Parent: [e1584c] (diff)

Child: [3109a3] (diff)

Download this file

indexer.cpp    588 lines (530 with data), 17.8 kB

  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
#ifndef lint
static char rcsid[] = "@(#$Id: indexer.cpp,v 1.66 2008-07-28 08:42:52 dockes Exp $ (C) 2004 J.F.Dockes";
#endif
/*
* 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.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the
* Free Software Foundation, Inc.,
* 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*/
#ifdef HAVE_CONFIG_H
#include "autoconfig.h"
#endif
#include <stdio.h>
#include <sys/stat.h>
#include <unistd.h>
#include <errno.h>
#include <cstring>
#include <fnmatch.h>
#include <iostream>
#include <list>
#include <map>
#include <algorithm>
#include "pathut.h"
#include "conftree.h"
#include "rclconfig.h"
#include "fstreewalk.h"
#include "rcldb.h"
#include "readfile.h"
#include "indexer.h"
#include "csguess.h"
#include "transcode.h"
#include "debuglog.h"
#include "internfile.h"
#include "smallut.h"
#include "wipedir.h"
#ifdef RCL_USE_ASPELL
#include "rclaspell.h"
#endif
#ifndef NO_NAMESPACES
using namespace std;
#endif /* NO_NAMESPACES */
#ifndef deleteZ
#define deleteZ(X) {delete X;X = 0;}
#endif
DbIndexer::~DbIndexer() {
// Maybe clean up temporary directory
if (m_tmpdir.length()) {
wipedir(m_tmpdir);
if (rmdir(m_tmpdir.c_str()) < 0) {
LOGERR(("DbIndexer::~DbIndexer: cannot clear temp dir %s\n",
m_tmpdir.c_str()));
}
}
m_db.close();
}
list<string> DbIndexer::getStemmerNames()
{
return Rcl::Db::getStemmerNames();
}
// Index each directory in the topdirs for a given db
bool DbIndexer::indexDb(bool resetbefore, list<string> *topdirs)
{
if (!init(resetbefore))
return false;
if (m_updater) {
m_updater->status.reset();
m_updater->status.dbtotdocs = m_db.docCnt();
}
m_walker.setSkippedPaths(m_config->getSkippedPaths());
for (list<string>::const_iterator it = topdirs->begin();
it != topdirs->end(); it++) {
LOGDEB(("DbIndexer::index: Indexing %s into %s\n", it->c_str(),
m_dbdir.c_str()));
// Set the current directory in config so that subsequent
// getConfParams() will get local values
m_config->setKeyDir(*it);
// Adjust the "follow symlinks" option
bool follow;
if (m_config->getConfParam("followLinks", &follow) && follow) {
m_walker.setOpts(FsTreeWalker::FtwFollow);
} else {
m_walker.setOpts(FsTreeWalker::FtwOptNone);
}
int abslen;
if (m_config->getConfParam("idxabsmlen", &abslen))
m_db.setAbstractParams(abslen, -1, -1);
// Set up skipped patterns for this subtree. This probably should be
// done in the directory change code in processone() instead.
m_walker.setSkippedNames(m_config->getSkippedNames());
// Walk the directory tree
if (m_walker.walk(*it, *this) != FsTreeWalker::FtwOk) {
LOGERR(("DbIndexer::index: error while indexing %s: %s\n",
it->c_str(), m_walker.getReason().c_str()));
return false;
}
}
if (m_updater) {
m_updater->status.fn.erase();
m_updater->status.phase = DbIxStatus::DBIXS_PURGE;
m_updater->update();
}
// Get rid of all database entries that don't exist in the
// filesystem anymore.
m_db.purge();
createStemmingDatabases();
createAspellDict();
// The close would be done in our destructor, but we want status here
if (m_updater) {
m_updater->status.phase = DbIxStatus::DBIXS_CLOSING;
m_updater->status.fn.erase();
m_updater->update();
}
if (!m_db.close()) {
LOGERR(("DbIndexer::index: error closing database in %s\n",
m_dbdir.c_str()));
return false;
}
if (!m_missingExternal.empty()) {
string missing;
stringsToString(m_missingExternal, missing);
LOGINFO(("DbIndexer::index missing helper program(s): %s\n",
missing.c_str()));
}
return true;
}
// Create stemming databases. We also remove those which are not
// configured.
