Files
qt-creator/src/libs/utils/filesearch.cpp

655 lines
23 KiB
C++
Raw Normal View History

/****************************************************************************
2008-12-02 12:01:29 +01:00
**
** Copyright (C) 2015 The Qt Company Ltd.
** Contact: http://www.qt.io/licensing
2008-12-02 12:01:29 +01:00
**
** This file is part of Qt Creator.
2008-12-02 12:01:29 +01:00
**
** Commercial License Usage
** Licensees holding valid commercial Qt licenses may use this file in
** accordance with the commercial license agreement provided with the
** Software or, alternatively, in accordance with the terms contained in
** a written agreement between you and The Qt Company. For licensing terms and
** conditions see http://www.qt.io/terms-conditions. For further information
** use the contact form at http://www.qt.io/contact-us.
**
** GNU Lesser General Public License Usage
** Alternatively, this file may be used under the terms of the GNU Lesser
** General Public License version 2.1 or version 3 as published by the Free
** Software Foundation and appearing in the file LICENSE.LGPLv21 and
** LICENSE.LGPLv3 included in the packaging of this file. Please review the
** following information to ensure the GNU Lesser General Public License
** requirements will be met: https://www.gnu.org/licenses/lgpl.html and
** http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.
**
** In addition, as a special exception, The Qt Company gives you certain additional
** rights. These rights are described in The Qt Company LGPL Exception
2010-12-17 16:01:08 +01:00
** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
**
****************************************************************************/
2008-12-02 14:09:21 +01:00
2008-12-02 12:01:29 +01:00
#include "filesearch.h"
#include "runextensions.h"
2008-12-02 12:01:29 +01:00
#include <QCoreApplication>
#include <QMutex>
#include <QRegularExpression>
#include <QTextCodec>
2008-12-02 12:01:29 +01:00
#include <cctype>
using namespace Utils;
2008-12-02 12:01:29 +01:00
2009-04-30 16:09:57 +02:00
static inline QString msgCanceled(const QString &searchTerm, int numMatches, int numFilesSearched)
{
return QCoreApplication::translate("Utils::FileSearch",
2009-04-30 16:09:57 +02:00
"%1: canceled. %n occurrences found in %2 files.",
0, numMatches).arg(searchTerm).arg(numFilesSearched);
2009-04-30 16:09:57 +02:00
}
static inline QString msgFound(const QString &searchTerm, int numMatches, int numFilesSearched)
{
return QCoreApplication::translate("Utils::FileSearch",
2009-04-30 16:09:57 +02:00
"%1: %n occurrences found in %2 files.",
0, numMatches).arg(searchTerm).arg(numFilesSearched);
2009-04-30 16:09:57 +02:00
}
2008-12-02 12:01:29 +01:00
namespace {
2008-12-02 14:09:21 +01:00
const int MAX_LINE_SIZE = 400;
QString clippedText(const QString &text, int maxLength)
{
if (text.length() > maxLength)
return text.left(maxLength) + QChar(0x2026); // '...'
return text;
}
// returns success
bool openStream(const QString &filePath, QTextCodec *encoding, QTextStream *stream, QFile *file,
QString *tempString,
const QMap<QString, QString> &fileToContentsMap)
{
if (fileToContentsMap.contains(filePath)) {
*tempString = fileToContentsMap.value(filePath);
stream->setString(tempString);
} else {
file->setFileName(filePath);
if (!file->open(QIODevice::ReadOnly))
return false;
stream->setDevice(file);
stream->setCodec(encoding);
}
return true;
}
class FileSearch
2008-12-02 14:09:21 +01:00
{
public:
FileSearch(const QString &searchTerm, QTextDocument::FindFlags flags,
QMap<QString, QString> fileToContentsMap);
FileSearchResultList operator()(QFutureInterface<FileSearchResultList> futureInterface,
const FileIterator::Item &item) const;
2008-12-02 14:09:21 +01:00
private:
QMap<QString, QString> fileToContentsMap;
QString searchTermLower;
QString searchTermUpper;
int termMaxIndex;
const QChar *termData;
const QChar *termDataLower;
const QChar *termDataUpper;
