#include "unicode/utypes.h"
#if !UCONFIG_NO_REGULAR_EXPRESSIONS
#include "unicode/regex.h"
#include "unicode/uclean.h"
#include "uassert.h"
#include "uhash.h"
#include "uvector.h"
#include "uvectr32.h"
#include "uvectr64.h"
#include "regexcmp.h"
#include "regeximp.h"
#include "regexst.h"
U_NAMESPACE_BEGIN
RegexPattern::RegexPattern() {
init();
}
RegexPattern::RegexPattern(const RegexPattern &other) : UObject(other) {
init();
*this = other;
}
RegexPattern &RegexPattern::operator = (const RegexPattern &other) {
if (this == &other) {
return *this;
}
zap();
init();
fDeferredStatus = other.fDeferredStatus;
if (U_FAILURE(fDeferredStatus)) {
return *this;
}
if (other.fPatternString == NULL) {
fPatternString = NULL;
fPattern = utext_clone(fPattern, other.fPattern, FALSE, TRUE, &fDeferredStatus);
} else {
fPatternString = new UnicodeString(*(other.fPatternString));
if (fPatternString == NULL) {
fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
} else {
fPattern = utext_openConstUnicodeString(NULL, fPatternString, &fDeferredStatus);
}
}
if (U_FAILURE(fDeferredStatus)) {
return *this;
}
fFlags = other.fFlags;
fLiteralText = other.fLiteralText;
fMinMatchLen = other.fMinMatchLen;
fFrameSize = other.fFrameSize;
fDataSize = other.fDataSize;
fStaticSets = other.fStaticSets;
fStaticSets8 = other.fStaticSets8;
fStartType = other.fStartType;
fInitialStringIdx = other.fInitialStringIdx;
fInitialStringLen = other.fInitialStringLen;
*fInitialChars = *other.fInitialChars;
fInitialChar = other.fInitialChar;
*fInitialChars8 = *other.fInitialChars8;
fNeedsAltInput = other.fNeedsAltInput;
fCompiledPat->assign(*other.fCompiledPat, fDeferredStatus);
fGroupMap->assign(*other.fGroupMap, fDeferredStatus);
int32_t i;
int32_t numSets = other.fSets->size();
fSets8 = new Regex8BitSet[numSets];
if (fSets8 == NULL) {
fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
return *this;
}
for (i=1; i<numSets; i++) {
if (U_FAILURE(fDeferredStatus)) {
return *this;
}
UnicodeSet *sourceSet = (UnicodeSet *)other.fSets->elementAt(i);
UnicodeSet *newSet = new UnicodeSet(*sourceSet);
if (newSet == NULL) {
fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
break;
}
fSets->addElement(newSet, fDeferredStatus);
fSets8[i] = other.fSets8[i];
}
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;
}
void RegexPattern::init() {
fFlags = 0;
fCompiledPat = 0;
fLiteralText.remove();
fSets = NULL;
fSets8 = NULL;
fDeferredStatus = U_ZERO_ERROR;
fMinMatchLen = 0;
fFrameSize = 0;
fDataSize = 0;
fGroupMap = NULL;
fStaticSets = NULL;
fStaticSets8 = NULL;
fStartType = START_NO_INFO;
fInitialStringIdx = 0;
fInitialStringLen = 0;
fInitialChars = NULL;
fInitialChar = 0;
fInitialChars8 = NULL;
fNeedsAltInput = FALSE;
fNamedCaptureMap = NULL;
fPattern = NULL; fPatternString = NULL; fCompiledPat = new UVector64(fDeferredStatus);
fGroupMap = new UVector32(fDeferredStatus);
fSets = new UVector(fDeferredStatus);
fInitialChars = new UnicodeSet;
fInitialChars8 = new Regex8BitSet;
fNamedCaptureMap = uhash_open(uhash_hashUnicodeString, uhash_compareUnicodeString, uhash_compareLong, &fDeferredStatus);
if (U_FAILURE(fDeferredStatus)) {
return;
}
if (fCompiledPat == NULL || fGroupMap == NULL || fSets == NULL ||
