Annotation of parser3/src/main/pa_string.C, revision 1.98
1.45 paf 1: /** @file
1.55 paf 2: Parser: string class. @see untasize_t.C.
1.46 paf 3:
1.36 paf 4: Copyright (c) 2001 ArtLebedev Group (http://www.artlebedev.com)
1.46 paf 5:
1.37 paf 6: Author: Alexander Petrosyan <paf@design.ru> (http://design.ru/paf)
1.4 paf 7: */
1.98 ! parser 8: static const char *RCSId="$Id: pa_string.C,v 1.97 2001/08/01 12:08:40 parser Exp $";
1.4 paf 9:
1.48 paf 10: #include "pa_config_includes.h"
1.1 paf 11:
1.70 paf 12: #include "pcre.h"
1.82 parser 13: #include "internal.h"
1.70 paf 14:
1.13 paf 15: #include "pa_pool.h"
1.12 paf 16: #include "pa_string.h"
1.5 paf 17: #include "pa_hash.h"
1.22 paf 18: #include "pa_exception.h"
1.53 paf 19: #include "pa_common.h"
1.60 paf 20: #include "pa_array.h"
21: #include "pa_globals.h"
1.61 paf 22: #include "pa_table.h"
1.60 paf 23:
1.75 paf 24: String::String(Pool& apool, const char *src, size_t src_size, bool tainted) :
1.94 parser 25: Pooled(apool),
26: forigins_mode(false) {
1.28 paf 27: last_chunk=&head;
28: head.count=CR_PREALLOCATED_COUNT;
1.5 paf 29: append_here=head.rows;
1.2 paf 30: head.preallocated_link=0;
1.28 paf 31: link_row=&head.rows[head.count];
1.8 paf 32: fused_rows=fsize=0;
1.41 paf 33:
34: if(src)
1.75 paf 35: if(tainted)
36: APPEND_TAINTED(src, src_size, 0, 0);
1.41 paf 37: else
1.75 paf 38: APPEND_CLEAN(src, src_size, 0, 0);
1.1 paf 39: }
40:
1.94 parser 41: String::String(const String& src) :
42: Pooled(src.pool()),
43: forigins_mode(false) {
1.8 paf 44: head.count=CR_PREALLOCATED_COUNT;
45:
1.55 paf 46: size_t src_used_rows=src.fused_rows;
1.8 paf 47: if(src_used_rows<=head.count) {
1.55 paf 48: // all new rows fit size_to preallocated area
1.98 ! parser 49: last_chunk=&head;
1.55 paf 50: size_t curr_chunk_rows=head.count;
1.8 paf 51: memcpy(head.rows, src.head.rows, sizeof(Chunk::Row)*src_used_rows);
52: append_here=&head.rows[src_used_rows];
53: link_row=&head.rows[curr_chunk_rows];
54: } else {
55: // warning:
1.10 paf 56: // heavily relies on the fact
57: // "preallocated area is the same for all strings"
1.8 paf 58: //
59: // info:
60: // allocating only enough mem to fit src string rows
61: // next append would allocate a new chunk
62: //
1.55 paf 63: // new rows don't fit size_to preallocated area: splitting size_to two chunks
1.8 paf 64: // preallocated chunk src to constructing head
65: memcpy(head.rows, src.head.rows, sizeof(Chunk::Row)*head.count);
1.55 paf 66: // remaining rows size_to new_chunk
67: size_t curr_chunk_rows=src_used_rows-head.count;
1.98 ! parser 68: last_chunk=static_cast<Chunk *>(
1.55 paf 69: malloc(sizeof(size_t)+sizeof(Chunk::Row)*curr_chunk_rows+sizeof(Chunk *)));
1.98 ! parser 70: last_chunk->count=curr_chunk_rows;
! 71: head.preallocated_link=last_chunk;
! 72: append_here=link_row=&last_chunk->rows[last_chunk->count];
1.8 paf 73:
74: Chunk *old_chunk=src.head.preallocated_link;
1.98 ! parser 75: Chunk::Row *new_rows=last_chunk->rows;
! 76: size_t rows_left_to_copy=last_chunk->count;
1.8 paf 77: while(true) {
1.55 paf 78: size_t old_count=old_chunk->count;
1.8 paf 79: Chunk *next_chunk=old_chunk->rows[old_count].link;
80: if(next_chunk) {