bool caseSensitive;
bool wholeWord;
};
2008-12-02 14:09:21 +01:00
class FileSearchRegExp
{
public:
FileSearchRegExp(const QString &searchTerm, QTextDocument::FindFlags flags,
QMap<QString, QString> fileToContentsMap);
FileSearchRegExp(const FileSearchRegExp &other);
FileSearchResultList operator()(QFutureInterface<FileSearchResultList> futureInterface,
const FileIterator::Item &item) const;
2008-12-02 14:09:21 +01:00
private:
QRegularExpressionMatch doGuardedMatch(const QString &line, int offset) const;
QMap<QString, QString> fileToContentsMap;
QRegularExpression expression;
mutable QMutex mutex;
};
2008-12-02 14:09:21 +01:00
FileSearch::FileSearch(const QString &searchTerm, QTextDocument::FindFlags flags,
QMap<QString, QString> fileToContentsMap)
{
this->fileToContentsMap = fileToContentsMap;
caseSensitive = (flags & QTextDocument::FindCaseSensitively);
wholeWord = (flags & QTextDocument::FindWholeWords);
searchTermLower = searchTerm.toLower();
searchTermUpper = searchTerm.toUpper();
termMaxIndex = searchTerm.length() - 1;
termData = searchTerm.constData();
termDataLower = searchTermLower.constData();
termDataUpper = searchTermUpper.constData();
}
FileSearchResultList FileSearch::operator()(QFutureInterface<FileSearchResultList> futureInterface,
const FileIterator::Item &item) const
{
FileSearchResultList results;
if (futureInterface.isCanceled())
return results;
QFile file;
QTextStream stream;
QString tempString;
if (!openStream(item.filePath, item.encoding, &stream, &file, &tempString, fileToContentsMap))
return results;
int lineNr = 0;
while (!stream.atEnd()) {
++lineNr;
const QString chunk = stream.readLine();
const QString resultItemText = clippedText(chunk, MAX_LINE_SIZE);
int chunkLength = chunk.length();
const QChar *chunkPtr = chunk.constData();
const QChar *chunkEnd = chunkPtr + chunkLength - 1;
for (const QChar *regionPtr = chunkPtr; regionPtr + termMaxIndex <= chunkEnd; ++regionPtr) {
const QChar *regionEnd = regionPtr + termMaxIndex;
if ( /* optimization check for start and end of region */
// case sensitive
(caseSensitive && *regionPtr == termData[0]
&& *regionEnd == termData[termMaxIndex])
||
// case insensitive
(!caseSensitive && (*regionPtr == termDataLower[0]
|| *regionPtr == termDataUpper[0])
&& (*regionEnd == termDataLower[termMaxIndex]
|| *regionEnd == termDataUpper[termMaxIndex]))
) {
bool equal = true;
2008-12-02 14:09:21 +01:00
// whole word check
const QChar *beforeRegion = regionPtr - 1;
const QChar *afterRegion = regionEnd + 1;
if (wholeWord
&& (((beforeRegion >= chunkPtr)
&& (beforeRegion->isLetterOrNumber()
|| ((*beforeRegion) == QLatin1Char('_'))))
||
((afterRegion <= chunkEnd)
&& (afterRegion->isLetterOrNumber()
|| ((*afterRegion) == QLatin1Char('_'))))
)) {
equal = false;
} else {
// check all chars
int regionIndex = 1;
for (const QChar *regionCursor = regionPtr + 1;
regionCursor < regionEnd;
++regionCursor, ++regionIndex) {
if ( // case sensitive
(caseSensitive
&& *regionCursor != termData[regionIndex])
||
// case insensitive
(!caseSensitive
&& *regionCursor != termDataLower[regionIndex]
&& *regionCursor != termDataUpper[regionIndex])
) {
equal = false;
2008-12-02 12:01:29 +01:00
}
2008-12-02 14:09:21 +01:00
}
}
if (equal) {
results << FileSearchResult(item.filePath, lineNr, resultItemText,
regionPtr - chunkPtr, termMaxIndex + 1,
QStringList());
regionPtr += termMaxIndex; // another +1 done by for-loop
2008-12-02 12:01:29 +01:00
}
}
}
if (futureInterface.isPaused())
futureInterface.waitForResume();
if (futureInterface.isCanceled())
break;
}
if (file.isOpen())
file.close();
return results;
}
FileSearchRegExp::FileSearchRegExp(const QString &searchTerm, QTextDocument::FindFlags flags,
QMap<QString, QString> fileToContentsMap)
{
this->fileToContentsMap = fileToContentsMap;
QString term = searchTerm;
if (flags & QTextDocument::FindWholeWords)
term = QString::fromLatin1("\\b%1\\b").arg(term);