fInitialChars == NULL || fInitialChars8 == NULL || fNamedCaptureMap == NULL) {
fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
return;
}
fSets->addElement((int32_t)0, fDeferredStatus);
uhash_setKeyDeleter(fNamedCaptureMap, uprv_deleteUObject);
}
void RegexPattern::zap() {
delete fCompiledPat;
fCompiledPat = NULL;
int i;
for (i=1; i<fSets->size(); i++) {
UnicodeSet *s;
s = (UnicodeSet *)fSets->elementAt(i);
if (s != NULL) {
delete s;
}
}
delete fSets;
fSets = NULL;
delete[] fSets8;
fSets8 = NULL;
delete fGroupMap;
fGroupMap = NULL;
delete fInitialChars;
fInitialChars = NULL;
delete fInitialChars8;
fInitialChars8 = NULL;
if (fPattern != NULL) {
utext_close(fPattern);
fPattern = NULL;
}
if (fPatternString != NULL) {
delete fPatternString;
fPatternString = NULL;
}
uhash_close(fNamedCaptureMap);
fNamedCaptureMap = NULL;
}
RegexPattern::~RegexPattern() {
zap();
}
RegexPattern *RegexPattern::clone() const {
RegexPattern *copy = new RegexPattern(*this);
return copy;
}
UBool RegexPattern::operator ==(const RegexPattern &other) const {
if (this->fFlags == other.fFlags && this->fDeferredStatus == other.fDeferredStatus) {
if (this->fPatternString != NULL && other.fPatternString != NULL) {
return *(this->fPatternString) == *(other.fPatternString);
} else if (this->fPattern == NULL) {
if (other.fPattern == NULL) {
return TRUE;
}
} else if (other.fPattern != NULL) {
UTEXT_SETNATIVEINDEX(this->fPattern, 0);
UTEXT_SETNATIVEINDEX(other.fPattern, 0);
return utext_equals(this->fPattern, other.fPattern);
}
}
return FALSE;
}
RegexPattern * U_EXPORT2
RegexPattern::compile(const UnicodeString ®ex,
uint32_t flags,
UParseError &pe,
UErrorCode &status)
{
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;
}
if (U_FAILURE(This->fDeferredStatus)) {
status = This->fDeferredStatus;
delete This;
return NULL;
}
This->fFlags = flags;
RegexCompile compiler(This, status);
compiler.compile(regex, pe, status);
if (U_FAILURE(status)) {
delete This;
This = NULL;
}
return This;
}
RegexPattern * U_EXPORT2
RegexPattern::compile(UText *regex,
uint32_t flags,
UParseError &pe,
UErrorCode &status)
{
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;
}
if (U_FAILURE(This->fDeferredStatus)) {
status = This->fDeferredStatus;
delete This;
return NULL;
}
This->fFlags = flags;
RegexCompile compiler(This, status);
compiler.compile(regex, pe, status);
if (U_FAILURE(status)) {
delete This;
This = NULL;
}
return This;
}
RegexPattern * U_EXPORT2
RegexPattern::compile(const UnicodeString ®ex,
UParseError &pe,
UErrorCode &err)
{
return compile(regex, 0, pe, err);
}
RegexPattern * U_EXPORT2
RegexPattern::compile(UText *regex,
UParseError &pe,
UErrorCode &err)
{
return compile(regex, 0, pe, err);
}
RegexPattern * U_EXPORT2
RegexPattern::compile(const UnicodeString ®ex,
uint32_t flags,
UErrorCode &err)
{
UParseError pe;
return compile(regex, flags, pe, err);
}
RegexPattern * U_EXPORT2
RegexPattern::compile(UText *regex,
uint32_t flags,
UErrorCode &err)
{
UParseError pe;
return compile(regex, flags, pe, err);
}
uint32_t RegexPattern::flags() const {
return fFlags;
}
RegexMatcher *RegexPattern::matcher(const UnicodeString &input,
UErrorCode &status) const {
