1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621
|
//////////////////////////////////////////////////////////////////////////////
//////////////////////////////////////////////////////////////////////////////
//
// copyright : (C) 2008 by Eran Ifrah
// file name : search_thread.cpp
//
// -------------------------------------------------------------------------
// A
// _____ _ _ _ _
// / __ \ | | | | (_) |
// | / \/ ___ __| | ___| | _| |_ ___
// | | / _ \ / _ |/ _ \ | | | __/ _ )
// | \__/\ (_) | (_| | __/ |___| | || __/
// \____/\___/ \__,_|\___\_____/_|\__\___|
//
// F i l e
//
// This program is free software; you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation; either version 2 of the License, or
// (at your option) any later version.
//
//////////////////////////////////////////////////////////////////////////////
//////////////////////////////////////////////////////////////////////////////
#include "search_thread.h"
#include "clFilesCollector.h"
#include "clWildMatch.hpp"
#include "dirtraverser.h"
#include "file_logger.h"
#include "fileutils.h"
#include "macros.h"
#include <algorithm>
#include <iostream>
#include <set>
#include <wx/dir.h>
#include <wx/event.h>
#include <wx/fontmap.h>
#include <wx/log.h>
#include <wx/stopwatch.h>
#include <wx/tokenzr.h>
#include <wx/txtstrm.h>
#include <wx/wfstream.h>
#if !wxUSE_GUI
#include "cl_command_event.h" // Needed for the definition of wxCommandEvent
#endif
wxDEFINE_EVENT(wxEVT_SEARCH_THREAD_MATCHFOUND, wxCommandEvent);
wxDEFINE_EVENT(wxEVT_SEARCH_THREAD_SEARCHEND, wxCommandEvent);
wxDEFINE_EVENT(wxEVT_SEARCH_THREAD_SEARCHCANCELED, wxCommandEvent);
wxDEFINE_EVENT(wxEVT_SEARCH_THREAD_SEARCHSTARTED, wxCommandEvent);
#define SEND_ST_EVENT() \
if(owner) { \
wxPostEvent(owner, event); \
} else if(m_notifiedWindow) { \
wxPostEvent(m_notifiedWindow, event); \
}
//----------------------------------------------------------------
// SearchData
//----------------------------------------------------------------
namespace
{
bool is_word_char(wxChar ch) { return ch == '_' || wxIsalnum(ch); }
// Minumum of 10ms between events that this thread is sending to the main thread
constexpr long MIN_SEND_INTERVAL_MS = 1;
size_t send_count = 0;
} // namespace
const wxString& SearchData::GetExtensions() const { return m_validExt; }
SearchData& SearchData::operator=(const SearchData& rhs) { return Copy(rhs); }
SearchData& SearchData::Copy(const SearchData& other)
{
if(this == &other) {
return *this;
}
m_findString = other.m_findString.c_str();
m_flags = other.m_flags;
m_validExt = other.m_validExt.c_str();
m_rootDirs = other.m_rootDirs;
m_newTab = other.m_newTab;
m_owner = other.m_owner;
m_encoding = other.m_encoding.c_str();
m_replaceWith = other.m_replaceWith;
m_excludePatterns.clear();
m_excludePatterns.insert(m_excludePatterns.end(), other.m_excludePatterns.begin(), other.m_excludePatterns.end());
m_files.clear();
m_files.reserve(other.m_files.size());
m_file_scanner_flags = other.m_file_scanner_flags;
for(size_t i = 0; i < other.m_files.size(); ++i) {
m_files.Add(other.m_files.Item(i).c_str());
}
return *this;
}
//----------------------------------------------------------------
// SearchThread
//----------------------------------------------------------------
SearchThread::SearchThread()
: WorkerThread()
, m_reExpr(wxT(""))
{
m_stopWatch.Start();
}
SearchThread::~SearchThread() {}
wxRegEx& SearchThread::GetRegex(const wxString& expr, bool matchCase)
