File: pa_common.C

package info (click to toggle)
parser 3.4.5-4
  • links: PTS, VCS
  • area: main
  • in suites: buster
  • size: 7,552 kB
  • sloc: cpp: 32,375; sh: 11,487; ansic: 10,849; yacc: 1,361; makefile: 248; awk: 5
file content (1495 lines) | stat: -rw-r--r-- 41,303 bytes parent folder | download
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
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
/** @file
	Parser: commonly functions.

	Copyright (c) 2000-2017 Art. Lebedev Studio (http://www.artlebedev.com)
	Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)

 * BASE64 part
 *  Authors: Michael Zucchi <notzed@ximian.com>
 *           Jeffrey Stedfast <fejj@ximian.com>
 *
 *  Copyright 2000-2004 Ximian, Inc. (www.ximian.com)
 *
 *  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 Street #330, Boston, MA 02111-1307, USA.
 *
 */

#include "pa_common.h"
#include "pa_exception.h"
#include "pa_hash.h"
#include "pa_globals.h"
#include "pa_charsets.h"
#include "pa_http.h"
#include "pa_request_charsets.h"
#include "pcre.h"
#include "pa_request.h"

#include "pa_idna.h"
#include "pa_convert_utf.h"

#ifdef _MSC_VER
#include <windows.h>
#include <direct.h>
#endif

#ifdef _MSC_VER
#define pa_mkdir(path, mode) _mkdir(path)
#else
#define pa_mkdir(path, mode) mkdir(path, mode)
#endif

volatile const char * IDENT_PA_COMMON_C="$Id: pa_common.C,v 1.303 2017/02/07 22:00:41 moko Exp $" IDENT_PA_COMMON_H IDENT_PA_HASH_H IDENT_PA_ARRAY_H IDENT_PA_STACK_H; 

// some maybe-undefined constants

#ifndef _O_TEXT
#	define _O_TEXT 0
#endif
#ifndef _O_BINARY
#	define _O_BINARY 0
#endif

#ifdef HAVE_FTRUNCATE
#	define PA_O_TRUNC 0
#else
#	ifdef _O_TRUNC
#		define PA_O_TRUNC _O_TRUNC
#	else
#		error you must have either ftruncate function or _O_TRUNC bit declared
#	endif
#endif

// defines for globals

#define FILE_STATUS_NAME  "status"

// globals

const String file_status_name(FILE_STATUS_NAME);

// forwards

const UTF16* pa_utf16_encode(const char* in, Charset& source_charset);

// functions

#ifdef _MSC_VER

int pa_stat(const char *pathname, struct stat *buffer){
	const UTF16* utf16name=pa_utf16_encode(pathname, pa_thread_request().charsets.source());
	return _wstat64((const wchar_t *)utf16name, buffer);
}

int pa_open(const char *pathname, int flags, int mode){
	const UTF16* utf16name=pa_utf16_encode(pathname, pa_thread_request().charsets.source());
	return _wopen((const wchar_t *)utf16name, flags, mode);
}

FILE *pa_fopen(const char *pathname, const char *mode){
	const UTF16* utf16name=pa_utf16_encode(pathname, pa_thread_request().charsets.source());
	const UTF16* utf16mode=pa_utf16_encode(mode, pa_thread_request().charsets.source());
	return _wfopen((const wchar_t *)utf16name, (const wchar_t *)utf16mode);
}

#endif

char* file_read_text(Request_charsets& charsets, const String& file_spec, bool fail_on_read_problem, HashStringValue* params, bool transcode_result) {
	File_read_result file=file_read(charsets, file_spec, true, params, fail_on_read_problem, 0, 0, 0, transcode_result);
	return file.success?file.str:0;
}

char* file_load_text(Request& r, const String& file_spec, bool fail_on_read_problem, HashStringValue* params, bool transcode_result) {
	File_read_result file=file_load(r, file_spec, true, params, fail_on_read_problem, 0, 0, 0, transcode_result);
	return file.success?file.str:0;
}

/// these options were handled but not checked elsewhere, now check them
int pa_get_valid_file_options_count(HashStringValue& options) {
	int result=0;
	if(options.get(PA_SQL_LIMIT_NAME))
		result++;
	if(options.get(PA_SQL_OFFSET_NAME))
		result++;
	if(options.get(PA_COLUMN_SEPARATOR_NAME))
		result++;
	if(options.get(PA_COLUMN_ENCLOSER_NAME))
		result++;
	if(options.get(PA_CHARSET_NAME))
		result++;
	return result;
}

#ifndef DOXYGEN
struct File_read_action_info {
	char **data; size_t *data_size;
	char* buf; size_t offset; size_t count;
}; 
#endif

static void file_read_action(struct stat& finfo, int f, const String& file_spec, void *context) {
	File_read_action_info& info=*static_cast<File_read_action_info *>(context); 
	size_t to_read_size=info.count;
	if(!to_read_size)
		to_read_size=check_file_size(finfo.st_size, file_spec);
	if(to_read_size) {
		if(info.offset)
			lseek(f, info.offset, SEEK_SET);
		*info.data=info.buf ? info.buf : (char *)pa_malloc_atomic(to_read_size+1);
		ssize_t result=read(f, *info.data, to_read_size);
		if(result<0)
			throw Exception("file.read", &file_spec, "read failed: %s (%d)", strerror(errno), errno);
		*info.data_size=result;
	} else { // empty file
		// for both, text and binary: for text we need that terminator, for binary we need nonzero pointer to be able to save such files
		*info.data=(char *)pa_malloc_atomic(1);
		*(char*)(*info.data)=0;
		*info.data_size=0;
		return;
	}
}

File_read_result file_read(Request_charsets& charsets, const String& file_spec, 
			bool as_text, HashStringValue *params,
			bool fail_on_read_problem,
			char* buf, size_t offset, size_t count, bool transcode_text_result) {
	File_read_result result={false, 0, 0, 0};
	if(params){
		int valid_options=pa_get_valid_file_options_count(*params);
		if(valid_options!=params->count())
			throw Exception(PARSER_RUNTIME, 0, CALLED_WITH_INVALID_OPTION);
	}