RegexMatcher *retMatcher = matcher(status);
if (retMatcher != NULL) {
retMatcher->fDeferredStatus = status;
retMatcher->reset(input);
}
return retMatcher;
}
RegexMatcher *RegexPattern::matcher(UErrorCode &status) const {
RegexMatcher *retMatcher = NULL;
if (U_FAILURE(status)) {
return NULL;
}
if (U_FAILURE(fDeferredStatus)) {
status = fDeferredStatus;
return NULL;
}
retMatcher = new RegexMatcher(this);
if (retMatcher == NULL) {
status = U_MEMORY_ALLOCATION_ERROR;
return NULL;
}
return retMatcher;
}
UBool U_EXPORT2 RegexPattern::matches(const UnicodeString ®ex,
const UnicodeString &input,
UParseError &pe,
UErrorCode &status) {
if (U_FAILURE(status)) {return FALSE;}
UBool retVal;
RegexPattern *pat = NULL;
RegexMatcher *matcher = NULL;
pat = RegexPattern::compile(regex, 0, pe, status);
matcher = pat->matcher(input, status);
retVal = matcher->matches(status);
delete matcher;
delete pat;
return retVal;
}
UBool U_EXPORT2 RegexPattern::matches(UText *regex,
UText *input,
UParseError &pe,
UErrorCode &status) {
if (U_FAILURE(status)) {return FALSE;}
UBool retVal = FALSE;
RegexPattern *pat = NULL;
RegexMatcher *matcher = NULL;
pat = RegexPattern::compile(regex, 0, pe, status);
matcher = pat->matcher(status);
if (U_SUCCESS(status)) {
matcher->reset(input);
retVal = matcher->matches(status);
}
delete matcher;
delete pat;
return retVal;
}
UnicodeString RegexPattern::pattern() const {
if (fPatternString != NULL) {
return *fPatternString;
} else if (fPattern == NULL) {
return UnicodeString();
} else {
UErrorCode status = U_ZERO_ERROR;
int64_t nativeLen = utext_nativeLength(fPattern);
int32_t len16 = utext_extract(fPattern, 0, nativeLen, NULL, 0, &status); UnicodeString result;
status = U_ZERO_ERROR;
UChar *resultChars = result.getBuffer(len16);
utext_extract(fPattern, 0, nativeLen, resultChars, len16, &status); result.releaseBuffer(len16);
return result;
}
}
UText *RegexPattern::patternText(UErrorCode &status) const {
if (U_FAILURE(status)) {return NULL;}
status = U_ZERO_ERROR;
if (fPattern != NULL) {
return fPattern;
} else {
RegexStaticSets::initGlobals(&status);
return RegexStaticSets::gStaticSets->fEmptyText;
}
}
int32_t RegexPattern::groupNumberFromName(const UnicodeString &groupName, UErrorCode &status) const {
if (U_FAILURE(status)) {
return 0;
}
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);
}
int32_t RegexPattern::split(const UnicodeString &input,
UnicodeString dest[],
int32_t destCapacity,
UErrorCode &status) const
{
if (U_FAILURE(status)) {
return 0;
};
RegexMatcher m(this);
int32_t r = 0;
if (U_SUCCESS(m.fDeferredStatus)) {
r = m.split(input, dest, destCapacity, status);
}
return r;
}
int32_t RegexPattern::split(UText *input,
UText *dest[],
int32_t destCapacity,
UErrorCode &status) const
{
if (U_FAILURE(status)) {
return 0;
};
RegexMatcher m(this);
int32_t r = 0;
if (U_SUCCESS(m.fDeferredStatus)) {
r = m.split(input, dest, destCapacity, status);
}
return r;
}
void RegexPattern::dumpOp(int32_t index) const {
(void)index; #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 *)) {
pinnedType = 0;
}
printf("%4d %08x %-15s ", index, op, opNames[pinnedType]);
switch (type) {
case URX_NOP:
case URX_DOTANY:
