-
+
if (U_FAILURE(status)) {
delete This;
This = NULL;
}
-
+
return This;
}
RegexCompile compiler(This, status);
compiler.compile(regex, pe, status);
-
+
if (U_FAILURE(status)) {
delete This;
This = NULL;
int64_t nativeLen = utext_nativeLength(fPattern);
int32_t len16 = utext_extract(fPattern, 0, nativeLen, NULL, 0, &status); // buffer overflow error
UnicodeString result;
-
+
status = U_ZERO_ERROR;
UChar *resultChars = result.getBuffer(len16);
utext_extract(fPattern, 0, nativeLen, resultChars, len16, &status); // unterminated warning
result.releaseBuffer(len16);
-
+
return result;
}
}
}
+//--------------------------------------------------------------------------------
+//
+// groupNumberFromName()
+//
+//--------------------------------------------------------------------------------
+int32_t RegexPattern::groupNumberFromName(const UnicodeString &groupName, UErrorCode &status) const {
+ if (U_FAILURE(status)) {
+ return 0;
+ }
+
+ // No need to explicitly check for syntactically valid names.
+ // Invalid ones will never be in the map, and the lookup will fail.
+
+ int32_t number = uhash_geti(fNamedCaptureMap, &groupName);
+ if (number == 0) {
+ status = U_REGEX_INVALID_CAPTURE_GROUP_NAME;
+ }
+ return number;
+}
+
+int32_t RegexPattern::groupNumberFromName(const char *groupName, int32_t nameLength, UErrorCode &status) const {
+ if (U_FAILURE(status)) {
+ return 0;
+ }
+ UnicodeString name(groupName, nameLength, US_INV);
+ return groupNumberFromName(name, status);
+}
+
//---------------------------------------------------------------------
//
}
-
//---------------------------------------------------------------------
//
// dump Output the compiled form of the pattern.
// Debugging function only.
//
//---------------------------------------------------------------------
-#if defined(REGEX_DEBUG)
void RegexPattern::dumpOp(int32_t index) const {
+ (void)index; // Suppress warnings in non-debug build.
+#if defined(REGEX_DEBUG)
static const char * const opNames[] = {URX_OPCODE_NAMES};
int32_t op = fCompiledPat->elementAti(index);
int32_t val = URX_VAL(op);
int32_t type = URX_TYPE(op);
int32_t pinnedType = type;
- if ((uint32_t)pinnedType >= sizeof(opNames)/sizeof(char *)) {
+ if ((uint32_t)pinnedType >= UPRV_LENGTHOF(opNames)) {
pinnedType = 0;
}
- REGEX_DUMP_DEBUG_PRINTF(("%4d %08x %-15s ", index, op, opNames[pinnedType]));
+ printf("%4d %08x %-15s ", index, op, opNames[pinnedType]);
switch (type) {
case URX_NOP:
case URX_DOTANY:
case URX_LBN_END:
case URX_LOOP_C:
case URX_LOOP_DOT_I:
+ case URX_BACKSLASH_H:
+ case URX_BACKSLASH_R:
+ case URX_BACKSLASH_V:
// types with an integer operand field.
- REGEX_DUMP_DEBUG_PRINTF(("%d", val));
+ printf("%d", val);
break;
case URX_ONECHAR:
case URX_ONECHAR_I:
- REGEX_DUMP_DEBUG_PRINTF(("%c", val<256?val:'?'));
+ if (val < 0x20) {
+ printf("%#x", val);
+ } else {
+ printf("'%s'", CStr(UnicodeString(val))());
+ }
break;
case URX_STRING:
int32_t lengthOp = fCompiledPat->elementAti(index+1);
U_ASSERT(URX_TYPE(lengthOp) == URX_STRING_LEN);
int32_t length = URX_VAL(lengthOp);
- int32_t i;
- for (i=val; i<val+length; i++) {
- UChar c = fLiteralText[i];
- if (c < 32 || c >= 256) {c = '.';}
- REGEX_DUMP_DEBUG_PRINTF(("%c", c));
- }
+ UnicodeString str(fLiteralText, val, length);
+ printf("%s", CStr(str)());
}
break;
UnicodeString s;
UnicodeSet *set = (UnicodeSet *)fSets->elementAt(val);
set->toPattern(s, TRUE);
- for (int32_t i=0; i<s.length(); i++) {
- REGEX_DUMP_DEBUG_PRINTF(("%c", s.charAt(i)));
- }
+ printf("%s", CStr(s)());
}
break;
{
UnicodeString s;
if (val & URX_NEG_SET) {
- REGEX_DUMP_DEBUG_PRINTF(("NOT "));
+ printf("NOT ");
val &= ~URX_NEG_SET;
}
UnicodeSet *set = fStaticSets[val];
set->toPattern(s, TRUE);
- for (int32_t i=0; i<s.length(); i++) {
- REGEX_DUMP_DEBUG_PRINTF(("%c", s.charAt(i)));
- }
+ printf("%s", CStr(s)());
}
break;
default:
- REGEX_DUMP_DEBUG_PRINTF(("??????"));
+ printf("??????");
break;
}
- REGEX_DUMP_DEBUG_PRINTF(("\n"));
-}
+ printf("\n");
#endif
+}
+void RegexPattern::dumpPattern() const {
#if defined(REGEX_DEBUG)
-U_CAPI void U_EXPORT2
-RegexPatternDump(const RegexPattern *This) {
int index;
- int i;
-
- REGEX_DUMP_DEBUG_PRINTF(("Original Pattern: "));
- UChar32 c = utext_next32From(This->fPattern, 0);
- while (c != U_SENTINEL) {
- if (c<32 || c>256) {
- c = '.';
- }
- REGEX_DUMP_DEBUG_PRINTF(("%c", c));
-
- c = UTEXT_NEXT32(This->fPattern);
- }
- REGEX_DUMP_DEBUG_PRINTF(("\n"));
- REGEX_DUMP_DEBUG_PRINTF((" Min Match Length: %d\n", This->fMinMatchLen));
- REGEX_DUMP_DEBUG_PRINTF((" Match Start Type: %s\n", START_OF_MATCH_STR(This->fStartType)));
- if (This->fStartType == START_STRING) {
- REGEX_DUMP_DEBUG_PRINTF((" Initial match string: \""));
- for (i=This->fInitialStringIdx; i<This->fInitialStringIdx+This->fInitialStringLen; i++) {
- REGEX_DUMP_DEBUG_PRINTF(("%c", This->fLiteralText[i])); // TODO: non-printables, surrogates.