	File_read_action_info info={&result.str, &result.length, buf, offset, count}; 

	result.success=file_read_action_under_lock(file_spec, "read", file_read_action, &info, as_text, fail_on_read_problem); 

	if(as_text){
		if(result.success){
			Charset* asked_charset=0;
			if(params)
				if(Value* vcharset_name=params->get(PA_CHARSET_NAME))
					asked_charset=&pa_charsets.get(vcharset_name->as_string());

			asked_charset=pa_charsets.checkBOM(result.str, result.length, asked_charset);

			if(result.length && transcode_text_result && asked_charset){ // length must be checked because transcode returns CONST string in case length==0, which contradicts hacking few lines below
				String::C body=String::C(result.str, result.length);
				body=Charset::transcode(body, *asked_charset, charsets.source());

				result.str=const_cast<char*>(body.str); // hacking a little
				result.length=body.length;
			}
		}
		if(result.length)
			fix_line_breaks(result.str, result.length);
	}

	return result;
}

File_read_result file_load(Request& r, const String& file_spec, 
			bool as_text, HashStringValue *params,
			bool fail_on_read_problem,
			char* buf, size_t offset, size_t count, bool transcode_text_result) {

	File_read_result result={false, 0, 0, 0};
	if(file_spec.starts_with("http://")) {
		if(offset || count)
			throw Exception(PARSER_RUNTIME, 0, "offset and load options are not supported for HTTP:// file load");

		// fail on read problem
		File_read_http_result http=pa_internal_file_read_http(r, file_spec, as_text, params, transcode_text_result);
		result.success=true;
		result.str=http.str;
		result.length=http.length;
		result.headers=http.headers; 
	} else
		result=
			file_read(r.charsets, file_spec, as_text, params, fail_on_read_problem, buf, offset, count, transcode_text_result);

	return result;
}


#ifdef PA_SAFE_MODE 
void check_safe_mode(struct stat finfo, const String& file_spec, const char* fname) {
	if(finfo.st_uid/*foreign?*/!=geteuid() 
		&& finfo.st_gid/*foreign?*/!=getegid()) 
		throw Exception(PARSER_RUNTIME,
			&file_spec,
			"parser is in safe mode: reading files of foreign group and user disabled "
			"[recompile parser with --disable-safe-mode configure option], "
			"actual filename '%s', fuid(%d)!=euid(%d) or fgid(%d)!=egid(%d)",
			fname, finfo.st_uid, geteuid(), finfo.st_gid, getegid()
		);
}
#else
void check_safe_mode(struct stat, const String&, const char*) {
}
#endif



bool file_read_action_under_lock(const String& file_spec, 
				const char* action_name, File_read_action action, void *context, 
				bool as_text, 
				bool fail_on_read_problem) {
	const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); 
	int f;

	// first open, next stat:
	// directory update of NTFS hard links performed on open.
	// ex: 
	//   a.html:^test[] and b.html hardlink to a.html
	//   user inserts ! before ^test in a.html
	//   directory entry of b.html in NTFS not updated at once, 
	//   they delay update till open, so we would receive "!^test[" string
	//   if would do stat, next open.
	// later: it seems, even this does not help sometimes
	if((f=pa_open(fname, O_RDONLY|(as_text?_O_TEXT:_O_BINARY)))>=0) {
		try {
			if(pa_lock_shared_blocking(f)!=0)
				throw Exception("file.lock", &file_spec, "shared lock failed: %s (%d), actual filename '%s'", strerror(errno), errno, fname);

			struct stat finfo;
			if(pa_fstat(f, &finfo)!=0)
				throw Exception("file.missing", // hardly possible: we just opened it OK
					&file_spec, "stat failed: %s (%d), actual filename '%s'", strerror(errno), errno, fname);

			check_safe_mode(finfo, file_spec, fname);

			action(finfo, f, file_spec, context); 
		} catch(...) {
			pa_unlock(f);close(f); 
			if(fail_on_read_problem)
				rethrow;
			return false;
		} 

		pa_unlock(f);close(f); 
		return true;
	} else {
		if(fail_on_read_problem)
			throw Exception(errno==EACCES ? "file.access" : (errno==ENOENT || errno==ENOTDIR || errno==ENODEV) ? "file.missing" : 0,
				&file_spec, "%s failed: %s (%d), actual filename '%s'", action_name, strerror(errno), errno, fname);
		return false;
	}
}

void create_dir_for_file(const String& file_spec) {
	size_t pos_after=1;
	size_t pos_before;
	while((pos_before=file_spec.pos('/', pos_after))!=STRING_NOT_FOUND) {
		pa_mkdir(file_spec.mid(0, pos_before).taint_cstr(String::L_FILE_SPEC), 0775); 
		pos_after=pos_before+1;
	}
}

bool file_write_action_under_lock(
				const String& file_spec, 
				const char* action_name,
				File_write_action action,
				void *context, 
				bool as_text, 
				bool do_append, 
				bool do_block, 
				bool fail_on_lock_problem) {
	const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); 
	int f;
	if(access(fname, W_OK)!=0) // no
		create_dir_for_file(file_spec); 

	if((f=pa_open(fname, 
		O_CREAT|O_RDWR
		|(as_text?_O_TEXT:_O_BINARY)
		|(do_append?O_APPEND:PA_O_TRUNC), 0664))>=0) {
		if((do_block?pa_lock_exclusive_blocking(f):pa_lock_exclusive_nonblocking(f))!=0) {
			Exception e("file.lock", &file_spec, "shared lock failed: %s (%d), actual filename '%s'", strerror(errno), errno, fname);
			close(f); 
			if(fail_on_lock_problem)
				throw e;
			return false;
		}

		try {
#if (defined(HAVE_FCHMOD) && defined(PA_SAFE_MODE))
			struct stat finfo;
			if(pa_fstat(f, &finfo)==0 && finfo.st_mode & 0111)
				fchmod(f, finfo.st_mode & 0666/*clear executable bits*/); // backward: ignore errors if any
#endif
			action(f, context);
		} catch(...) {
#ifdef HAVE_FTRUNCATE
			if(!do_append)
				ftruncate(f, lseek(f, 0, SEEK_CUR)); // one can not use O_TRUNC, read lower
#endif
			pa_unlock(f);close(f); 
			rethrow;
		}
		
