RegularExpression.cpp [plain text]
#include "config.h"
#include "RegularExpression.h"
#include "Yarr.h"
#include "YarrInterpreter.h"
#include <wtf/Assertions.h>
#include <wtf/BumpPointerAllocator.h>
namespace JSC { namespace Yarr {
class RegularExpression::Private : public RefCounted<RegularExpression::Private> {
public:
static Ref<Private> create(const String& pattern, TextCaseSensitivity caseSensitivity, MultilineMode multilineMode)
{
return adoptRef(*new Private(pattern, caseSensitivity, multilineMode));
}
int lastMatchLength { -1 };
unsigned m_numSubpatterns;
std::unique_ptr<JSC::Yarr::BytecodePattern> m_regExpByteCode;
private:
Private(const String& pattern, TextCaseSensitivity caseSensitivity, MultilineMode multilineMode)
: m_regExpByteCode(compile(pattern, caseSensitivity, multilineMode))
{
}
std::unique_ptr<JSC::Yarr::BytecodePattern> compile(const String& patternString, TextCaseSensitivity caseSensitivity, MultilineMode multilineMode)
{
RegExpFlags flags = NoFlags;
if (caseSensitivity == TextCaseInsensitive)
flags = static_cast<RegExpFlags>(flags | FlagIgnoreCase);
if (multilineMode == MultilineEnabled)
flags = static_cast<RegExpFlags>(flags | FlagMultiline);
JSC::Yarr::YarrPattern pattern(patternString, flags, m_constructionErrorCode);
if (JSC::Yarr::hasError(m_constructionErrorCode)) {
LOG_ERROR("RegularExpression: YARR compile failed with '%s'", JSC::Yarr::errorMessage(m_constructionErrorCode));
return nullptr;
}
m_numSubpatterns = pattern.m_numSubpatterns;
return JSC::Yarr::byteCompile(pattern, &m_regexAllocator);
}
BumpPointerAllocator m_regexAllocator;
JSC::Yarr::ErrorCode m_constructionErrorCode { Yarr::ErrorCode::NoError };
};
RegularExpression::RegularExpression(const String& pattern, TextCaseSensitivity caseSensitivity, MultilineMode multilineMode)
: d(Private::create(pattern, caseSensitivity, multilineMode))
{
}
RegularExpression::RegularExpression(const RegularExpression& re)
: d(re.d)
{
}
RegularExpression::~RegularExpression()
{
}
RegularExpression& RegularExpression::operator=(const RegularExpression& re)
{
d = re.d;
return *this;
}
int RegularExpression::match(const String& str, int startFrom, int* matchLength) const
{
if (!d->m_regExpByteCode)
return -1;
if (str.isNull())
return -1;
int offsetVectorSize = (d->m_numSubpatterns + 1) * 2;
unsigned* offsetVector;
Vector<unsigned, 32> nonReturnedOvector;
nonReturnedOvector.grow(offsetVectorSize);
offsetVector = nonReturnedOvector.data();
ASSERT(offsetVector);
for (unsigned j = 0, i = 0; i < d->m_numSubpatterns + 1; j += 2, i++)
offsetVector[j] = JSC::Yarr::offsetNoMatch;
unsigned result;
if (str.length() <= INT_MAX)
result = JSC::Yarr::interpret(d->m_regExpByteCode.get(), str, startFrom, offsetVector);
else {
result = JSC::Yarr::offsetNoMatch;
}
if (result == JSC::Yarr::offsetNoMatch) {
d->lastMatchLength = -1;
return -1;
}
d->lastMatchLength = offsetVector[1] - offsetVector[0];
if (matchLength)
*matchLength = d->lastMatchLength;
return offsetVector[0];
}
int RegularExpression::searchRev(const String& str) const
{
int start = 0;
int pos;
int lastPos = -1;
int lastMatchLength = -1;
do {
int matchLength;
pos = match(str, start, &matchLength);
if (pos >= 0) {
if (pos + matchLength > lastPos + lastMatchLength) {
lastPos = pos;
lastMatchLength = matchLength;
}
start = pos + 1;
}
} while (pos != -1);
d->lastMatchLength = lastMatchLength;
return lastPos;
}
int RegularExpression::matchedLength() const
{
return d->lastMatchLength;
}
void replace(String& string, const RegularExpression& target, const String& replacement)
{
int index = 0;
while (index < static_cast<int>(string.length())) {
int matchLength;
index = target.match(string, index, &matchLength);
if (index < 0)
break;
string.replace(index, matchLength, replacement);
index += replacement.length();
if (!matchLength)
break; }
}
bool RegularExpression::isValid() const
{
return d->m_regExpByteCode.get();
}
} }