- }
- REGEX_DUMP_DEBUG_PRINTF(("\"\n"));
- } else if (This->fStartType == START_SET) {
- int32_t numSetChars = This->fInitialChars->size();
- if (numSetChars > 20) {
- numSetChars = 20;
- }
- REGEX_DUMP_DEBUG_PRINTF((" Match First Chars : "));
- for (i=0; i<numSetChars; i++) {
- UChar32 c = This->fInitialChars->charAt(i);
- if (0x20<c && c <0x7e) {
- REGEX_DUMP_DEBUG_PRINTF(("%c ", c));
+ UnicodeString patStr;
+ for (UChar32 c = utext_next32From(fPattern, 0); c != U_SENTINEL; c = utext_next32(fPattern)) {
+ patStr.append(c);
+ }
+ printf("Original Pattern: \"%s\"\n", CStr(patStr)());
+ printf(" Min Match Length: %d\n", fMinMatchLen);
+ printf(" Match Start Type: %s\n", START_OF_MATCH_STR(fStartType));
+ if (fStartType == START_STRING) {
+ UnicodeString initialString(fLiteralText,fInitialStringIdx, fInitialStringLen);
+ printf(" Initial match string: \"%s\"\n", CStr(initialString)());
+ } else if (fStartType == START_SET) {
+ UnicodeString s;
+ fInitialChars->toPattern(s, TRUE);
+ printf(" Match First Chars: %s\n", CStr(s)());
+
+ } else if (fStartType == START_CHAR) {
+ printf(" First char of Match: ");
+ if (fInitialChar > 0x20) {
+ printf("'%s'\n", CStr(UnicodeString(fInitialChar))());
} else {
- REGEX_DUMP_DEBUG_PRINTF(("%#x ", c));
+ printf("%#x\n", fInitialChar);
}
- }
- if (numSetChars < This->fInitialChars->size()) {
- REGEX_DUMP_DEBUG_PRINTF((" ..."));
- }
- REGEX_DUMP_DEBUG_PRINTF(("\n"));
+ }
- } else if (This->fStartType == START_CHAR) {
- REGEX_DUMP_DEBUG_PRINTF((" First char of Match : "));
- if (0x20 < This->fInitialChar && This->fInitialChar<0x7e) {
- REGEX_DUMP_DEBUG_PRINTF(("%c\n", This->fInitialChar));
- } else {
- REGEX_DUMP_DEBUG_PRINTF(("%#x\n", This->fInitialChar));
- }
+ printf("Named Capture Groups:\n");
+ if (uhash_count(fNamedCaptureMap) == 0) {
+ printf(" None\n");
+ } else {
+ int32_t pos = UHASH_FIRST;
+ const UHashElement *el = NULL;
+ while ((el = uhash_nextElement(fNamedCaptureMap, &pos))) {
+ const UnicodeString *name = (const UnicodeString *)el->key.pointer;
+ int32_t number = el->value.integer;
+ printf(" %d\t%s\n", number, CStr(*name)());
+ }
}
- REGEX_DUMP_DEBUG_PRINTF(("\nIndex Binary Type Operand\n" \
- "-------------------------------------------\n"));
- for (index = 0; index<This->fCompiledPat->size(); index++) {
- This->dumpOp(index);
+ printf("\nIndex Binary Type Operand\n" \
+ "-------------------------------------------\n");
+ for (index = 0; index<fCompiledPat->size(); index++) {
+ dumpOp(index);
}
- REGEX_DUMP_DEBUG_PRINTF(("\n\n"));
-}
+ printf("\n\n");
#endif
+}
projects / apple / icu.git / blobdiff