#ifdef HAVE_FTRUNCATE
		if(!do_append)
			ftruncate(f, lseek(f, 0, SEEK_CUR)); // O_TRUNC truncates even exclusevely write-locked file [thanks to Igor Milyakov <virtan@rotabanner.com> for discovering]
#endif
		pa_unlock(f);close(f); 
		return true;
	} else
		throw Exception(errno==EACCES ? "file.access" : 0, &file_spec, "%s failed: %s (%d), actual filename '%s'", action_name, strerror(errno), errno, fname);
	// here should be nothing, see rethrow above
}

#ifndef DOXYGEN
struct File_write_action_info {
	const char* str;
	size_t length;
}; 
#endif

static void file_write_action(int f, void *context) {
	File_write_action_info& info=*static_cast<File_write_action_info *>(context); 
	if(info.length) {
		ssize_t written=write(f, info.str, info.length); 
		if(written<0)
			throw Exception("file.write", 0, "write failed: %s (%d)",  strerror(errno), errno); 
		if((size_t)written!=info.length)
			throw Exception("file.write", 0, "write failed: %u of %u bytes written", written, info.length);
	}
}

void file_write(
				Request_charsets& charsets,
				const String& file_spec,
				const char* data,
				size_t size, 
				bool as_text, 
				bool do_append,
				Charset* asked_charset) {

	if(as_text && asked_charset){
		String::C body=String::C(data, size);
		body=Charset::transcode(body, charsets.source(), *asked_charset);
		data=body.str;
		size=body.length;
	};

	File_write_action_info info={data, size}; 

	file_write_action_under_lock(
		file_spec, 
		"write",
		file_write_action,
		&info, 
		as_text, 
		do_append); 
}

static size_t get_dir(char* fname, size_t helper_length){
	bool dir=false;
	size_t pos=0;
	for(pos=helper_length; pos; pos--){
		char c=fname[pos-1];
		if(c=='/' || c=='\\'){
			fname[pos-1]=0;
			dir=true;
		} else if(dir) break;
	}
	return pos;
}

static bool entry_readable(char* fname, bool need_dir) {
	if(need_dir){
		size_t size=strlen(fname);
		while(size) {
			char c=fname[size-1];
			if(c=='/' || c=='\\')
				fname[--size]=0;
			else
				break;
		}
	}

	struct stat finfo;
	if(access(fname, R_OK)==0 && entry_exists(fname, &finfo)) {
		bool is_dir=(finfo.st_mode&S_IFDIR) != 0;
		return is_dir==need_dir;
	}
	return false;
}

static bool entry_readable(const String& file_spec, bool need_dir) {
	return entry_readable(file_spec.taint_cstrm(String::L_FILE_SPEC), need_dir);
}

// throws nothing! [this is required in file_move & file_delete]
static void rmdir(const String& file_spec, size_t pos_after) {
	char* dir_spec=file_spec.taint_cstrm(String::L_FILE_SPEC);
	size_t length=strlen(dir_spec);
	while( (length=get_dir(dir_spec, length)) && (length > pos_after) ){
#ifdef _MSC_VER
		if(!entry_readable(dir_spec, true))
			break;
		DWORD attrs=GetFileAttributes(dir_spec);
		if(
			(attrs==INVALID_FILE_ATTRIBUTES)
			|| !(attrs & FILE_ATTRIBUTE_DIRECTORY)
			|| (attrs & FILE_ATTRIBUTE_REPARSE_POINT)
		)
			break;
#endif
		if( rmdir(dir_spec) )
			break;
	};
}

bool file_delete(const String& file_spec, bool fail_on_problem, bool keep_empty_dirs) {
	const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); 
	if(unlink(fname)!=0) {
		if(fail_on_problem)
			throw Exception(errno==EACCES?"file.access":errno==ENOENT?"file.missing":0,
				&file_spec, "unlink failed: %s (%d), actual filename '%s'", strerror(errno), errno, fname);
		else
			return false;
	}

	if(!keep_empty_dirs)
		rmdir(file_spec, 1); 

	return true;
}

void file_move(const String& old_spec, const String& new_spec, bool keep_empty_dirs) {
	const char* old_spec_cstr=old_spec.taint_cstr(String::L_FILE_SPEC); 
	const char* new_spec_cstr=new_spec.taint_cstr(String::L_FILE_SPEC); 
	
	create_dir_for_file(new_spec); 

	if(rename(old_spec_cstr, new_spec_cstr)!=0)
		throw Exception(errno==EACCES ? "file.access" : errno==ENOENT ? "file.missing" : 0,
			&old_spec, "rename failed: %s (%d), actual filename '%s' to '%s'", strerror(errno), errno, old_spec_cstr, new_spec_cstr);

	if(!keep_empty_dirs)
		rmdir(old_spec, 1); 
}


bool entry_exists(const char* fname, struct stat *afinfo) {
	struct stat lfinfo;
	bool result=pa_stat(fname, &lfinfo)==0;
	if(afinfo)
		*afinfo=lfinfo;
	return result;
}

bool entry_exists(const String& file_spec) {
	const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); 
	return entry_exists(fname, 0); 
}

bool file_exist(const String& file_spec) {
	return entry_readable(file_spec, false); 
}

bool dir_exists(const String& file_spec) {
	return entry_readable(file_spec, true); 
}

const String* file_exist(const String& path, const String& name) {
	String& result=*new String(path);
	if(path.last_char() != '/')
		result << "/"; 
	result << name;
	return file_exist(result)?&result:0;
}