const QRegularExpression::PatternOptions patternOptions = (flags & QTextDocument::FindCaseSensitively)
? QRegularExpression::NoPatternOption : QRegularExpression::CaseInsensitiveOption;
expression = QRegularExpression(term, patternOptions);
}
FileSearchRegExp::FileSearchRegExp(const FileSearchRegExp &other)
: fileToContentsMap(other.fileToContentsMap),
expression(other.expression)
{
}
QRegularExpressionMatch FileSearchRegExp::doGuardedMatch(const QString &line, int offset) const
{
QMutexLocker lock(&mutex);
return expression.match(line, offset);
}
FileSearchResultList FileSearchRegExp::operator()(QFutureInterface<FileSearchResultList> futureInterface,
const FileIterator::Item &item) const
{
FileSearchResultList results;
if (futureInterface.isCanceled())
return results;
QFile file;
QTextStream stream;
QString tempString;
if (!openStream(item.filePath, item.encoding, &stream, &file, &tempString, fileToContentsMap))
return results;
int lineNr = 0;
QString line;
QRegularExpressionMatch match;
while (!stream.atEnd()) {
++lineNr;
line = stream.readLine();
const QString resultItemText = clippedText(line, MAX_LINE_SIZE);
int lengthOfLine = line.size();
int pos = 0;
while ((match = doGuardedMatch(line, pos)).hasMatch()) {
pos = match.capturedStart();
results << FileSearchResult(item.filePath, lineNr, resultItemText,
pos, match.capturedLength(),
match.capturedTexts());
if (match.capturedLength() == 0)
break;
pos += match.capturedLength();
if (pos >= lengthOfLine)
break;
}
if (futureInterface.isPaused())
futureInterface.waitForResume();
if (futureInterface.isCanceled())
break;
2008-12-02 12:01:29 +01:00
}
if (file.isOpen())
file.close();
return results;
2008-12-02 14:09:21 +01:00
}
2008-12-02 12:01:29 +01:00
struct SearchState
2008-12-02 14:09:21 +01:00
{
SearchState(const QString &term, FileIterator *iterator) : searchTerm(term), files(iterator) {}
QString searchTerm;
FileIterator *files = 0;
FileSearchResultList cachedResults;
int numFilesSearched = 0;
int numMatches = 0;
};
SearchState initFileSearch(QFutureInterface<FileSearchResultList> &futureInterface,
const QString &searchTerm, FileIterator *files)
{
futureInterface.setProgressRange(0, files->maxProgress());
futureInterface.setProgressValueAndText(files->currentProgress(), msgFound(searchTerm, 0, 0));
return SearchState(searchTerm, files);
}
void collectSearchResults(QFutureInterface<FileSearchResultList> &futureInterface,
SearchState &state,
const FileSearchResultList &results)
{
state.numMatches += results.size();
state.cachedResults << results;
state.numFilesSearched += 1;
if (futureInterface.isProgressUpdateNeeded()
|| futureInterface.progressValue() == 0 /*workaround for regression in Qt*/) {
if (!state.cachedResults.isEmpty()) {
futureInterface.reportResult(state.cachedResults);
state.cachedResults.clear();
}
futureInterface.setProgressRange(0, state.files->maxProgress());
futureInterface.setProgressValueAndText(state.files->currentProgress(),
msgFound(state.searchTerm,
state.numMatches,
state.numFilesSearched));
}
}
void cleanUpFileSearch(QFutureInterface<FileSearchResultList> &futureInterface,
SearchState &state)
{
if (!state.cachedResults.isEmpty()) {
futureInterface.reportResult(state.cachedResults);
state.cachedResults.clear();
2008-12-02 12:01:29 +01:00
}
if (futureInterface.isCanceled()) {
futureInterface.setProgressValueAndText(state.files->currentProgress(),
msgCanceled(state.searchTerm,
state.numMatches,
state.numFilesSearched));
} else {
futureInterface.setProgressValueAndText(state.files->currentProgress(),
msgFound(state.searchTerm,
state.numMatches,
state.numFilesSearched));
}
delete state.files;
}
2008-12-02 12:01:29 +01:00
} // namespace
QFuture<FileSearchResultList> Utils::findInFiles(const QString &searchTerm, FileIterator *files,
QTextDocument::FindFlags flags, QMap<QString, QString> fileToContentsMap)
2008-12-02 12:01:29 +01:00
{
return mapReduce<FileSearchResultList>(std::cref(*files),