81: // not last source chunk
82: // taking it all
83: memcpy(new_rows, old_chunk->rows, sizeof(Chunk::Row)*old_count);
84: new_rows+=old_count;
85: rows_left_to_copy-=old_count;
86:
87: old_chunk=next_chunk;
88: } else {
89: // the last source chunk
90: // taking only those rows of chunk that _left_to_copy
91: memcpy(new_rows, old_chunk->rows, sizeof(Chunk::Row)*rows_left_to_copy);
92: break;
93: }
94: }
1.5 paf 95: }
1.8 paf 96: link_row->link=0;
97: fused_rows=src_used_rows;
98: fsize=src.fsize;
1.94 parser 99: }
100:
101: void String::expand() {
102: size_t new_chunk_count=last_chunk->count+CR_GROW_COUNT;
103: last_chunk=static_cast<Chunk *>(
104: malloc(sizeof(size_t)+sizeof(Chunk::Row)*new_chunk_count+sizeof(Chunk *)));
105: last_chunk->count=new_chunk_count;
106: link_row->link=last_chunk;
107: append_here=last_chunk->rows;
108: link_row=&last_chunk->rows[last_chunk->count];
109: link_row->link=0;
1.5 paf 110: }
1.28 paf 111:
1.42 paf 112: String& String::append(const String& src, Untaint_lang lang, bool forced) {
1.60 paf 113: const Chunk *chunk=&src.head;
1.40 paf 114: do {
1.60 paf 115: const Chunk::Row *row=chunk->rows;
116: for(size_t i=0; i<chunk->count; i++, row++) {
117: if(row==src.append_here)
1.40 paf 118: goto break2;
1.60 paf 119:
120: APPEND(row->item.ptr, row->item.size,
121: (lang!=UL_PASS_APPENDED && (row->item.lang==UL_TAINTED || forced))?lang:row->item.lang,
122: row->item.origin.file, row->item.origin.line);
1.40 paf 123: }
124: chunk=row->link;
125: } while(chunk);
126: break2:
1.60 paf 127: return *this;
1.34 paf 128: }
1.60 paf 129:
1.13 paf 130: String& String::real_append(STRING_APPEND_PARAMS) {
1.9 paf 131: if(!src)
132: return *this;
1.26 paf 133: if(!size)
134: size=strlen(src);
135: if(!size)
1.9 paf 136: return *this;
137:
1.1 paf 138: if(chunk_is_full())
139: expand();
140:
141: append_here->item.ptr=src;
1.26 paf 142: fsize+=append_here->item.size=size;
1.52 paf 143: append_here->item.lang=lang;
1.13 paf 144: #ifndef NO_STRING_ORIGIN
1.14 paf 145: append_here->item.origin.file=file;
146: append_here->item.origin.line=line;
1.13 paf 147: #endif
1.8 paf 148: append_here++; fused_rows++;
1.1 paf 149:
150: return *this;
1.97 parser 151: }
152:
153: char String::first_char() const {
154: if(!fused_rows)
155: THROW(0, 0,
156: this,
157: "getting first char of empty string");
158:
159: return *head.rows[0].item.ptr;
1.1 paf 160: }
161:
1.16 paf 162: uint String::hash_code() const {
1.7 paf 163: uint result=0;
1.5 paf 164:
1.16 paf 165: const Chunk *chunk=&head;
1.5 paf 166: do {
1.16 paf 167: const Chunk::Row *row=chunk->rows;
1.55 paf 168: for(size_t i=0; i<chunk->count; i++) {
1.5 paf 169: if(row==append_here)
170: goto break2;
171:
1.6 paf 172: result=Hash::generic_code(result, row->item.ptr, row->item.size);
1.5 paf 173: row++;
174: }
175: chunk=row->link;
176: } while(chunk);
177: break2:
178: return result;
179: }
180:
1.60 paf 181: /// @todo move 'lang' skipping to pos
182: int String::cmp(int& partial, const String& src,
183: size_t this_offset, Untaint_lang lang) const {
1.59 paf 184: partial=-1;
1.55 paf 185: this_offset=min(this_offset, size()-1);
186:
1.16 paf 187: const Chunk *a_chunk=&head;