bool file_executable(const String& file_spec) {
	return access(file_spec.taint_cstr(String::L_FILE_SPEC), X_OK)==0;
}

bool file_stat(const String& file_spec, uint64_t& rsize, time_t& ratime, time_t& rmtime, time_t& rctime, bool fail_on_read_problem) {
	const char* fname=file_spec.taint_cstr(String::L_FILE_SPEC); 
	struct stat finfo;
	if(pa_stat(fname, &finfo)!=0) {
		if(fail_on_read_problem)
			throw Exception("file.missing", &file_spec, "getting file size failed: %s (%d), real filename '%s'", strerror(errno), errno, fname);
		else
			return false;
	}
	rsize=finfo.st_size;
	ratime=(time_t)finfo.st_atime;
	rmtime=(time_t)finfo.st_mtime;
	rctime=(time_t)finfo.st_ctime;
	return true;
}

size_t check_file_size(uint64_t size, const String& file_spec){
	if(size > pa_file_size_limit)
		throw Exception(PARSER_RUNTIME, &file_spec, "content size of %.15g bytes exceeds the limit (%.15g bytes)", (double)size, (double)pa_file_size_limit);
	return (size_t)size;
}

/** 
	String related functions
*/

bool capitalized(const char* s){
	bool upper=true;
	for(const char* c=s; *c; c++){
		if(*c != (upper ? toupper((unsigned char)*c) : tolower((unsigned char)*c)))
			return false;
		upper=strchr("-_ ", *c) != 0;
	}
	return true;
}

const char* capitalize(const char* s){
	if(!s || capitalized(s))
		return s;

	char* result=pa_strdup(s);
	if(result){
		bool upper=true;
		for(char* c=result; *c; c++){
			*c=upper ? (char)toupper((unsigned char)*c) : (char)tolower((unsigned char)*c);
			upper=strchr("-_ ", *c) != 0;
		}
	}
	return (const char*)result;
}

char *str_lower(const char *s, size_t helper_length){
	char *result=pa_strdup(s, helper_length);
	for(char* c=result; *c; c++)
		*c=(char)tolower((unsigned char)*c);
	return result;
}

char *str_upper(const char *s, size_t helper_length){
	char *result=pa_strdup(s, helper_length);
	for(char* c=result; *c; c++)
		*c=(char)toupper((unsigned char)*c);
	return result;
}

void fix_line_breaks(char *str, size_t& length) {
	//_asm int 3;
	const char* const eob=str+length;
	char* dest=str;
	// fix DOS: \r\n -> \n
	// fix Macintosh: \r -> \n
	char* bol=str;
	while(char* eol=(char*)memchr(bol, '\r', eob -bol)) {
		size_t len=eol-bol;
		if(dest!=bol)
			memmove(dest, bol, len); 
		dest+=len;
		*dest++='\n'; 

		if(&eol[1]<eob && eol[1]=='\n') { // \r, \n = DOS
			bol=eol+2;
			length--; 
		} else // \r, not \n = Macintosh
			bol=eol+1;
	}
	// last piece without \r
	if(dest!=bol)
		memmove(dest, bol, eob-bol); 
	str[length]=0; // terminating
}

/**
	scans for @a delim[default \n] in @a *row_ref, 
	@return piece of line before it or end of string, if no @a delim found
	assigns @a *row_ref to point right after delimiter if there were one
	or to zero if no @a delim were found.
*/

char* getrow(char* *row_ref, char delim) {
	char* result=*row_ref;
	if(result) {
		*row_ref=strchr(result, delim); 
		if(*row_ref) 
			*((*row_ref)++)=0; 
		else if(!*result) 
			return 0;
	}
	return result;
}

char* lsplit(char* string, char delim) {
	if(string) {
		char* v=strchr(string, delim); 
		if(v) {
			*v=0;
			return v+1;
		}
	}
	return 0;
}

char* lsplit(char* *string_ref, char delim) {
	char* result=*string_ref;
	char* next=lsplit(*string_ref, delim); 
	*string_ref=next;
	return result;
}

char* rsplit(char* string, char delim) {
	if(string) {
		char* v=strrchr(string, delim); 
		if(v) {
			*v=0;
			return v+1;
		}
	}
	return NULL;	
}


// format: %[flags][width][.precision]type	http://msdn.microsoft.com/ru-ru/library/56e442dc(en-us,VS.80).aspx
//		flags: '-', '+', ' ', '#', '0'		http://msdn.microsoft.com/ru-ru/library/8aky45ct(en-us,VS.80).aspx
//		width, precision: non negative decimal number
enum FormatType {
	FormatInvalid,
	FormatInt,
	FormatUInt,
	FormatDouble
};
FormatType format_type(const char* fmt){
	enum FormatState {
		Percent, 
		Flags, 
		Width,
		Precision,
		Done
	} state=Percent;

	FormatType result=FormatInvalid;

	const char* pos=fmt;
	while(char c=*(pos++)){
		switch(state){
			case Percent:
				if(c=='%'){
					state=Flags;
				} else {
					return FormatInvalid; // 1st char must be '%' only
				}
				break;
			case Flags:
				if(strchr("-+ #0", c)!=0){
					break;
				}
				// go to the next step
			case Width:
				if(c=='.'){
					state=Precision;
					break;
				}
				// go to the next step
			case Precision:
				if(c>='0' && c<='9'){
					if(state == Flags) state=Width; // no more flags
					break;
				} else if(c=='d' || c=='i'){
					result=FormatInt;
				} else if(strchr("feEgG", c)!=0){
					result=FormatDouble;
				} else if(strchr("uoxX", c)!=0){
					result=FormatUInt;
				} else {
					return FormatInvalid; // invalid char
				}
				state=Done;
				break;
			case Done:
				return FormatInvalid; // no chars allowed after 'type'
		}
	}
	return result;
}


const char* format(double value, const char* fmt) {
	char local_buf[MAX_NUMBER];
	int size=-1;