case URX_DOTANY_ALL:
case URX_FAIL:
case URX_CARET:
case URX_DOLLAR:
case URX_BACKSLASH_G:
case URX_BACKSLASH_X:
case URX_END:
case URX_DOLLAR_M:
case URX_CARET_M:
break;
case URX_RESERVED_OP:
case URX_START_CAPTURE:
case URX_END_CAPTURE:
case URX_STATE_SAVE:
case URX_JMP:
case URX_JMP_SAV:
case URX_JMP_SAV_X:
case URX_BACKSLASH_B:
case URX_BACKSLASH_BU:
case URX_BACKSLASH_D:
case URX_BACKSLASH_Z:
case URX_STRING_LEN:
case URX_CTR_INIT:
case URX_CTR_INIT_NG:
case URX_CTR_LOOP:
case URX_CTR_LOOP_NG:
case URX_RELOC_OPRND:
case URX_STO_SP:
case URX_LD_SP:
case URX_BACKREF:
case URX_STO_INP_LOC:
case URX_JMPX:
case URX_LA_START:
case URX_LA_END:
case URX_BACKREF_I:
case URX_LB_START:
case URX_LB_CONT:
case URX_LB_END:
case URX_LBN_CONT:
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:
printf("%d", val);
break;
case URX_ONECHAR:
case URX_ONECHAR_I:
printf("%c", val<256?val:'?');
break;
case URX_STRING:
case URX_STRING_I:
{
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 = '.';}
printf("%c", c);
}
}
break;
case URX_SETREF:
case URX_LOOP_SR_I:
{
UnicodeString s;
UnicodeSet *set = (UnicodeSet *)fSets->elementAt(val);
set->toPattern(s, TRUE);
for (int32_t i=0; i<s.length(); i++) {
printf("%c", s.charAt(i));
}
}
break;
case URX_STATIC_SETREF:
case URX_STAT_SETREF_N:
{
UnicodeString s;
if (val & URX_NEG_SET) {
printf("NOT ");
val &= ~URX_NEG_SET;
}
UnicodeSet *set = fStaticSets[val];
set->toPattern(s, TRUE);
for (int32_t i=0; i<s.length(); i++) {
printf("%c", s.charAt(i));
}
}
break;
default:
printf("??????");
break;
}
printf("\n");
#endif
}
void RegexPattern::dumpPattern() const {
#if defined(REGEX_DEBUG)
int index;
int i;
printf("Original Pattern: ");
UChar32 c = utext_next32From(fPattern, 0);
while (c != U_SENTINEL) {
if (c<32 || c>256) {
c = '.';
}
printf("%c", c);
c = UTEXT_NEXT32(fPattern);
}
printf("\n");
printf(" Min Match Length: %d\n", fMinMatchLen);
printf(" Match Start Type: %s\n", START_OF_MATCH_STR(fStartType));
if (fStartType == START_STRING) {
printf(" Initial match string: \"");
for (i=fInitialStringIdx; i<fInitialStringIdx+fInitialStringLen; i++) {
printf("%c", fLiteralText[i]); }
printf("\"\n");
} else if (fStartType == START_SET) {
int32_t numSetChars = fInitialChars->size();
if (numSetChars > 20) {
numSetChars = 20;
}
printf(" Match First Chars : ");
for (i=0; i<numSetChars; i++) {
UChar32 c = fInitialChars->charAt(i);
if (0x20<c && c <0x7e) {
printf("%c ", c);
} else {
printf("%#x ", c);
}
}
if (numSetChars < fInitialChars->size()) {
printf(" ...");
}
printf("\n");
} else if (fStartType == START_CHAR) {
printf(" First char of Match : ");
if (0x20 < fInitialChar && fInitialChar<0x7e) {
printf("%c\n", fInitialChar);
} else {
printf("%#x\n", 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;
char s[100];
name->extract(0, 99, s, sizeof(s), US_INV); int32_t number = el->value.integer;
printf(" %d\t%s\n", number, s);
}
}
printf("\nIndex Binary Type Operand\n" \
"-------------------------------------------\n");
for (index = 0; index<fCompiledPat->size(); index++) {
dumpOp(index);
}
printf("\n\n");
#endif
}
UOBJECT_DEFINE_RTTI_IMPLEMENTATION(RegexPattern)
U_NAMESPACE_END
#endif // !UCONFIG_NO_REGULAR_EXPRESSIONS