{
if(m_reExpr == expr && matchCase == m_matchCase) {
return m_regex;
} else {
m_reExpr = expr;
m_matchCase = matchCase;
#ifndef __WXMAC__
int flags = wxRE_ADVANCED;
#else
int flags = wxRE_DEFAULT;
#endif
if(!matchCase)
flags |= wxRE_ICASE;
m_regex.Compile(m_reExpr, flags);
}
return m_regex;
}
void SearchThread::PerformSearch(const SearchData& data) { Add(new SearchData(data)); }
void SearchThread::ProcessRequest(ThreadRequest* req)
{
FileLogger::RegisterThread(wxThread::GetCurrentId(), "Search Thread");
wxStopWatch sw;
m_summary = SearchSummary();
DoSearchFiles(req);
m_summary.SetElapsedTime(sw.Time());
SearchData* sd = (SearchData*)req;
m_summary.SetFindWhat(sd->GetFindString());
m_summary.SetReplaceWith(sd->GetReplaceWith());
// Send search end event
SendEvent(wxEVT_SEARCH_THREAD_SEARCHEND, sd->GetOwner());
}
void SearchThread::GetFiles(const SearchData* data, wxArrayString& files)
{
wxStopWatch sw;
clDEBUG() << "Building list of files ..." << endl;
wxStringSet_t scannedFiles;
const wxArrayString& rootDirs = data->GetRootDirs();
files = data->GetFiles();
// Populate "scannedFiles" with list of files to scan
scannedFiles.insert(files.begin(), files.end());
files.reserve(5000);
clDEBUG() << "Scanning directories..." << endl;
sw.Start();
clFileExtensionMatcher ext_matcher{ data->GetExtensions() };
clPathExcluder path_excluder{ data->GetExcludePatterns() };
wxStringSet_t visited_dirs;
for(size_t i = 0; i < rootDirs.size(); ++i) {
clDEBUG() << " scanning root directory:" << rootDirs.Item(i) << endl;
// collect only unique files that are matching the pattern
auto on_files = [&](const wxArrayString& paths) {
files.reserve(files.size() + paths.size());
for(const wxString& fullpath : paths) {
if(scannedFiles.insert(fullpath).second && ext_matcher.matches(fullpath)) {
files.Add(fullpath);
}
}
};
// do not traverse into excluded directories or directories that
// we already visited
auto on_folder = [&](const wxString& fullpath) -> bool {
return
// first time visiting this directory
visited_dirs.insert(fullpath).second &&
// is not excluded
!path_excluder.is_exclude_path(fullpath);
};
// make sure it's really a dir (not a fifo, etc.)
clFilesScanner scanner;
scanner.ScanWithCallbacks(rootDirs.Item(i), on_folder, on_files, data->GetFileScannerFlags());
clDEBUG() << " scanning root directory:" << rootDirs.Item(i) << "..done" << endl;
}
wxString duration;
duration << sw.Time() / 1000 << "." << sw.Time() % 1000;
clDEBUG() << "Scanning directories... done (" << duration << ")" << endl;
clDEBUG() << "Found" << files.size() << "files" << endl;
// sort the files found
clDEBUG() << "Sorting the matches..." << endl;
files.Sort([](const wxString& f1, const wxString& f2) -> int { return f1.CmpNoCase(f2); });
clDEBUG() << "Sorting the matches... done" << endl;
}
void SearchThread::DoSearchFiles(ThreadRequest* req)
{
SearchData* data = static_cast<SearchData*>(req);
// Get all files
if(data->GetFindString().IsEmpty()) {
SendEvent(wxEVT_SEARCH_THREAD_SEARCHSTARTED, data->GetOwner());
return;
}
StopSearch(false);
wxArrayString fileList;
GetFiles(data, fileList);
wxStopWatch sw;
// Send startup message to main thread
if(m_notifiedWindow || data->GetOwner()) {
wxCommandEvent event(wxEVT_SEARCH_THREAD_SEARCHSTARTED, GetId());
event.SetClientData(new SearchData(*data));
if(data->GetOwner()) {
::wxPostEvent(data->GetOwner(), event);
} else {
// since we are in if ( m_notifiedWindow || data->GetOwner() ) block...
