Chameleon

Chameleon Commit Details

Date:2014-01-08 09:02:30 (10 years 2 months ago)
Author:Chuck Fry
Commit:2321
Parents: 2320
Message:EXPERIMENTAL optimizations to XML parser - run at own risk!
Changes:
M/branches/chucko/i386/libsaio/xml.c

File differences

branches/chucko/i386/libsaio/xml.c
340340
341341
342342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
343
344
345
360346
361
347
362348
363349
364350
365
351
366352
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
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
396383
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
421408
422
423
424
425
426
427
428
409
410
411
412
413
414
415
429416
430
431
432
433
417
418
419
420
421
422
423
424
425
426
434427
435428
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
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
464458
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
489483
490
491
492
493
494
495
484
485
486
487
488
489
496490
497
491
498492
499
500
501
502
503
504
505
506
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528508
529509
530510
531511
532512
513
533514
534515
535516
536517
537518
538
539
540
519
520
521
541522
542
523
543524
544
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
545547
546
548
547549
548
550
551
552
553
554
549555
550
556
551557
552558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
553576
554577
555578
......
877900
878901
879902
880
903
904
905
881906
882
907
908
883909
884
885
886
887
888
889
890
891
892
910
911
912
913
914
915
916
893917
894
895
896
897
898
899
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
900939
901940
902941
if (length == -1) return -1;
pos = length;
if (!strncmp(tagName, kXMLTagPList, 6))
{
length = 0;
}
/***** dict ****/
else if (!strcmp(tagName, kXMLTagDict))
{
length = ParseTagList(buffer + pos, tag, kTagTypeDict, 0);
}
else if (!strncmp(tagName, kXMLTagDict, strlen(kXMLTagDict)) && tagName[strlen(tagName)-1] == '/')
{
length = ParseTagList(buffer + pos, tag, kTagTypeDict, 1);
}
else if (!strncmp(tagName, kXMLTagDict " ", strlen(kXMLTagDict " ")))
{
length = ParseTagList(buffer + pos, tag, kTagTypeDict, 0);
}
/* Check most common cases first, make them fast */
/***** key ****/
else if (!strcmp(tagName, kXMLTagKey))
if (!strcmp(tagName, kXMLTagKey))
{
length = ParseTagKey(buffer + pos, tag);
}
/***** string ****/
else if (!strcmp(tagName, kXMLTagString))
{
length = ParseTagString(buffer + pos, tag);
}
else if (!strncmp(tagName, kXMLTagString " ", strlen(kXMLTagString " ")))
{
// TODO: save tag if if found
if(!strncmp(tagName + strlen(kXMLTagString " "), kXMLStringID, strlen(kXMLStringID)))
{
// ID=
int id = 0;
int cnt = strlen(kXMLTagString " " kXMLStringID "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagString " " kXMLStringID "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseStringID error (0x%x)\n", *val);
getchar();
return -1;
}
}
length = ParseTagString(buffer + pos, tag);
else if (!strncmp(tagName, kXMLTagString, strlen(kXMLTagString))) {
if (!tagName[strlen(kXMLTagString)]) /* <string> */
{
length = ParseTagString(buffer + pos, tag);
}
else if (' ' == tagName[strlen(kXMLTagString)]) /* <string ...> */
{
// TODO: save tag if if found
if(!strncmp(tagName + strlen(kXMLTagString " "), kXMLStringID, strlen(kXMLStringID))) /* <string ID= ...> */
{
// ID=
int id = 0;
int cnt = strlen(kXMLTagString " " kXMLStringID "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagString " " kXMLStringID "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseStringID error (0x%x)\n", *val);
getchar();
return -1;
}
}
length = ParseTagString(buffer + pos, tag);
SaveRefString(buffer + pos, id);
}
else if(!strncmp(tagName + strlen(kXMLTagString " "), kXMLStringIDRef, strlen(kXMLStringIDRef)))
{
// IDREF=
int id = 0;
int cnt = strlen(kXMLTagString " " kXMLStringIDRef "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagString " " kXMLStringIDRef "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseStringIDREF error (0x%x)\n", *val);
getchar();