	if(fmt && strlen(fmt)){
		switch(format_type(fmt)){
			case FormatDouble:
				size=snprintf(local_buf, sizeof(local_buf), fmt, value); 
				break;
			case FormatInt:
				size=snprintf(local_buf, sizeof(local_buf), fmt, (int)value); 
				break;
			case FormatUInt:
				size=snprintf(local_buf, sizeof(local_buf), fmt, (uint)value); 
				break;
			case FormatInvalid:
				throw Exception(PARSER_RUNTIME, 0, "Incorrect format string '%s' was specified.", fmt);
		}
	} else
		size=snprintf(local_buf, sizeof(local_buf), "%d", (int)value);

	if(size < 0 || size >= MAX_NUMBER-1){ // on win32 we manually reduce max size while printing
		throw Exception(PARSER_RUNTIME, 0, "Error occur white executing snprintf with format string '%s'.", fmt);
	}

	return pa_strdup(local_buf, (size_t)size);
}

size_t stdout_write(const void *buf, size_t size) {
#ifdef WIN32
	size_t to_write = size;
	do{
		int chunk_written=fwrite(buf, 1, min((size_t)8*0x400, size), stdout); 
		if(chunk_written<=0)
			break;
		size-=chunk_written;
		buf=((const char*)buf)+chunk_written;
	} while(size>0); 

	return to_write-size;
#else
	return fwrite(buf, 1, size, stdout); 
#endif
}

enum EscapeState {
	EscapeRest, 
	EscapeFirst, 
	EscapeSecond,
	EscapeUnicode
};

// @todo prescan for reduce required size (unescaped sting in 1 byte charset requires less memory usually)
char* unescape_chars(const char* cp, int len, Charset* charset, bool js){
	char* s=new(PointerFreeGC) char[len+1]; // must be enough (%uXXXX==6 bytes, max utf-8 char length==6 bytes)
	char* dst=s;
	EscapeState escapeState=EscapeRest;
	uint escapedValue=0;
	int srcPos=0;
	short int jsCnt=0;
	while(srcPos<len){
		uchar c=(uchar)cp[srcPos]; 
		if(c=='%' || (c=='\\' && js)){
			escapeState=EscapeFirst;
		} else {
			switch(escapeState) {
				case EscapeRest:
					if(c=='+' && !js){
						*dst++=' ';
					} else {
						*dst++=c;
					}
					break;
				case EscapeFirst:
					if(charset && c=='u'){
						// escaped unicode value: %u0430
						jsCnt=0;
						escapedValue=0;
						escapeState=EscapeUnicode;
					} else {
						if(isxdigit(c)){
							escapedValue=hex_value[c] << 4;
							escapeState=EscapeSecond;
						} else {
							*dst++=c;
							escapeState=EscapeRest;
						}
					}
					break;
				case EscapeSecond:
					if(isxdigit(c)){
						escapedValue+=hex_value[c]; 
						*dst++=(char)escapedValue;
					}
					escapeState=EscapeRest;
					break;
				case EscapeUnicode:
					if(isxdigit(c)){
						escapedValue=(escapedValue << 4) + hex_value[c];
						if(++jsCnt==4){
							// transcode utf8 char to client charset (we can lost some chars here)
							charset->store_Char((XMLByte*&)dst, (XMLCh)escapedValue, '?');
							escapeState=EscapeRest;
						}
					} else {
						// not full unicode value
						escapeState=EscapeRest;
					}
					break;
			}
		}

		srcPos++;
	}

	*dst=0; // zero-termination
	return s;
}

char *search_stop(char*& current, char cstop_at) {
	// sanity check
	if(!current)
		return 0;

	// skip leading WS
	while(*current==' ' || *current=='\t')
		current++;
	if(!*current)
		return current=0;

	char *result=current;
	if(char *pstop_at=strchr(current, cstop_at)) {
		*pstop_at=0;
		current=pstop_at+1;
	} else
		current=0;
	return result;
}

#ifdef WIN32
void back_slashes_to_slashes(char* s) {
	if(s)
		for(; *s; s++)
			if(*s=='\\')
				*s='/'; 
}
#endif

size_t strpos(const char *str, const char *substr) {
	const char *p = strstr(str, substr);
	return (p==0)?STRING_NOT_FOUND:p-str;
}

int remove_crlf(char* start, char* end) {
	char* from=start;
	char* to=start;
	bool skip=false;
	while(from < end){
		switch(*from){
			case '\n':
			case '\r':
			case '\t':
			case ' ':
				if(!skip){
					*to=' ';
					to++;
					skip=true;
				}
				break;
			default:
				if(from != to)
					*to=*from;
				to++;
				skip=false;
		}
		from++;
	}
	return to-start;
}

const char* hex_digits="0123456789ABCDEF";

const char* hex_string(unsigned char* bytes, size_t size, bool upcase) {
	char *bytes_hex=new(PointerFreeGC) char [size*2/*byte->hh*/+1/*for zero-teminator*/];
	unsigned char *src=bytes;
	unsigned char *end=bytes+size;
	char *dest=bytes_hex;

	const char *hex=upcase? hex_digits : "0123456789abcdef";

	for(; src<end; src++) {
		*dest++=hex[*src/0x10];
		*dest++=hex[*src%0x10];
	}
	*dest=0;

	return bytes_hex;
}

/* mime64 functions are from libgmime[http://spruce.sourceforge.net/gmime/] lib */
/*
 *  Authors: Michael Zucchi <notzed@helixcode.com>
 *           Jeffrey Stedfast <fejj@helixcode.com>
 *
 *  Copyright 2000 Helix Code, Inc. (www.helixcode.com)
 *
 *  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 Street #330, Boston, MA 02111-1307, USA.
 *
 */
static const char *base64_alphabet =
	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";