::wxPostEvent(m_notifiedWindow, event);
}
}
for(size_t i = 0; i < fileList.Count(); i++) {
m_summary.SetNumFileScanned((int)i + 1);
// give user chance to cancel the search ...
if(TestStopSearch()) {
// Send cancel event
SendEvent(wxEVT_SEARCH_THREAD_SEARCHCANCELED, data->GetOwner());
StopSearch(false);
break;
}
DoSearchFile(fileList.Item(i), data);
}
}
bool SearchThread::TestStopSearch()
{
bool stop = false;
{
wxCriticalSectionLocker locker(m_cs);
stop = m_stopSearch;
}
return stop;
}
void SearchThread::StopSearch(bool stop)
{
wxCriticalSectionLocker locker(m_cs);
m_stopSearch = stop;
}
void SearchThread::DoSearchFile(const wxString& fileName, const SearchData* data)
{
// Process single lines
int lineNumber = 1;
if(!wxFileName::FileExists(fileName)) {
return;
}
// ignore binary executables
if(FileUtils::IsBinaryExecutable(fileName)) {
return;
}
size_t size = FileUtils::GetFileSize(fileName);
if(size == 0) {
return;
}
wxString fileData;
fileData.Alloc(size);
#if wxUSE_GUI
// support for other encoding
wxFontEncoding enc = wxFontMapper::GetEncodingFromName(data->GetEncoding().c_str());
wxCSConv fontEncConv(enc);
if(!FileUtils::ReadFileContent(fileName, fileData, fontEncConv)) {
m_summary.GetFailedFiles().Add(fileName);
return;
}
#else
if(!FileUtils::ReadFileContent(fileName, fileData, wxConvLibc)) {
m_summary.GetFailedFiles().Add(fileName);
return;
}
#endif
wxArrayString lines = ::wxStringTokenize(fileData, wxT("\n"), wxTOKEN_RET_EMPTY_ALL);
int lineOffset = 0;
if(data->IsRegularExpression()) {
// regular expression search
for(const wxString& line : lines) {
// Read the next line
DoSearchLineRE(line, lineNumber, lineOffset, fileName, data);
lineOffset += line.Length() + 1;
lineNumber++;
}
} else {
// simple search
wxString findString;
wxArrayString filters;
findString = data->GetFindString();
if(data->IsEnablePipeSupport()) {
if(data->GetFindString().Find('|') != wxNOT_FOUND) {
findString = data->GetFindString().BeforeFirst('|');
wxString filtersString = data->GetFindString().AfterFirst('|');
filters = ::wxStringTokenize(filtersString, "|", wxTOKEN_STRTOK);
if(!data->IsMatchCase()) {
for(size_t i = 0; i < filters.size(); ++i) {
filters.Item(i).MakeLower();
}
}
}
}
// Dont search for empty strings
if(findString.empty()) {
return;
}
if(!data->IsMatchCase()) {
findString.MakeLower();
}
for(const wxString& line : lines) {
DoSearchLine(line, lineNumber, lineOffset, fileName, data, findString, filters);
lineOffset += line.Length() + 1;
lineNumber++;
}
}
if(m_results.empty() == false) {
SendEvent(wxEVT_SEARCH_THREAD_MATCHFOUND, data->GetOwner());
}
}
void SearchThread::DoSearchLineRE(const wxString& line, const int lineNum, const int lineOffset,
const wxString& fileName, const SearchData* data)
{
wxRegEx& re = GetRegex(data->GetFindString(), data->IsMatchCase());
size_t col = 0;
int iCorrectedCol = 0;
int iCorrectedLen = 0;
wxString modLine = line;
if(re.IsValid()) {
while(re.Matches(modLine)) {
size_t start, len;
re.GetMatch(&start, &len);
col += start;
// Notify our match
// correct search Pos and Length owing to non plain ASCII multibyte characters
iCorrectedCol = FileUtils::UTF8Length(line.c_str(), col);
iCorrectedLen = FileUtils::UTF8Length(line.c_str(), col + len) - iCorrectedCol;
SearchResult result;
result.SetPosition(lineOffset + col);
result.SetColumnInChars((int)col);
result.SetColumn(iCorrectedCol);
result.SetLineNumber(lineNum);
result.SetPattern(line);
result.SetFileName(fileName);
result.SetLenInChars((int)len);
result.SetLen(iCorrectedLen);