return -1;
}
}
char* str = GetRefString(id);
SaveRefString(buffer + pos, id);
}
else if(!strncmp(tagName + strlen(kXMLTagString " "), kXMLStringIDRef, strlen(kXMLStringIDRef))) /* <string IDREF= ...> */
{
// IDREF=
int id = 0;
int cnt = strlen(kXMLTagString " " kXMLStringIDRef "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagString " " kXMLStringIDRef "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseStringIDREF error (0x%x)\n", *val);
getchar();
return -1;
}
}
char* str = GetRefString(id);
TagPtr tmpTag = NewTag();
tmpTag->type = kTagTypeString;
tmpTag->string = str;
tmpTag->tag = 0;
tmpTag->tagNext = 0;
tmpTag->offset = buffer_start ? buffer - buffer_start + pos : 0;
*tag = tmpTag;
TagPtr tmpTag = NewTag();
tmpTag->type = kTagTypeString;
tmpTag->string = str;
tmpTag->tag = 0;
tmpTag->tagNext = 0;
tmpTag->offset = buffer_start ? buffer - buffer_start + pos : 0;
*tag = tmpTag;
length = 0;
//printf("Located IDREF, id = %d, string = %s\n", id, str);
}
}
length = 0;
//printf("Located IDREF, id = %d, string = %s\n", id, str);
}
}
else /* unrecognized <string...> */
{
*tag = 0;
length = 0;
}
}
/***** integer ****/
else if (!strcmp(tagName, kXMLTagInteger))
{
length = ParseTagInteger(buffer + pos, tag);
}
else if (!strncmp(tagName, kXMLTagInteger " ", strlen(kXMLTagInteger " ")))
{
if(!strncmp(tagName + strlen(kXMLTagInteger " "), kXMLStringID, strlen(kXMLStringID)))
{
// ID=
int id = 0;
int cnt = strlen(kXMLTagInteger " " kXMLStringID "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagInteger " " kXMLStringID "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseIntegerID error (0x%x)\n", *val);
getchar();
return -1;
}
}
length = ParseTagInteger(buffer + pos, tag);
else if (!strncmp(tagName, kXMLTagInteger, strlen(kXMLTagInteger))) {
if (!tagName[strlen(kXMLTagInteger)]) /* <integer> */
{
length = ParseTagInteger(buffer + pos, tag);
}
else if (' ' == tagName[strlen(kXMLTagInteger)]) /* <integer ...> */
{
if(!strncmp(tagName + strlen(kXMLTagInteger " "), kXMLStringID, strlen(kXMLStringID))) /* <integer ID=...> */
{
// ID=
int id = 0;
int cnt = strlen(kXMLTagInteger " " kXMLStringID "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagInteger " " kXMLStringID "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseIntegerID error (0x%x)\n", *val);
getchar();
return -1;
}
}
length = ParseTagInteger(buffer + pos, tag);
SaveRefString((*tag)->string, id);
}
else if(!strncmp(tagName + strlen(kXMLTagInteger " "), kXMLStringIDRef, strlen(kXMLStringIDRef)))
{
// IDREF=
int id = 0;
int cnt = strlen(kXMLTagInteger " " kXMLStringIDRef "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagInteger " " kXMLStringIDRef "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseStringIDREF error (0x%x)\n", *val);
getchar();
return -1;
}
}
int integer = (int)GetRefString(id);
SaveRefString((*tag)->string, id);
}
else if(!strncmp(tagName + strlen(kXMLTagInteger " "), kXMLStringIDRef, strlen(kXMLStringIDRef))) /* <integer IDREF=...> */
{
// IDREF=
int id = 0;
int cnt = strlen(kXMLTagInteger " " kXMLStringIDRef "\"") + 1;
while ((tagName[cnt] != '\0') && (tagName[cnt] != '"')) cnt++;
tagName[cnt] = 0;
char* val = tagName + strlen(kXMLTagInteger " " kXMLStringIDRef "\"");
while(*val)
{
if ((*val >= '0' && *val <= '9'))// 0 - 9
{
id = (id * 10) + (*val++ - '0');
}
else
{
printf("ParseStringIDREF error (0x%x)\n", *val);
getchar();
return -1;
}
}
int integer = (int)GetRefString(id);
TagPtr tmpTag = NewTag();
tmpTag->type = kTagTypeInteger;
tmpTag->string = (char*) integer;
tmpTag->tag = 0;
tmpTag->tagNext = 0;
tmpTag->offset = buffer_start ? buffer - buffer_start + pos : 0;
TagPtr tmpTag = NewTag();
tmpTag->type = kTagTypeInteger;
tmpTag->string = (char*) integer;
tmpTag->tag = 0;
tmpTag->tagNext = 0;
tmpTag->offset = buffer_start ? buffer - buffer_start + pos : 0;
*tag = tmpTag;
*tag = tmpTag;
length = 0;
//printf("Located IDREF, id = %d, string = %s\n", id, str);
}
else
{
length = ParseTagInteger(buffer + pos, tag);
}
}
length = 0;