bool DbIndexer::createStemmingDatabases()
{
string slangs;
if (m_config->getConfParam("indexstemminglanguages", slangs)) {
list<string> langs;
stringToStrings(slangs, langs);
// Get the list of existing stem dbs from the database (some may have
// been manually created, we just keep those from the config
list<string> dblangs = m_db.getStemLangs();
list<string>::const_iterator it;
for (it = dblangs.begin(); it != dblangs.end(); it++) {
if (find(langs.begin(), langs.end(), *it) == langs.end())
m_db.deleteStemDb(*it);
}
for (it = langs.begin(); it != langs.end(); it++) {
if (m_updater) {
m_updater->status.phase = DbIxStatus::DBIXS_STEMDB;
m_updater->status.fn = *it;
m_updater->update();
}
m_db.createStemDb(*it);
}
}
return true;
}
bool DbIndexer::init(bool resetbefore, bool rdonly)
{
if (m_tmpdir.empty() || access(m_tmpdir.c_str(), 0) < 0) {
string reason;
if (!maketmpdir(m_tmpdir, reason)) {
LOGERR(("DbIndexer: cannot create temporary directory: %s\n",
reason.c_str()));
return false;
}
}
Rcl::Db::OpenMode mode = rdonly ? Rcl::Db::DbRO :
resetbefore ? Rcl::Db::DbTrunc : Rcl::Db::DbUpd;
if (!m_db.open(m_dbdir, m_config->getStopfile(), mode)) {
LOGERR(("DbIndexer: error opening database in %s\n", m_dbdir.c_str()));
return false;
}
return true;
}
bool DbIndexer::createStemDb(const string &lang)
{
if (!init(false, true))
return false;
return m_db.createStemDb(lang);
}
// The language for the aspell dictionary is handled internally by the aspell
// module, either from a configuration variable or the NLS environment.
bool DbIndexer::createAspellDict()
{
LOGDEB2(("DbIndexer::createAspellDict()\n"));
#ifdef RCL_USE_ASPELL
// For the benefit of the real-time indexer, we only initialize
// noaspell from the configuration once. It can then be set to
// true if dictionary generation fails, which avoids retrying
// it forever.
static int noaspell = -12345;
if (noaspell == -12345) {
noaspell = false;
m_config->getConfParam("noaspell", &noaspell);
}
if (noaspell)
return true;
if (!init(false, true))
return false;
Aspell aspell(m_config);
string reason;
if (!aspell.init(reason)) {
LOGERR(("DbIndexer::createAspellDict: aspell init failed: %s\n",
reason.c_str()));
noaspell = true;
return false;
}
LOGDEB(("DbIndexer::createAspellDict: creating dictionary\n"));
if (!aspell.buildDict(m_db, reason)) {
LOGERR(("DbIndexer::createAspellDict: aspell buildDict failed: %s\n",
reason.c_str()));
noaspell = true;
return false;
}
// The close would be done in our destructor, but we want status here
if (!m_db.close()) {
LOGERR(("DbIndexer::indexfiles: error closing database in %s\n",
m_dbdir.c_str()));
noaspell = true;
return false;
}
#endif
return true;
}
/**
* Index individual files, out of a full tree run. No database purging
*/
bool DbIndexer::indexFiles(const list<string> &filenames)
{
bool called_init = false;
list<string>::const_iterator it;
for (it = filenames.begin(); it != filenames.end(); it++) {
string dir = path_getfather(*it);
m_config->setKeyDir(dir);
int abslen;
if (m_config->getConfParam("idxabsmlen", &abslen))
m_db.setAbstractParams(abslen, -1, -1);
struct stat stb;
if (lstat(it->c_str(), &stb) != 0) {
LOGERR(("DbIndexer::indexFiles: lstat(%s): %s", it->c_str(),
strerror(errno)));
continue;
}
// If we get to indexing directory names one day, will need to test
// against dbdir here to avoid modification loops (with rclmon).
if (!S_ISREG(stb.st_mode)) {
LOGDEB2(("DbIndexer::indexFiles: %s: not a regular file\n",
it->c_str()));
continue;
}
static string lstdir;
static list<string> skpl;
if (lstdir.compare(dir)) {
LOGDEB(("Recomputing list of skipped names\n"));
skpl = m_config->getSkippedNames();
lstdir = dir;
}
if (!skpl.empty()) {
list<string>::const_iterator skit;
string fn = path_getsimple(*it);
for (skit = skpl.begin(); skit != skpl.end(); skit++) {
if (fnmatch(skit->c_str(), fn.c_str(), 0) == 0) {
LOGDEB(("Skipping [%s] :matches skip list\n", fn.c_str()));
goto skipped;
}
}
}
// Defer opening db until really needed.
if (!called_init) {
if (!init())
return false;
called_init = true;
}
if (processone(*it, &stb, FsTreeWalker::FtwRegular) !=
FsTreeWalker::FtwOk) {
LOGERR(("DbIndexer::indexFiles: processone failed\n"));
return false;
}
skipped:
false; // Need a statement here to make compiler happy ??