result.SetFlags(data->m_flags);
result.SetFindWhat(data->GetFindString());
wxArrayString regexCaptures;
for(size_t i = 0; i < re.GetMatchCount(); ++i) {
regexCaptures.Add(re.GetMatch(modLine, i));
}
result.SetRegexCaptures(regexCaptures);
// Make sure our match is not on a comment
m_results.push_back(result);
m_summary.SetNumMatchesFound(m_summary.GetNumMatchesFound() + 1);
col += len;
// adjust the line
if(line.Length() - col <= 0)
break;
modLine = modLine.Right(line.Length() - col);
}
}
}
void SearchThread::DoSearchLine(const wxString& line, const int lineNum, const int lineOffset, const wxString& fileName,
const SearchData* data, const wxString& findWhat, const wxArrayString& filters)
{
wxString modLine = line;
if(!data->IsMatchCase()) {
modLine.MakeLower();
}
int pos = 0;
int col = 0;
int iCorrectedCol = 0;
int iCorrectedLen = 0;
while(pos != wxNOT_FOUND) {
pos = modLine.Find(findWhat);
if(pos != wxNOT_FOUND) {
col += pos;
// Pipe support
bool allFiltersOK = true;
if(!filters.IsEmpty()) {
// Apply the filters
for(size_t i = 0; i < filters.size() && allFiltersOK; ++i) {
allFiltersOK = (modLine.Find(filters.Item(i)) != wxNOT_FOUND);
}
}
// Pipe filtes OK?
if(!allFiltersOK)
return;
// we have a match
if(data->IsMatchWholeWord()) {
// make sure that the word before is not in the wordChars map
if(pos > 0 && is_word_char(modLine.GetChar(pos - 1))) {
if(!AdjustLine(modLine, pos, findWhat)) {
break;
} else {
col += (int)findWhat.Length();
continue;
}
}
// if we have more characters to the right, make sure that the first char does not match any
// in the wordCharsMap
if(pos + findWhat.Length() <= modLine.Length()) {
wxChar nextCh = modLine.GetChar(pos + findWhat.Length());
if(is_word_char(nextCh)) {
if(!AdjustLine(modLine, pos, findWhat)) {
break;
} else {
col += (int)findWhat.Length();
continue;
}
}
}
}
// Notify our match
// correct search Pos and Length owing to non plain ASCII multibyte characters
iCorrectedCol = FileUtils::UTF8Length(line.c_str(), col);
iCorrectedLen = FileUtils::UTF8Length(findWhat.c_str(), findWhat.Length());
SearchResult result;
result.SetPosition(lineOffset + col);
result.SetColumnInChars(col);
result.SetColumn(iCorrectedCol);
result.SetLineNumber(lineNum);
// Dont use match pattern larger than 500 chars
result.SetPattern(line.length() > 500 ? line.Mid(0, 500) : line);
result.SetFileName(fileName);
result.SetLenInChars((int)findWhat.Length());
result.SetLen(iCorrectedLen);
result.SetFindWhat(data->GetFindString());
result.SetFlags(data->m_flags);
m_results.push_back(result);
m_summary.SetNumMatchesFound(m_summary.GetNumMatchesFound() + 1);
if(!AdjustLine(modLine, pos, findWhat)) {
break;
}
col += (int)findWhat.Length();
}
}
}
bool SearchThread::AdjustLine(wxString& line, int& pos, const wxString& findString)
{
// adjust the current line
if(line.Length() - (pos + findString.Length()) >= findString.Length()) {
line = line.Right(line.Length() - (pos + findString.Length()));
pos += (int)findString.Length();
return true;
} else {
return false;
}
}
void SearchThread::SendEvent(wxEventType type, wxEvtHandler* owner)
{
if(!m_notifiedWindow && !owner)
return;
wxCommandEvent event(type, GetId());
if(type == wxEVT_SEARCH_THREAD_MATCHFOUND) {
// match found and we scanned 10 files
event.SetClientData(new SearchResultList(m_results));
m_results.clear();
SEND_ST_EVENT();
} else if((type == wxEVT_SEARCH_THREAD_SEARCHEND) || (type == wxEVT_SEARCH_THREAD_SEARCHCANCELED)) {
// search eneded, if we got any matches "buffered" send them before the