[searchTerm, files](QFutureInterface<FileSearchResultList> &futureInterface) {
return initFileSearch(futureInterface, searchTerm, files);
},
FileSearch(searchTerm, flags, fileToContentsMap),
&collectSearchResults,
&cleanUpFileSearch);
2008-12-02 12:01:29 +01:00
}
QFuture<FileSearchResultList> Utils::findInFilesRegExp(const QString &searchTerm, FileIterator *files,
QTextDocument::FindFlags flags, QMap<QString, QString> fileToContentsMap)
2008-12-02 12:01:29 +01:00
{
return mapReduce<FileSearchResultList>(std::cref(*files),
[searchTerm, files](QFutureInterface<FileSearchResultList> &futureInterface) {
return initFileSearch(futureInterface, searchTerm, files);
},
FileSearchRegExp(searchTerm, flags, fileToContentsMap),
&collectSearchResults,
&cleanUpFileSearch);
2008-12-02 12:01:29 +01:00
}
QString Utils::expandRegExpReplacement(const QString &replaceText, const QStringList &capturedTexts)
{
2010-11-03 11:37:04 +01:00
// handles \1 \\ \& & \t \n
QString result;
2010-10-19 08:55:55 +02:00
const int numCaptures = capturedTexts.size() - 1;
for (int i = 0; i < replaceText.length(); ++i) {
QChar c = replaceText.at(i);
if (c == QLatin1Char('\\') && i < replaceText.length() - 1) {
c = replaceText.at(++i);
if (c == QLatin1Char('\\')) {
result += QLatin1Char('\\');
} else if (c == QLatin1Char('&')) {
result += QLatin1Char('&');
} else if (c == QLatin1Char('t')) {
result += QLatin1Char('\t');
2010-11-03 11:37:04 +01:00
} else if (c == QLatin1Char('n')) {
result += QLatin1Char('\n');
} else if (c.isDigit()) {
int index = c.unicode()-'1';
if (index < numCaptures) {
result += capturedTexts.at(index+1);
} else {
result += QLatin1Char('\\');
result += c;
}
} else {
result += QLatin1Char('\\');
result += c;
}
} else if (c == QLatin1Char('&')) {
result += capturedTexts.at(0);
} else {
result += c;
}
}
return result;
}
namespace Utils {
namespace Internal {
QString matchCaseReplacement(const QString &originalText, const QString &replaceText)
{
if (originalText.isEmpty() || replaceText.isEmpty())
return replaceText;
//Now proceed with actual case matching
bool firstIsUpperCase = originalText.at(0).isUpper();
bool firstIsLowerCase = originalText.at(0).isLower();
bool restIsLowerCase = true; // to be verified
bool restIsUpperCase = true; // to be verified
for (int i = 1; i < originalText.length(); ++i) {
if (originalText.at(i).isUpper())
restIsLowerCase = false;
else if (originalText.at(i).isLower())
restIsUpperCase = false;
if (!restIsLowerCase && !restIsUpperCase)
return replaceText; // mixed
}
if (restIsLowerCase) {
QString res = replaceText.toLower();
if (firstIsUpperCase)
res.replace(0, 1, res.at(0).toUpper());
return res;
}
if (restIsUpperCase) {
QString res = replaceText.toUpper();
if (firstIsLowerCase)
res.replace(0, 1, res.at(0).toLower());
return res;
}
return replaceText; // mixed
}
}
QString matchCaseReplacement(const QString &originalText, const QString &replaceText)
{
if (originalText.isEmpty())
return replaceText;
//Find common prefix & suffix: these will be unaffected
const int replaceTextLen = replaceText.length();
const int originalTextLen = originalText.length();
int prefixLen = 0;
for (; prefixLen < replaceTextLen && prefixLen < originalTextLen; ++prefixLen)
if (replaceText.at(prefixLen).toLower() != originalText.at(prefixLen).toLower())
break;
int suffixLen = 0;
for (; suffixLen < replaceTextLen - prefixLen && suffixLen < originalTextLen - prefixLen; ++suffixLen)
if (replaceText.at(replaceTextLen - 1 - suffixLen).toLower() != originalText.at(originalTextLen- 1 - suffixLen).toLower())
break;
//keep prefix and suffix, and do actual replacement on the 'middle' of the string
return originalText.left(prefixLen)
+ Internal::matchCaseReplacement(originalText.mid(prefixLen, originalTextLen - prefixLen - suffixLen),
replaceText.mid(prefixLen, replaceTextLen - prefixLen - suffixLen))
+ originalText.right(suffixLen);
}
// #pragma mark -- FileIterator