188: const Chunk *b_chunk=&src.head;
189: const Chunk::Row *a_row=a_chunk->rows;
190: const Chunk::Row *b_row=b_chunk->rows;
1.55 paf 191: size_t a_offset=this_offset;
192: size_t b_offset=0;
1.9 paf 193: Chunk::Row *a_end=append_here;
194: Chunk::Row *b_end=src.append_here;
1.55 paf 195: size_t a_countdown=a_chunk->count;
196: size_t b_countdown=b_chunk->count;
197: size_t result;
1.60 paf 198: size_t pos=0;
1.33 paf 199:
1.83 parser 200: bool a_break=size()==0;
1.91 parser 201: bool b_break=src.size()==0;
1.83 parser 202: if(!(a_break || b_break)) while(true) {
1.55 paf 203: if(pos+a_row->item.size > this_offset) {
1.71 paf 204: if(lang!=UL_UNSPECIFIED && a_row->item.lang!=lang)
1.60 paf 205: return -1; // wrong lang -- bail out
206:
1.55 paf 207: int size_diff=
208: (a_row->item.size-a_offset)-
209: (b_row->item.size-b_offset);
210:
211: if(size_diff==0) { // a has same size as b
1.60 paf 212: result=memcmp(a_row->item.ptr+a_offset, b_row->item.ptr+b_offset,
213: a_row->item.size-a_offset);
1.55 paf 214: if(result)
215: return result;
1.60 paf 216: pos+=a_row->item.size;
1.55 paf 217: a_row++; a_countdown--; a_offset=0;
218: b_row++; b_countdown--; b_offset=0;
219: } else if (size_diff>0) { // a longer
1.60 paf 220: result=memcmp(a_row->item.ptr+a_offset, b_row->item.ptr+b_offset,
221: b_row->item.size-b_offset);
1.55 paf 222: if(result)
223: return result;
224: a_offset+=b_row->item.size-b_offset;
225: b_row++; b_countdown--; b_offset=0;
226: } else { // b longer
1.60 paf 227: result=memcmp(a_row->item.ptr+a_offset, b_row->item.ptr+b_offset,
228: a_row->item.size-a_offset);
1.55 paf 229: if(result)
230: return result;
231: b_offset+=a_row->item.size-a_offset;
1.60 paf 232: pos+=a_row->item.size;
1.55 paf 233: a_row++; a_countdown--; a_offset=0;
234: }
1.83 parser 235: if(b_break=b_row==b_end) {
236: a_break=a_row==a_end;
237: break;
238: }
1.55 paf 239: if(!b_countdown) {
240: b_chunk=b_row->link;
241: b_row=b_chunk->rows;
242: b_countdown=b_chunk->count;
243: }
244: } else {
1.60 paf 245: a_offset-=a_row->item.size;
246: pos+=a_row->item.size;
247: a_row++; a_countdown--;
1.9 paf 248: }
249:
1.83 parser 250: if(a_break=a_row==a_end) {
251: b_break=b_row==b_end;
252: break;
253: }
1.11 paf 254: if(!a_countdown) {
1.9 paf 255: a_chunk=a_row->link;
256: a_row=a_chunk->rows;
1.11 paf 257: a_countdown=a_chunk->count;
1.9 paf 258: }
1.27 paf 259: }
1.55 paf 260: if(a_break==b_break) { // ended simultaneously
261: partial=0; return 0;
262: } else if(a_break) { // first bytes equal, but a ended before b
263: partial=1; return -1;
264: } else {
265: partial=2; return +1;
266: }
1.27 paf 267: }
268:
1.60 paf 269: /// @todo move 'lang' skipping to pos
1.59 paf 270: int String::cmp(int& partial, const char* b_ptr, size_t src_size,
1.60 paf 271: size_t this_offset, Untaint_lang lang) const {
1.59 paf 272: partial=-1;
1.50 paf 273: size_t b_size=src_size?src_size:b_ptr?strlen(b_ptr):0;
1.59 paf 274: this_offset=min(this_offset, size()-1);
1.27 paf 275:
276: const Chunk *a_chunk=&head;
277: const Chunk::Row *a_row=a_chunk->rows;
1.59 paf 278: size_t a_offset=this_offset;
1.55 paf 279: size_t b_offset=0;
1.27 paf 280: Chunk::Row *a_end=append_here;