}
// The close would be done in our destructor, but we want status here
if (!m_db.close()) {
LOGERR(("DbIndexer::indexfiles: error closing database in %s\n",
m_dbdir.c_str()));
return false;
}
return true;
}
/** Purge docs for given files out of the database */
bool DbIndexer::purgeFiles(const list<string> &filenames)
{
if (!init())
return false;
list<string>::const_iterator it;
for (it = filenames.begin(); it != filenames.end(); it++) {
if (!m_db.purgeFile(*it)) {
LOGERR(("DbIndexer::purgeFiles: Database error\n"));
return false;
}
}
// The close would be done in our destructor, but we want status here
if (!m_db.close()) {
LOGERR(("DbIndexer::purgefiles: error closing database in %s\n",
m_dbdir.c_str()));
return false;
}
return true;
}
/// This method gets called for every file and directory found by the
/// tree walker.
///
/// It checks with the db if the file has changed and needs to be
/// reindexed. If so, it calls internfile() which will identify the
/// file type and call an appropriate handler to convert the document into
/// internal format, which we then add to the database.
///
/// Accent and majuscule handling are performed by the db module when doing
/// the actual indexing work. The Rcl::Doc created by internfile()
/// mostly contains pretty raw utf8 data.
FsTreeWalker::Status
DbIndexer::processone(const std::string &fn, const struct stat *stp,
FsTreeWalker::CbFlag flg)
{
if (m_updater && !m_updater->update()) {
return FsTreeWalker::FtwStop;
}
// If we're changing directories, possibly adjust parameters (set
// the current directory in configuration object)
if (flg == FsTreeWalker::FtwDirEnter ||
flg == FsTreeWalker::FtwDirReturn) {
m_config->setKeyDir(fn);
int abslen;
if (m_config->getConfParam("idxabsmlen", &abslen))
m_db.setAbstractParams(abslen, -1, -1);
if (flg == FsTreeWalker::FtwDirReturn)
return FsTreeWalker::FtwOk;
}
// Check db up to date ? Doing this before file type
// identification means that, if usesystemfilecommand is switched
// from on to off it may happen that some files which are now
// without mime type will not be purged from the db, resulting
// in possible 'cannot intern file' messages at query time...
char cbuf[100];
// Document signature
sprintf(cbuf, "%ld%ld", (long)stp->st_size, (long)stp->st_mtime);
string sig = cbuf;
if (!m_db.needUpdate(fn, sig)) {
LOGDEB(("processone: up to date: %s\n", fn.c_str()));
if (m_updater) {
m_updater->status.fn = fn;
if (!m_updater->update()) {
return FsTreeWalker::FtwStop;
}
}
return FsTreeWalker::FtwOk;
}
FileInterner interner(fn, stp, m_config, m_tmpdir);
// File name transcoded to utf8 for indexation.
string charset = m_config->getDefCharset(true);
// If this fails, the file name won't be indexed, no big deal
// Note that we used to do the full path here, but I ended up believing
// that it made more sense to use only the file name
string utf8fn; int ercnt;
if (!transcode(path_getsimple(fn), utf8fn, charset, "UTF-8", &ercnt)) {
LOGERR(("processone: fn transcode failure from [%s] to UTF-8: %s\n",
charset.c_str(), path_getsimple(fn).c_str()));
} else if (ercnt) {
LOGDEB(("processone: fn transcode %d errors from [%s] to UTF-8: %s\n",
ercnt, charset.c_str(), path_getsimple(fn).c_str()));
}
LOGDEB2(("processone: fn transcoded from [%s] to [%s] (%s->%s)\n",
path_getsimple(fn).c_str(), utf8fn.c_str(), charset.c_str(), "UTF-8"));
FileInterner::Status fis = FileInterner::FIAgain;
bool hadNullIpath = false;
Rcl::Doc doc;
const string plus = "+";
char ascdate[20];
sprintf(ascdate, "%ld", long(stp->st_mtime));
while (fis == FileInterner::FIAgain) {
doc.erase();
string ipath;
fis = interner.internfile(doc, ipath);
if (fis == FileInterner::FIError) {
list<string> ext = interner.getMissingExternal();
m_missingExternal.merge(ext);
m_missingExternal.unique();
// We used to return at this point.
//
// The nice side was that if a filter failed because of a
// lacking supporting app, the file would be indexed once
// the app was installed.
//
// The not so nice point was that the file name was not
// indexed.