// the summary event
if(m_results.empty() == false) {
wxCommandEvent evt(wxEVT_SEARCH_THREAD_MATCHFOUND, GetId());
evt.SetClientData(new SearchResultList(m_results));
if(owner) {
wxPostEvent(owner, evt);
} else if(m_notifiedWindow) {
wxPostEvent(m_notifiedWindow, evt);
}
}
m_results.clear();
// Now send the summary event
event.SetClientData(type == wxEVT_SEARCH_THREAD_SEARCHEND ? new SearchSummary(m_summary) : nullptr);
SEND_ST_EVENT();
}
// to avoid flooding the UI with search events, sleep for 1ms after each
// send_event call
++send_count;
if(send_count >= 10) {
wxThread::Sleep(1);
send_count = 0;
}
}
void SearchThread::FilterFiles(wxArrayString& files, const SearchData* data)
{
wxArrayString tmpFiles;
std::set<wxString> uniqueFiles;
const wxArrayString& excludePatterns = data->GetExcludePatterns();
const wxString& mask = data->GetExtensions();
std::for_each(files.begin(), files.end(), [&](const wxString& filename) {
if(uniqueFiles.count(filename))
return;
uniqueFiles.insert(filename);
if(FileUtils::WildMatch(mask, filename) && !FileUtils::WildMatch(excludePatterns, filename)) {
tmpFiles.Add(filename);
}
});
files.swap(tmpFiles);
files.Sort([](const wxString& f1, const wxString& f2) -> int { return f1.CmpNoCase(f2); });
}
static SearchThread* gs_SearchThread = NULL;
void SearchThreadST::Free()
{
if(gs_SearchThread) {
delete gs_SearchThread;
}
gs_SearchThread = NULL;
}
SearchThread* SearchThreadST::Get()
{
if(gs_SearchThread == NULL)
gs_SearchThread = new SearchThread;
return gs_SearchThread;
}
JSONItem SearchResult::ToJSON() const
{
JSONItem json = JSONItem::createObject();
json.addProperty("file", m_fileName);
json.addProperty("line", m_lineNumber);
json.addProperty("col", m_column);
json.addProperty("pos", m_position);
json.addProperty("pattern", m_pattern);
json.addProperty("len", m_len);
json.addProperty("flags", m_flags);
json.addProperty("columnInChars", m_columnInChars);
json.addProperty("lenInChars", m_lenInChars);
json.addProperty("regexCaptures", m_regexCaptures);
return json;
}
void SearchResult::FromJSON(const JSONItem& json)
{
m_position = json.namedObject("pos").toInt(m_position);
m_column = json.namedObject("col").toInt(m_column);
m_lineNumber = json.namedObject("line").toInt(m_lineNumber);
m_pattern = json.namedObject("pattern").toString(m_pattern);
m_fileName = json.namedObject("file").toString(m_fileName);
m_len = json.namedObject("len").toInt(m_len);
m_flags = json.namedObject("flags").toSize_t(m_flags);
m_columnInChars = json.namedObject("columnInChars").toInt(m_columnInChars);
m_lenInChars = json.namedObject("lenInChars").toInt(m_lenInChars);
m_regexCaptures = json.namedObject("regexCaptures").toArrayString();
}
JSONItem SearchSummary::ToJSON() const
{
JSONItem json = JSONItem::createObject();
json.addProperty("filesScanned", m_fileScanned);
json.addProperty("matchesFound", m_matchesFound);
json.addProperty("elapsed", m_elapsed);
json.addProperty("failedFiles", m_failedFiles);
json.addProperty("findWhat", m_findWhat);
json.addProperty("replaceWith", m_replaceWith);
return json;
}
void SearchSummary::FromJSON(const JSONItem& json)
{
m_fileScanned = json.namedObject("filesScanned").toInt(m_fileScanned);
m_matchesFound = json.namedObject("matchesFound").toInt(m_matchesFound);
m_elapsed = json.namedObject("elapsed").toInt(m_elapsed);
m_failedFiles = json.namedObject("failedFiles").toArrayString();
m_findWhat = json.namedObject("findWhat").toString();
m_replaceWith = json.namedObject("replaceWith").toString();
}
|