void FileIterator::advance(FileIterator::const_iterator *it) const
{
if (it->m_index < 0) // == end
return;
++it->m_index;
const_cast<FileIterator *>(this)->update(it->m_index);
if (it->m_index < currentFileCount()) {
it->m_item.filePath = fileAt(it->m_index);
it->m_item.encoding = codecAt(it->m_index);
} else {
it->m_index = -1; // == end
it->m_item.filePath.clear();
it->m_item.encoding = 0;
}
}
FileIterator::const_iterator FileIterator::begin() const
{
const_cast<FileIterator *>(this)->update(0);
if (currentFileCount() == 0)
return end();
return FileIterator::const_iterator(this,
FileIterator::Item(fileAt(0), codecAt(0)),
0/*index*/);
}
FileIterator::const_iterator FileIterator::end() const
{
return FileIterator::const_iterator(this, FileIterator::Item(QString(), 0),
-1/*end*/);
}
// #pragma mark -- FileListIterator
FileListIterator::FileListIterator(const QStringList &fileList,
const QList<QTextCodec *> encodings)
: m_files(fileList),
m_encodings(encodings),
m_maxIndex(-1)
{
}
void FileListIterator::update(int requestedIndex)
{
if (requestedIndex > m_maxIndex)
m_maxIndex = requestedIndex;
}
int FileListIterator::currentFileCount() const
{
return m_files.size();
}
QString FileListIterator::fileAt(int index) const
{
return m_files.at(index);
}
QTextCodec *FileListIterator::codecAt(int index) const
{
return m_encodings.at(index);
}
int FileListIterator::maxProgress() const
{
return m_files.size();
}
int FileListIterator::currentProgress() const
{
return m_maxIndex + 1;
}
QTextCodec *FileListIterator::encodingAt(int index) const
{
if (index >= 0 && index < m_encodings.size())
return m_encodings.at(index);
return QTextCodec::codecForLocale();
}
// #pragma mark -- SubDirFileIterator
namespace {
const int MAX_PROGRESS = 1000;
}
SubDirFileIterator::SubDirFileIterator(const QStringList &directories, const QStringList &filters,
QTextCodec *encoding)
: m_filters(filters), m_progress(0)
{
m_encoding = (encoding == 0 ? QTextCodec::codecForLocale() : encoding);
qreal maxPer = qreal(MAX_PROGRESS) / directories.count();
foreach (const QString &directoryEntry, directories) {
if (!directoryEntry.isEmpty()) {
m_dirs.push(QDir(directoryEntry));
m_progressValues.push(maxPer);
m_processedValues.push(false);
}
}
}
void SubDirFileIterator::update(int index)
{
if (index < m_files.size())
return;
// collect files from the directories until we have enough for the given index
while (!m_dirs.isEmpty() && index >= m_files.size()) {
QDir dir = m_dirs.pop();
2010-10-19 08:55:55 +02:00
const qreal dirProgressMax = m_progressValues.pop();
const bool processed = m_processedValues.pop();
if (dir.exists()) {
QStringList subDirs;
if (!processed)
subDirs = dir.entryList(QDir::Dirs|QDir::Hidden|QDir::NoDotAndDotDot);
if (subDirs.isEmpty()) {
QStringList fileEntries = dir.entryList(m_filters,
QDir::Files|QDir::Hidden);
QStringListIterator it(fileEntries);
it.toBack();
while (it.hasPrevious()) {
const QString &file = it.previous();
m_files.append(dir.path()+ QLatin1Char('/') +file);
}
m_progress += dirProgressMax;
} else {
qreal subProgress = dirProgressMax/(subDirs.size()+1);
m_dirs.push(dir);
m_progressValues.push(subProgress);
m_processedValues.push(true);
QStringListIterator it(subDirs);
it.toBack();
while (it.hasPrevious()) {
const QString &directory = it.previous();
m_dirs.push(QDir(dir.path()+ QLatin1Char('/') + directory));
m_progressValues.push(subProgress);
m_processedValues.push(false);
}
}
} else {
m_progress += dirProgressMax;
}
}
if (index >= m_files.size())
m_progress = MAX_PROGRESS;
}
int SubDirFileIterator::currentFileCount() const
{
return m_files.size();
}
QString SubDirFileIterator::fileAt(int index) const
{
return m_files.at(index);
}
QTextCodec *SubDirFileIterator::codecAt(int index) const
{
Q_UNUSED(index)
return m_encoding;
}
int SubDirFileIterator::maxProgress() const
{
return MAX_PROGRESS;
}
int SubDirFileIterator::currentProgress() const
{
return qMin(qRound(m_progress), MAX_PROGRESS);
}
}