1.55 paf 281: size_t a_countdown=a_chunk->count;
1.60 paf 282: size_t pos=0;
1.52 paf 283:
1.83 parser 284: bool a_break=size()==0;
285: bool b_break=b_size==0;
286: if(!(a_break || b_break)) while(true) {
1.59 paf 287: if(pos+a_row->item.size > this_offset) {
1.71 paf 288: if(lang!=UL_UNSPECIFIED && a_row->item.lang!=lang)
1.60 paf 289: return -1; // wrong lang -- bail out
290:
1.59 paf 291: int size_diff=
292: (a_row->item.size-a_offset)-
293: (b_size-b_offset);
294:
295: if(size_diff==0) { // a has same size as b
296: if(size_t result=memcmp(a_row->item.ptr+a_offset, b_ptr+b_offset,
297: a_row->item.size-a_offset)!=0)
298: return result;
1.60 paf 299: pos+=a_row->item.size;
1.59 paf 300: a_row++; a_countdown--; a_offset=0;
301: b_break=true;
302: } else if (size_diff>0) { // a longer
303: if(size_t result=memcmp(a_row->item.ptr+a_offset, b_ptr+b_offset,
304: b_size-b_offset)!=0)
305: return result;
306: a_offset+=b_size-b_offset;
307: b_break=true;
308: } else { // b longer
309: if(size_t result=memcmp(a_row->item.ptr+a_offset, b_ptr+b_offset,
310: a_row->item.size-a_offset)!=0)
311: return result;
312: b_offset+=a_row->item.size-a_offset;
1.60 paf 313: pos+=a_row->item.size;
1.59 paf 314: a_row++; a_countdown--; a_offset=0;
315: }
316: } else {
1.60 paf 317: a_offset-=a_row->item.size;
318: pos+=a_row->item.size;
319: a_row++; a_countdown--;
1.27 paf 320: }
321:
1.86 parser 322: a_break=a_row==a_end;
323: if(a_break || b_break)
1.83 parser 324: break;
1.27 paf 325: if(!a_countdown) {
326: a_chunk=a_row->link;
327: a_row=a_chunk->rows;
328: a_countdown=a_chunk->count;
1.9 paf 329: }
330: }
1.55 paf 331: if(a_break==b_break) { // ended simultaneously
332: partial=0; return 0;
333: } else if(a_break) { // first bytes equal, but a ended before b
334: partial=1; return -1;
335: } else {
336: partial=2; return +1;
337: }
1.5 paf 338: }
1.46 paf 339:
340: #ifndef NO_STRING_ORIGIN
341: const Origin& String::origin() const {
1.96 parser 342: if(!fused_rows) {
343: static const Origin empty_origin={"empty string"};
344: return empty_origin;
345: }
1.46 paf 346:
1.49 paf 347: // determining origin by last appended piece
1.50 paf 348: // because first one frequently constant.
349: // ex: ^load[/file] "document_root" + "/file"
1.80 paf 350: // when last peice is constant,
351: // ex: parser_root_auto_path{dynamic} / auto.p{const}
352: // using first piece
353: Origin& last_origin=append_here[-1].item.origin;
354: return last_origin.file ? last_origin : head.rows[0].item.origin;
1.46 paf 355: }
356: #endif
1.53 paf 357:
1.69 paf 358: String& String::mid(size_t start, size_t finish) const {
1.53 paf 359: start=max(0, start);
360: finish=min(size(), finish);
1.60 paf 361: if(start==finish)
362: return *empty_string;
1.53 paf 363:
364: String& result=*NEW String(pool());
365:
366: size_t pos=0;
367: const Chunk *chunk=&head;
368: do {
369: const Chunk::Row *row=chunk->rows;
1.55 paf 370: for(size_t i=0; i<chunk->count; pos+=row->item.size, i++, row++) {
1.53 paf 371: if(row==append_here)
372: goto break2;
373:
1.60 paf 374: size_t item_finish=pos+row->item.size;
375: if(item_finish > start) { // started now or already?