/**
 * g_mime_utils_base64_encode_step:
 * @in: input stream
 * @inlen: length of the input
 * @out: output string
 * @state: holds the number of bits that are stored in @save
 * @save: leftover bits that have not yet been encoded
 *
 * Base64 encodes a chunk of data. Performs an 'encode step', only
 * encodes blocks of 3 characters to the output at a time, saves
 * left-over state in state and save (initialise to 0 on first
 * invocation).
 *
 * Returns the number of bytes encoded.
 **/

#define BASE64_GROUPS_IN_LINE 19

static size_t
g_mime_utils_base64_encode_step (const unsigned char *in, size_t inlen, unsigned char *out, int *state, int *save)
{
	register const unsigned char *inptr;
	register unsigned char *outptr;
	
	if (inlen <= 0)
		return 0;
	
	inptr = in;
	outptr = out;
	
	if (inlen + ((unsigned char *)save)[0] > 2) {
		const unsigned char *inend = in + inlen - 2;
		register int c1 = 0, c2 = 0, c3 = 0;
		register int already;
		
		already = *state;
		
		switch (((char *)save)[0]) {
		case 1:	c1 = ((unsigned char *)save)[1]; goto skip1;
		case 2:	c1 = ((unsigned char *)save)[1];
			c2 = ((unsigned char *)save)[2]; goto skip2;
		}
		
		/* yes, we jump into the loop, no i'm not going to change it, its beautiful! */
		while (inptr < inend) {
			c1 = *inptr++;
		skip1:
			c2 = *inptr++;
		skip2:
			c3 = *inptr++;
			*outptr++ = base64_alphabet [c1 >> 2];
			*outptr++ = base64_alphabet [(c2 >> 4) | ((c1 & 0x3) << 4)];
			*outptr++ = base64_alphabet [((c2 & 0x0f) << 2) | (c3 >> 6)];
			*outptr++ = base64_alphabet [c3 & 0x3f];
			/* this is a bit ugly ... */
			if ((++already) >= BASE64_GROUPS_IN_LINE) {
				*outptr++ = '\n';
				already = 0;
			}
		}
		
		((unsigned char *)save)[0] = 0;
		inlen = 2 - (inptr - inend);
		*state = already;
	}
	
	//d(printf ("state = %d, inlen = %d\n", (int)((char *)save)[0], inlen));
	
	if (inlen > 0) {
		register char *saveout;
		
		/* points to the slot for the next char to save */
		saveout = & (((char *)save)[1]) + ((char *)save)[0];
		
		/* inlen can only be 0 1 or 2 */
		switch (inlen) {
		case 2:	*saveout++ = *inptr++;
		case 1:	*saveout++ = *inptr++;
		}
		*(char *)save = *(char *)save+(char)inlen;
	}
	
	/*d(printf ("mode = %d\nc1 = %c\nc2 = %c\n",
		  (int)((char *)save)[0],
		  (int)((char *)save)[1],
		  (int)((char *)save)[2]));*/
	
	return (outptr - out);
}

/**
 * g_mime_utils_base64_encode_close:
 * @in: input stream
 * @inlen: length of the input
 * @out: output string
 * @state: holds the number of bits that are stored in @save
 * @save: leftover bits that have not yet been encoded
 *
 * Base64 encodes the input stream to the output stream. Call this
 * when finished encoding data with g_mime_utils_base64_encode_step to
 * flush off the last little bit.
 *
 * Returns the number of bytes encoded.
 **/
static size_t
g_mime_utils_base64_encode_close (const unsigned char *in, size_t inlen, unsigned char *out, int *state, int *save)
{
	unsigned char *outptr = out;
	int c1, c2;
	
	if (inlen > 0)
		outptr += g_mime_utils_base64_encode_step (in, inlen, outptr, state, save);
	
	c1 = ((unsigned char *)save)[1];
	c2 = ((unsigned char *)save)[2];
	
	switch (((unsigned char *)save)[0]) {
	case 2:
		outptr[2] = base64_alphabet [(c2 & 0x0f) << 2];
		goto skip;
	case 1:
		outptr[2] = '=';
	skip:
		outptr[0] = base64_alphabet [c1 >> 2];
		outptr[1] = base64_alphabet [c2 >> 4 | ((c1 & 0x3) << 4)];
		outptr[3] = '=';
		outptr += 4;
		break;
	}
	
	*outptr++ = 0;
	
	*save = 0;
	*state = 0;
	
	return (outptr - out);
}

static unsigned char gmime_base64_rank[256] = {
	255,255,255,255,255,255,255,255,255,254,254,255,255,254,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	254,255,255,255,255,255,255,255,255,255,255, 62,255,255,255, 63,
	 52, 53, 54, 55, 56, 57, 58, 59, 60, 61,255,255,255,  0,255,255,
	255,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14,
	 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,255,255,255,255,255,
	255, 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, 51,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
	255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,
};

/**
 * g_mime_utils_base64_decode_step:
 * @in: input stream
 * @inlen: max length of data to decode
 * @out: output stream
 * @state: holds the number of bits that are stored in @save
 * @save: leftover bits that have not yet been decoded
 * @strict: only base64 and whitespace chars are allowed
 *
 * Decodes a chunk of base64 encoded data.
 *
 * Returns the number of bytes decoded (which have been dumped in @out).
 **/
size_t
g_mime_utils_base64_decode_step(const unsigned char *in, size_t inlen, unsigned char *out, int *state, int *save, bool strict=false)
{
	const unsigned char *inptr;
	unsigned char *outptr;
	const unsigned char *inend;
	int saved;
	unsigned char c;
	int i;
	
	inend = in + inlen;
	outptr = out;
	
	/* convert 4 base64 bytes to 3 normal bytes */
	saved = *save;
	i = *state;
	inptr = in;
	while (inptr < inend) {
		c = gmime_base64_rank[*inptr++];
		switch(c) {
			case 0xff: // non-base64 and non-whitespace chars. not allowed in strict mode
				if(strict)
					throw Exception(BASE64_FORMAT, 0, "Invalid base64 char on position %d is detected", inptr-in-1);
			case 0xfe: // whitespace chars 0x09, 0x0A, 0x0D, 0x20 are allowed in any mode
				break;
			default:
				saved = (saved << 6) | c;
				i++;
				if (i == 4) {
					*outptr++ = (unsigned char)(saved >> 16);
					*outptr++ = (unsigned char)(saved >> 8);
					*outptr++ = (unsigned char)(saved);
					i = 0;
				}
		}
	}
	