//
// We now index at least the file name. We use a dirty
// hack to ensure that the indexing will be retried each
// time: the stored number as decimal ascii mtime is
// prefixed with a '+', which doesnt change its value for
// atoll() but is tested by rcldb::needUpdate()
// Reset the date as set by the handler if any
doc.fmtime.erase();
// Go through:
}
if (doc.fmtime.empty()) {
// Set the date if this was not done in the document handler
doc.fmtime = (fis == FileInterner::FIError) ? plus + ascdate :
ascdate;
}
// Internal access path for multi-document files
if (ipath.empty())
hadNullIpath = true;
else
doc.ipath = ipath;
// Note that the filter may have its own idea of the file name
// (ie: mail attachment)
if (doc.utf8fn.empty())
doc.utf8fn = utf8fn;
char cbuf[100];
sprintf(cbuf, "%ld", (long)stp->st_size);
doc.fbytes = cbuf;
// Document signature for up to date checks: concatenate mtime and
// size. Note: looking for changes only, no need to parseback so no
// need for reversible formatting
sprintf(cbuf, "%ld%ld", (long)stp->st_size, (long)stp->st_mtime);
doc.sig = cbuf;
// Add document to database
if (!m_db.add(fn, doc))
return FsTreeWalker::FtwError;
// Tell what we are doing and check for interrupt request
if (m_updater) {
if ((++(m_updater->status.docsdone) % 10) == 0) {
m_updater->status.fn = fn;
if (!ipath.empty())
m_updater->status.fn += "|" + ipath;
if (!m_updater->update()) {
return FsTreeWalker::FtwStop;
}
}
}
}
// If we had no instance with a null ipath, we create an empty
// document to stand for the file itself, to be used mainly for up
// to date checks. Typically this happens for an mbox file.
if (hadNullIpath == false) {
LOGDEB1(("Creating empty doc for file\n"));
Rcl::Doc fileDoc;
fileDoc.fmtime = ascdate;
fileDoc.utf8fn = utf8fn;
fileDoc.mimetype = interner.getMimetype();
char cbuf[100];
sprintf(cbuf, "%ld", (long)stp->st_size);
fileDoc.fbytes = cbuf;
// Document signature for up to date checks.
sprintf(cbuf, "%ld%ld", (long)stp->st_size, (long)stp->st_mtime);
fileDoc.sig = cbuf;
if (!m_db.add(fn, fileDoc))
return FsTreeWalker::FtwError;
}
return FsTreeWalker::FtwOk;
}
////////////////////////////////////////////////////////////////////////////
// ConIndexer methods: ConfIndexer is the top-level object, that can index
// multiple directories to multiple databases.
ConfIndexer::~ConfIndexer()
{
deleteZ(m_dbindexer);
}
bool ConfIndexer::index(bool resetbefore)
{
list<string> tdl = m_config->getTopdirs();
if (tdl.empty()) {
m_reason = "Top directory list (topdirs param.) not found in config"
"or Directory list parse error";
return false;
}
// Each top level directory to be indexed can be associated with a
// different database. We first group the directories by database:
// it is important that all directories for a database be indexed
// at once so that deleted file cleanup works
list<string>::iterator dirit;
map<string, list<string> > dbmap;
map<string, list<string> >::iterator dbit;
for (dirit = tdl.begin(); dirit != tdl.end(); dirit++) {
string dbdir;
string doctopdir = *dirit;
m_config->setKeyDir(doctopdir);
dbdir = m_config->getDbDir();
if (dbdir.empty()) {
LOGERR(("ConfIndexer::index: no database directory in "
"configuration for %s\n", doctopdir.c_str()));
m_reason = "No database directory set for " + doctopdir;
return false;
}
dbit = dbmap.find(dbdir);
if (dbit == dbmap.end()) {
list<string> l;
l.push_back(doctopdir);
dbmap[dbdir] = l;
} else {
dbit->second.push_back(doctopdir);
}
}
m_config->setKeyDir("");
// The dbmap now has dbdir as key and directory lists as values.
// Index each directory group in turn
for (dbit = dbmap.begin(); dbit != dbmap.end(); dbit++) {
//cout << dbit->first << " -> ";
//list<string>::const_iterator dit;
//for (dit = dbit->second.begin(); dit != dbit->second.end(); dit++) {
// cout << *dit << " ";
//}
//cout << endl;
m_dbindexer = new DbIndexer(m_config, dbit->first, m_updater);
if (!m_dbindexer->indexDb(resetbefore, &dbit->second)) {
deleteZ(m_dbindexer);
m_reason = "Failed indexing in " + dbit->first;
return false;
}
deleteZ(m_dbindexer);
}
return true;
}