376: bool started=result.size()==0; // started now?
377: bool finished=finish <= item_finish; // finished now?
1.53 paf 378: size_t offset=started?start-pos:0;
379: size_t size=finished?finish-pos:row->item.size;
380: result.APPEND(
381: row->item.ptr+offset, size-offset,
382: row->item.lang,
383: row->item.origin.file, row->item.origin.line);
384: if(finished)
385: goto break2;
386: }
387: }
388: chunk=row->link;
389: } while(chunk);
390: break2:
1.60 paf 391: // SAPI::log(pool(), "piece of '%s' from %d to %d is '%s'",
392: //cstr(), start, finish, result.cstr());
1.53 paf 393: return result;
1.54 paf 394: }
395:
1.60 paf 396: int String::pos(const String& substr,
397: size_t result, Untaint_lang lang) const {
1.58 paf 398: for(; result<size(); result++) {
1.60 paf 399: int partial; cmp(partial, substr, result, lang);
1.58 paf 400: if(
401: partial==0 || // full match
402: partial==2) // 'substr' starts 'this'+'result'
403: return result;
404: }
405:
406: return -1;
407: }
408:
1.60 paf 409: int String::pos(const char *substr, size_t substr_size,
410: size_t result, Untaint_lang lang) const {
1.57 paf 411: for(; result<size(); result++) {
1.60 paf 412: int partial; cmp(partial, substr, substr_size, result, lang);
1.55 paf 413: if(
414: partial==0 || // full match
415: partial==2) // 'substr' starts 'this'+'result'
416: return result;
417: }
418:
419: return -1;
1.60 paf 420: }
421:
422: void String::split(Array& result,
423: size_t* pos_after_ref,
424: const char *delim, size_t delim_size,
425: Untaint_lang lang, int limit) const {
426: if(delim_size) {
427: size_t pos_after=pos_after_ref?*pos_after_ref:0;
428: int pos_before;
429: // while we have 'delim'...
430: for(; (pos_before=pos(delim, delim_size, pos_after, lang))>=0 && limit; limit--) {
1.69 paf 431: result+=&mid(pos_after, pos_before);
1.60 paf 432: pos_after=pos_before+delim_size;
433: }
434: // last piece
435: if(pos_after<size() && limit) {
1.69 paf 436: result+=&mid(pos_after, size());
1.60 paf 437: pos_after=size();
438: }
439: if(pos_after_ref)
440: *pos_after_ref=pos_after;
441: } else { // empty delim
442: result+=this;
443: if(pos_after_ref)
444: *pos_after_ref+=size();
445: }
446: }
447:
448: void String::split(Array& result,
449: size_t* pos_after_ref,
450: const String& delim, Untaint_lang lang,
451: int limit) const {
452: if(delim.size()) {
453: size_t pos_after=pos_after_ref?*pos_after_ref:0;
454: int pos_before;
455: // while we have 'delim'...