	*save = saved;
	*state = i;
	
	/* quick scan back for '=' on the end somewhere */
	/* fortunately we can drop 1 output char for each trailing = (upto 2) */
	i = 2;
	while (inptr > in && i) {
		inptr--;
		if (gmime_base64_rank[*inptr] <= 0xfe) {
			if (*inptr == '=' && outptr > out)
				outptr--;
			i--;
		}
	}
	
	/* if i != 0 then there is a truncation error! */
	return (outptr - out);
}


char* pa_base64_encode(const char *in, size_t in_size){
	size_t new_size = ((in_size / 3 + 1) * 4);
	new_size += new_size / (BASE64_GROUPS_IN_LINE * 4)/*new lines*/ + 1/*zero terminator*/;
	char* result = new(PointerFreeGC) char[new_size];
	int state=0;
	int save=0;
#ifndef NDEBUG
	size_t filled=
#endif
		g_mime_utils_base64_encode_close ((const unsigned char*)in, in_size, (unsigned char*)result, &state, &save);

	//throw Exception(PARSER_RUNTIME, 0, "%d %d %d", in_size, new_size, filled);
	assert(filled <= new_size);

	return result;
}

struct File_base64_action_info {
	unsigned char** base64;
}; 

static void file_base64_file_action(struct stat& finfo, int f, const String& file_spec, void *context) {

	if(finfo.st_size) { 
		File_base64_action_info& info=*static_cast<File_base64_action_info *>(context);
		*info.base64=new(PointerFreeGC) unsigned char[check_file_size(finfo.st_size, file_spec) * 2 + 6]; 
		unsigned char* base64 = *info.base64;
		int state=0;
		int save=0;
		int nCount;
		do {
			unsigned char buffer[FILE_BUFFER_SIZE];
			nCount = file_block_read(f, buffer, sizeof(buffer));
			if( nCount ){
				size_t filled=g_mime_utils_base64_encode_step ((const unsigned char*)buffer, nCount, base64, &state, &save);
				base64+=filled;
			}
		} while(nCount > 0);
		g_mime_utils_base64_encode_close (0, 0, base64, &state, &save);
	}
}

char* pa_base64_encode(const String& file_spec){
	unsigned char* base64=0;
	File_base64_action_info info={&base64}; 

	file_read_action_under_lock(file_spec, "pa_base64_encode", file_base64_file_action, &info);

	return (char*)base64; 
}

void pa_base64_decode(const char *in, size_t in_size, char*& result, size_t& result_size, bool strict) {
	// every 4 base64 bytes are converted into 3 normal bytes
	// not full set (tail) of 4-bytes set is ignored
	size_t new_size=in_size/4*3;
	result=new(PointerFreeGC) char[new_size+1/*terminator*/];

	int state=0;
	int save=0;
	result_size=g_mime_utils_base64_decode_step ((const unsigned char*)in, in_size, (unsigned char*)result, &state, &save, strict);
	assert(result_size <= new_size);
	result[result_size]=0; // for text files

	if(strict && state!=0)
		throw Exception(BASE64_FORMAT, 0, "Unexpected end of chars");
}


int file_block_read(const int f, unsigned char* buffer, const size_t size){
	int nCount = read(f, buffer, size);
	if (nCount < 0)
		throw Exception("file.read", 0, "read failed: %s (%d)", strerror(errno), errno); 
	return nCount;
}

static unsigned long crc32Table[256];
static void InitCrc32Table()
{
	if(crc32Table[1] == 0){
		// This is the official polynomial used by CRC32 in PKZip.
		// Often times the polynomial shown reversed as 0x04C11DB7.
		static const unsigned long dwPolynomial = 0xEDB88320;

		for(int i = 0; i < 256; i++)
		{
			unsigned long dwCrc = i;
			for(int j = 8; j > 0; j--)
			{
				if(dwCrc & 1)
					dwCrc = (dwCrc >> 1) ^ dwPolynomial;
				else
					dwCrc >>= 1;
			}
			crc32Table[i] = dwCrc;
		}
	}
}

inline void CalcCrc32(const unsigned char byte, unsigned long &crc32)
{
	crc32 = ((crc32) >> 8) ^ crc32Table[(byte) ^ ((crc32) & 0x000000FF)];
}


unsigned long pa_crc32(const char *in, size_t in_size){
	unsigned long crc32=0xFFFFFFFF;

	InitCrc32Table();
	for(size_t i = 0; i<in_size; i++)
		CalcCrc32(in[i], crc32);

	return ~crc32; 
}

static void file_crc32_file_action(struct stat& finfo, int f, const String&, void *context) {
	unsigned long& crc32=*static_cast<unsigned long *>(context);
	if(finfo.st_size) {
		InitCrc32Table();
		int nCount=0;
		do {
			unsigned char buffer[FILE_BUFFER_SIZE];
			nCount = file_block_read(f, buffer, sizeof(buffer));
			for(int i = 0; i < nCount; i++) CalcCrc32(buffer[i], crc32);
		} while(nCount > 0);
	}
}

unsigned long pa_crc32(const String& file_spec){
	unsigned long crc32=0xFFFFFFFF;
	file_read_action_under_lock(file_spec, "crc32", file_crc32_file_action, &crc32);
	return ~crc32; 
}