//printf("Located IDREF, id = %d, string = %s\n", id, str);
}
else /* presume <integer ...>...</integer> */
{
length = ParseTagInteger(buffer + pos, tag);
}
}
else /* unrecognized <integer...> */
{
*tag = 0;
length = 0;
}
}
/***** data ****/
else if (!strcmp(tagName, kXMLTagData))
{
length = ParseTagData(buffer + pos, tag);
}
else if (!strncmp(tagName, kXMLTagData " ", strlen(kXMLTagData " ")))
{
length = ParseTagData(buffer + pos, tag);
}
else if (!strcmp(tagName, kXMLTagDate))
{
length = ParseTagDate(buffer + pos, tag);
}
/***** date ****/
else if (!strncmp(tagName, kXMLTagDate " ", strlen(kXMLTagDate " ")))
{
length = ParseTagDate(buffer + pos, tag);
}
/***** false ****/
else if (!strcmp(tagName, kXMLTagFalse))
{
length = ParseTagBoolean(buffer + pos, tag, kTagTypeFalse);
}
/***** true ****/
else if (!strcmp(tagName, kXMLTagTrue))
{
length = ParseTagBoolean(buffer + pos, tag, kTagTypeTrue);
}
/***** array ****/
else if (!strcmp(tagName, kXMLTagArray))
/***** plist ****/
else if (!strncmp(tagName, kXMLTagPList, 6))
{
length = ParseTagList(buffer + pos, tag, kTagTypeArray, 0);
length = 0;
}
else if (!strncmp(tagName, kXMLTagArray " ", strlen(kXMLTagArray " ")))
/***** dict ****/
else if (!strncmp(tagName, kXMLTagDict, strlen(kXMLTagDict))) {
if (!strncmp(tagName, kXMLTagDict, strlen(kXMLTagDict)) && tagName[strlen(tagName)-1] == '/') /* <dict.../> */
{
length = ParseTagList(buffer + pos, tag, kTagTypeDict, 1);
}
else if (!tagName[strlen(kXMLTagDict)] || ' ' == tagName[strlen(kXMLTagDict)]) /* <dict> or<dict ...> */
{
length = ParseTagList(buffer + pos, tag, kTagTypeDict, 0);
}
else /* unrecognized <dict...> */
{
*tag = 0;
length = 0;
}
}
/***** data ****/
else if ((!strncmp(tagName, kXMLTagData, strlen(kXMLTagData)))
&& (!tagName[strlen(kXMLTagData)] /* <data> */
|| ' ' == tagName[strlen(kXMLTagData)])) /* <data ...> */
{
length = ParseTagList(buffer + pos, tag, kTagTypeArray, 0);
length = ParseTagData(buffer + pos, tag);
}
else if (!strcmp(tagName, kXMLTagArray "/"))
/***** date ****/
else if ((!strncmp(tagName, kXMLTagDate, strlen(kXMLTagDate)))
&& (!tagName[strlen(kXMLTagDate)] /* <date> */
|| ' ' == tagName[strlen(kXMLTagDate)])) /* <date ...> */
{
length = ParseTagList(buffer + pos, tag, kTagTypeArray, 1);
length = ParseTagDate(buffer + pos, tag);
}
/***** array ****/
else if (!strncmp(tagName, kXMLTagArray, strlen(kXMLTagArray))) { /* <array...> */
char c = tagName[strlen(kXMLTagArray)];
if ('/' == c) /* <array/> */
{
length = ParseTagList(buffer + pos, tag, kTagTypeArray, 1);
}
else if ('\0' == c || ' ' == c) /* <array> or <array ...> */
{
length = ParseTagList(buffer + pos, tag, kTagTypeArray, 0);
}
else /* unrecognized <array...> */
{
*tag = 0;
length = 0;
}
}
/***** unknown ****/
else
{
static long
GetNextTag( char * buffer, char ** tag, long * start )
{
long cnt, cnt2;
char* tagStart = buffer;
char* tagEnd;
char c;
if (tag == 0) return -1;
if (!tag)
return -1;
// Find the start of the tag.
cnt = 0;
while ((buffer[cnt] != '\0') && (buffer[cnt] != '<')) cnt++;
if (buffer[cnt] == '\0') return -1;
// Find the end of the tag.
cnt2 = cnt + 1;
while ((buffer[cnt2] != '\0') && (buffer[cnt2] != '>')) cnt2++;
if (buffer[cnt2] == '\0') return -1;
/* Find the start of the tag. */
do {
c = *tagStart++;
if (c == '<') break;
} while (c);
if (!c)
return -1;
// Fix the tag data.
*tag = buffer + cnt + 1;
buffer[cnt2] = '\0';
if (start) *start = cnt;
return cnt2 + 1;
/* tagStart points just past the '<' */
/* Find the end of the tag. */
tagEnd = tagStart;
do {
c = *tagEnd++;
if (c == '>') break;
} while (c);
if (!c)
return -1;
/* tagEnd points just past the '>' */
/* Fix up tag data by nulling out the '>' */
--tagEnd;
*tagEnd = '\0';
*tag = tagStart; /* first char after initial '<' */
if (start)
*start = (long) (tagStart - buffer) - 1; /* offset of the initial '<' itself */
/* Return offset to the char after '>' */
return (long) (tagEnd - buffer) + 1;
}
//==========================================================================

Archive Download the corresponding diff file

Revision: 2321