456: for(; (pos_before=pos(delim, pos_after, lang))>=0 && limit; limit--) {
1.69 paf 457: result+=&mid(pos_after, pos_before);
1.60 paf 458: pos_after=pos_before+delim.size();
459: }
460: // last piece
461: if(pos_after<size() && limit) {
1.69 paf 462: result+=&mid(pos_after, size());
1.60 paf 463: pos_after=size();
464: }
465: if(pos_after_ref)
466: *pos_after_ref=pos_after;
467: } else { // empty delim
468: result+=this;
469: if(pos_after_ref)
470: *pos_after_ref+=size();
471: }
1.61 paf 472: }
473:
1.63 paf 474: static void regex_options(char *options, int *result){
475: struct Regex_option {
476: char key;
477: int clear, set;
478: int *result;
479: } regex_option[]={
480: {'i', 0, PCRE_CASELESS, result}, // a=A
1.79 paf 481: {'s', 0, PCRE_DOTALL, result}, // \n\n$ [default]
1.63 paf 482: {'x', 0, PCRE_EXTENDED, result}, // whitespace in regex ignored
483: {'m', PCRE_DOTALL, PCRE_MULTILINE, result}, // ^aaa\n$^bbb\n$
484: {'g', 0, true, result+1}, // many rows
485: {0},
486: };
487: result[0]=PCRE_EXTRA | PCRE_DOTALL;
488: result[1]=0;
489:
490: if(options)
491: for(Regex_option *o=regex_option; o->key; o++)
492: if(
493: strchr(options, o->key) ||
494: strchr(options, toupper(o->key))) {
495: *(o->result)&=~o->clear;
496: *(o->result)|=o->set;
497: }
498: }
499:
1.88 parser 500: /// @todo maybe need speedup: some option to remove pre/match/post string generation
1.77 paf 501: bool String::match(const unsigned char *pcre_tables,
502: const String *aorigin,
1.62 paf 503: const String& regexp,
1.63 paf 504: const String *options,
1.64 paf 505: Table **table,
1.95 parser 506: Row_action row_action, void *info,
507: bool *was_global) const {
1.64 paf 508:
1.73 paf 509: if(!regexp.size())
510: THROW(0, 0,
511: aorigin,
512: "regexp is empty");
1.68 paf 513: const char *pattern=regexp.cstr(UL_AS_IS);
1.62 paf 514: const char *errptr;
515: int erroffset;
1.63 paf 516: int option_bits[2]; regex_options(options?options->cstr():0, option_bits);
1.95 parser 517: if(was_global)
518: *was_global=option_bits[1]!=0;
1.63 paf 519: pcre *code=pcre_compile(pattern, option_bits[0],
1.62 paf 520: &errptr, &erroffset,
1.74 paf 521: pcre_tables);
1.62 paf 522:
1.67 paf 523: if(!code)
1.62 paf 524: THROW(0, 0,
1.69 paf 525: ®exp.mid(erroffset, regexp.size()),
1.74 paf 526: "regular expression syntax error - %s", errptr);
1.62 paf 527:
1.63 paf 528: int info_substrings=pcre_info(code, 0, 0);
529: if(info_substrings<0) {
530: (*pcre_free)(code);
531: THROW(0, 0,
1.73 paf 532: aorigin,
1.76 paf 533: "pcre_info error (%d)",
1.73 paf 534: info_substrings);
1.63 paf 535: }
536:
537: int startoffset=0;
1.68 paf 538: const char *subject=cstr(UL_AS_IS);
1.62 paf 539: int length=strlen(subject);
1.63 paf 540: int ovecsize;
541: int *ovector=(int *)malloc(sizeof(int)*
1.65 paf 542: (ovecsize=(1/*match*/+info_substrings)*3));
1.62 paf 543:
1.64 paf 544: { // create table
545: Array& columns=*NEW Array(pool());
546: columns+=string_pre_match_name;
547: columns+=string_match_name;
548: columns+=string_post_match_name;
549: for(int i=1; i<=info_substrings; i++) {
550: char *column=(char *)malloc(MAX_NUMBER);
551: snprintf(column, MAX_NUMBER, "%d", i);
552: columns+=NEW String(pool(), column); // .i column name
553: }
554: *table=NEW Table(pool(), aorigin, &columns);
1.62 paf 555: }
1.63 paf 556:
1.64 paf 557: int exec_option_bits=0;
1.63 paf 558: while(true) {
559: int exec_substrings=pcre_exec(code, 0,
560: subject, length, startoffset,
1.64 paf 561: exec_option_bits, ovector, ovecsize);
1.63 paf 562:
563: if(exec_substrings==PCRE_ERROR_NOMATCH) {
564: (*pcre_free)(code);