// content-type: xxx; charset=WE-NEED-THIS
// content-type: xxx; charset="WE-NEED-THIS"
// content-type: xxx; charset="WE-NEED-THIS";
Charset* detect_charset(const char* content_type){
	if(content_type){
		char* CONTENT_TYPE=str_upper(content_type);

		if(const char* begin=strstr(CONTENT_TYPE, "CHARSET=")){
			begin+=8; // skip "CHARSET="
			char* end=0;
			if(*begin && (*begin=='"' || *begin =='\'')){
				char quote=*begin;
				begin++;
				end=(char*)strchr(begin, quote);
			}
			if(!end)
				end=(char*)strchr(begin, ';');

			if(end)
				*end=0; // terminator

			return *begin ? &pa_charsets.get_direct(begin) : 0;
		}
	}
	return 0;
}

const UTF16* pa_utf16_encode(const char* in, Charset& source_charset){
	if(!in)
		return 0;

	String::C sIn(in,strlen(in));

	UTF16* utf16=(UTF16*)pa_malloc_atomic(sIn.length*2+2);
	UTF16* utf16_end=utf16;

	if(!source_charset.isUTF8())
		sIn=Charset::transcode(sIn, source_charset, pa_UTF8_charset);

	int status=pa_convertUTF8toUTF16((const UTF8**)&sIn.str, (const UTF8*)(sIn.str+sIn.length), &utf16_end, utf16+sIn.length, strictConversion);
	if(status != conversionOK)
		throw Exception("utf-16 encode", new String(in), "utf-16 conversion failed (%d)", status);

	*utf16_end=0;

	return utf16;
}

const char* pa_utf16_decode(const UTF16* in, Charset& asked_charset){
	if(!in)
		return 0;

	const UTF16* utf16_start=in;
	const UTF16* utf16_end;

	for(utf16_end=in; *utf16_end; utf16_end++);

	char *result = (char *)pa_malloc_atomic((utf16_end-in)*6+1);
	char *result_end = result;

	int status=pa_convertUTF16toUTF8(&utf16_start, utf16_end, (UTF8**)&result_end, (UTF8*)(result+(utf16_end-in)*6), strictConversion);

	if(status != conversionOK)
		throw Exception("utf-16 decode", 0, "utf conversion failed (%d)", status);

	*result_end='\0';

	if(asked_charset.isUTF8())
		return result;

	return Charset::transcode(result, pa_UTF8_charset, asked_charset).cstr();
}

static bool is_latin(const char *in){
	for(; *in; in++){
		if ((unsigned char)(*in) > 0x7F)
			return false;
	}
	return true;
}

#define MAX_IDNA_LENGTH 256

const char *pa_idna_encode(const char *in, Charset& source_charset){
	if(!in || is_latin(in))
		return in;

	uint32_t utf32[MAX_IDNA_LENGTH];
	uint32_t *utf32_end=utf32;

	String::C sIn(in,strlen(in));

	if(!source_charset.isUTF8())
		sIn=Charset::transcode(sIn, source_charset, pa_UTF8_charset);

	int status=pa_convertUTF8toUTF32((const UTF8**)&sIn.str, (const UTF8*)(sIn.str+sIn.length), &utf32_end, utf32+MAX_IDNA_LENGTH-1, strictConversion);
	if(status != conversionOK)
		throw Exception("idna encode", new String(in), "utf conversion failed (%d)", status);

	*utf32_end=0;

	char *result = (char *)pa_malloc(MAX_IDNA_LENGTH);
	status=pa_idna_to_ascii_4z(utf32, result, MAX_IDNA_LENGTH, 0);
	if(status != IDNA_SUCCESS)
		throw Exception("idna encode", new String(in), "encode failed: %s", pa_idna_strerror(status));

	return result;
}

const char *pa_idna_decode(const char *in, Charset &asked_charset){
	if(!in || !(*in))
		return in;

	uint32_t utf32[MAX_IDNA_LENGTH];
	const uint32_t *utf32_start=utf32;
	uint32_t *utf32_end;

	int status=pa_idna_to_unicode_4z(in, utf32, MAX_IDNA_LENGTH, 0);
	if(status != IDNA_SUCCESS)
		throw Exception("idna decode", new String(in), "decode failed: %s", pa_idna_strerror(status));

	for(utf32_end=utf32; *utf32_end; utf32_end++);

	char *result = (char *)pa_malloc(MAX_IDNA_LENGTH);
	char *result_end = result;

	status=pa_convertUTF32toUTF8(&utf32_start, utf32_end, (UTF8**)&result_end, (UTF8*)(result+MAX_IDNA_LENGTH-1), strictConversion);

	if(status != conversionOK)
		throw Exception("idna decode", new String(in), "utf conversion failed (%d)", status);

	*result_end='\0';

	if(!asked_charset.isUTF8())
		result = (char *)Charset::transcode(result, pa_UTF8_charset, asked_charset).cstr();

	return result;
}
/// must be last in this file
#undef vsnprintf
int pa_vsnprintf(char* b, size_t s, const char* f, va_list l) {
	if(!s)
		return 0;

	int r;
	// note: on win32 & maybe somewhere else
	// vsnprintf do not writes terminating 0 in 'buffer full' case, reducing
	// http://stackoverflow.com/questions/2915672/snprintf-and-visual-studio-2010
	--s;

	// clients do not check for negative 's', feature: ignore such prints
	if((ssize_t)s<0)
		return 0;

#ifdef _MSC_VER
	// win32: if the number of bytes to write exceeds buffer, then count bytes are written and -1 is returned
	r=_vsnprintf(b, s, f, l); 
	if(r<0) 
		r=s;
#else
	r=vsnprintf(b, s, f, l); 
	/*
	solaris: man vsnprintf

	The snprintf() function returns  the  number  of  characters
	formatted, that is, the number of characters that would have
	been written to the buffer if it were large enough.  If  the
	value  of  n  is  0  on a call to snprintf(), an unspecified
	value less than 1 is returned.
	*/

	if(r<0)
		r=0;
	else if((size_t)r>s)
		r=s;
#endif
	b[r]=0;
	return r;
}

int pa_snprintf(char* b, size_t s, const char* f, ...) {
	va_list l;
	va_start(l, f); 
	int r=pa_vsnprintf(b, s, f, l); 
	va_end(l); 
	return r;
}