1.67 paf 565: (*row_action)(**table, 0/*last time, no row*/, 0, 0, info);
1.63 paf 566: return option_bits[1]!=0; // global=true+table, not global=false
567: }
568:
569: if(exec_substrings<0) {
570: (*pcre_free)(code);
571: THROW(0, 0,
572: aorigin,
1.76 paf 573: "regular expression execute error (%d)",
1.63 paf 574: exec_substrings);
575: }
576:
577: Array& row=*NEW Array(pool());
1.81 paf 578: row+=&mid(0, ovector[0]); // .prematch column value
1.69 paf 579: row+=&mid(ovector[0], ovector[1]); // .match
1.81 paf 580: row+=&mid(ovector[1], size()); // .postmatch
1.63 paf 581:
582: for(int i=1; i<exec_substrings; i++) {
1.69 paf 583: // -1:-1 case handled peacefully by mid() itself
584: row+=&mid(ovector[i*2+0], ovector[i*2+1]); // .i column value
1.63 paf 585: }
586:
1.67 paf 587: (*row_action)(**table, &row, startoffset, ovector[0], info);
1.63 paf 588:
1.67 paf 589: if(!option_bits[1] || !(startoffset=ovector[1])) { // not global | going to hang
1.63 paf 590: (*pcre_free)(code);
1.67 paf 591: (*row_action)(**table, 0/*last time, no row*/, 0, 0, info);
1.63 paf 592: return true;
593: }
594:
595: /*
596: if(option_bits[0] & PCRE_MULTILINE)
1.64 paf 597: exec_option_bits|=PCRE_NOTBOL; // start of subject+startoffset not BOL
1.63 paf 598: */
599: }
1.82 parser 600: }
601:
602: String& String::change_case(Pool& pool, const unsigned char *tables,
603: Change_case_kind kind) const {
604: String& result=*new(pool) String(pool);
605:
606: const unsigned char *a;
607: const unsigned char *b;
608: switch(kind) {
609: case CC_UPPER:
610: a=tables+lcc_offset;
611: b=tables+fcc_offset;
612: break;
613: case CC_LOWER:
614: a=tables+lcc_offset;
615: b=0;
616: break;
617: default:
618: PTHROW(0, 0,
619: this,
620: "unknown change case kind #%d",
621: static_cast<int>(kind)); // never
622: a=b=0; // calm, compiler
623: break; // never
624: }
625:
626: const Chunk *chunk=&head;
627: do {
628: const Chunk::Row *row=chunk->rows;
629: for(size_t i=0; i<chunk->count; i++, row++) {
630: if(row==append_here)
631: goto break2;
632:
633: char *new_cstr=(char *)pool.malloc(row->item.size);
634: char *dest=new_cstr;
635: const char *src=row->item.ptr;
636: for(int size=row->item.size; size--; src++) {
637: unsigned char c=a[(unsigned char)*src];
638: if(b)
639: c=b[c];
640:
641: *dest++=(char)c;
642: }
643:
644: result.APPEND(new_cstr, row->item.size,
645: row->item.lang,
646: row->item.origin.file, row->item.origin.line);
647: }
648: chunk=row->link;
649: } while(chunk);
650: break2:
1.89 parser 651:
652: return result;
653: }
654:
1.90 parser 655: double String::as_double() const {
1.89 parser 656: double result;
657: const char *cstr=this->cstr();
658: char *error_pos=0;
659: // 0xABC
660: if(cstr[0]=='0' && (cstr[1]=='x' || cstr[1]=='X'))
661: result=(double)(unsigned long)strtol(cstr, &error_pos, 0);
662: else
663: result=strtod(cstr, &error_pos);
664:
665: if(error_pos && *error_pos)
666: THROW(0, 0,
667: this,
668: "invalid number (double)");
669:
670: return result;
671: }
1.90 parser 672: int String::as_int() const {
1.89 parser 673: int result;
674: const char *cstr=this->cstr();
675: char *error_pos=0;
676: // 0xABC
677: if(cstr[0]=='0' && (cstr[1]=='x' || cstr[1]=='X'))
678: result=(int)(unsigned long)strtol(cstr, &error_pos, 0);
679: else
680: result=(int)strtol(cstr, &error_pos, 0);
681:
682: if(error_pos && *error_pos)
683: THROW(0, 0,
684: this,
685: "invalid number (int)");
1.82 parser 686:
687: return result;
1.61 paf 688: }
E-mail:
![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to pa_string.C CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [parser3project] / parser3 / src / main