Annotation of parser3/src/main/pa_string.C, revision 1.172.2.21.2.14
1.45 paf 1: /** @file
1.172.2.21.2. (paf 2:): Parser: string class. @see untalength_t.C.
1.46 paf 3:
1.172.2.11 paf 4: Copyright (c) 2001-2003 ArtLebedev Group (http://www.artlebedev.com)
1.138 paf 5: Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)
1.164 paf 6: */
1.46 paf 7:
1.172.2.21.2. 4(paf 8:3): static const char* IDENT_STRING_C="$Date: 2003/03/20 16:57:43 $";
1.4 paf 9:
1.70 paf 10: #include "pcre.h"
11:
1.12 paf 12: #include "pa_string.h"
1.22 paf 13: #include "pa_exception.h"
1.61 paf 14: #include "pa_table.h"
1.101 parser 15: #include "pa_dictionary.h"
1.132 paf 16: #include "pa_charset.h"
1.60 paf 17:
1.172.2.2 paf 18: // helpers
1.139 paf 19:
1.172.2.2 paf 20: /// String::match uses this as replace & global search table columns
1.139 paf 21:
1.172.2.4 paf 22: const int MAX_MATCH_GROUPS=100;
23:
1.172.2.14 paf 24: class String_match_table_template_columns: public ArrayString {
1.172.2.2 paf 25: public:
1.172.2.4 paf 26: String_match_table_template_columns() {
1.172.2.21.2. (paf 27:): *this+=new String("prematch");
28:): *this+=new String("match");
29:): *this+=new String("postmatch");
1.172.2.4 paf 30: for(int i=0; i<MAX_MATCH_GROUPS; i++) {
1.172.2.21.2. (paf 31:): char *cname=new(PointerFreeGC) char[3/*strlen("100")*/+1/*terminating 0*/];
32:): *this+=new String(cname, sprintf(cname, "%d", 1+i));
1.172.2.2 paf 33: }
1.172.2.21 paf 34: }
1.172.2.4 paf 35: };
36:
1.172.2.21.2. (paf 37:): Table string_match_table_template(new String_match_table_template_columns);
1.172.2.2 paf 38:
1.172.2.21.2. (paf 39:): // String::ArrayFragment methods
40:):
3(paf 41:3): void String::ArrayFragment::append_positions(const ArrayFragment& src,
42:3): size_t substr_begin, size_t substr_end) {
1(paf 43:3): if(substr_begin==substr_end)
3(paf 44:3): return;
(paf 45:):
1(paf 46:3): size_t fragment_begin=0;
(paf 47:): size_t fragment_end;
1(paf 48:3): for(Array_iterator<element_type> i(src); ; fragment_begin=fragment_end) {
(paf 49:): const element_type& fragment=i.next();
1(paf 50:3): fragment_end=fragment_begin+fragment.length;
(paf 51:):
1(paf 52:3): if(fragment_begin<substr_begin) // not reached fragments which may include 'substr'?
(paf 53:): continue;
54:):
55:): // found first fragment including piece of 'substr'
56:): size_t piece_end=min(substr_end, fragment_end);
1(paf 57:3): *this+=Fragment(fragment.lang, piece_end-substr_begin);
(paf 58:):
59:): while(substr_end>fragment_end) { // are there more fragments including pieces of 'substr'?
1(paf 60:3): fragment_begin=fragment_end;
61:3):
(paf 62:): const element_type& fragment=i.next();
1(paf 63:3): fragment_end=fragment_begin+fragment.length;
(paf 64:):
65:): if(substr_end>fragment_end) // are there still more?
3(paf 66:3): append(Fragment(fragment.lang, fragment.length)); // appending whole fragment
(paf 67:): else { // no, it was last
3(paf 68:3): append(Fragment(fragment.lang, substr_end-fragment_begin));
69:3): return;
(paf 70:): }
71:): }
72:):
73:): break;
74:): }
75:): }
76:): /*
1(paf 77:3): void String::ArrayFragment::mid(ArrayFragment& result, size_t substr_begin, size_t substr_end) {
(paf 78:): }
79:): */
80:):
81:): // String methods
1.172.2.2 paf 82:
1.172.2.21.2. (paf 83:): String::String(const char* cstr, size_t helper_length, bool tainted): body(CORD_EMPTY) {
3(paf 84:3): append(cstr, helper_length, tainted?UL_TAINTED:UL_CLEAN);
1.1 paf 85: }
1.140 paf 86:
1.172.2.21.2. (paf 87:): String::String(const String& src): body(src.body), fragments(src.fragments) {}
1.28 paf 88:
1.172.2.21.2. 3(paf 89:3): String& String::append(const char* str, size_t helper_length, Untaint_lang lang) {
90:3): if(!str)
1.9 paf 91: return *this;
1.172.2.21.2. 3(paf 92:3): size_t use_length=helper_length?helper_length:strlen(str);
0(paf 93:3): if(!use_length)
1.9 paf 94: return *this;
1.122 paf 95:
1.172.2.21.2. 3(paf 96:3): body=CORD_cat_char_star(body, str, use_length);
0(paf 97:3): fragments+=Fragment(lang, use_length);
1.1 paf 98:
99: return *this;
100: }
101:
1.172.2.21.2. (paf 102:): static int CORD_batched_iter_fn_generic_hash_code(const char* s, void * client_data) {
103:): uint& result=*static_cast<uint*>(client_data);
104:): generic_hash_code(result, s);
105:): return 0;
106:): };
1.16 paf 107: uint String::hash_code() const {
1.7 paf 108: uint result=0;
1.172.2.21.2. (paf 109:): CORD_iter5(body, 0, 0, CORD_batched_iter_fn_generic_hash_code, &result);
1.5 paf 110: return result;
111: }
112:
1.172.2.21.2. 1(paf 113:3): /// @todo check in doc: whether it documents NOW bad situation "abc".mid(-1, 3) =were?="ab"
114:3): const String& String::mid(size_t substr_begin, size_t substr_end) const {
(paf 115:): String& result=*new String;
1.33 paf 116:
1.172.2.21.2. 1(paf 117:3): size_t self_length=length();
118:3): substr_begin=max(min(substr_begin, self_length), (size_t)0);
119:3): substr_end=min(max(substr_end, substr_begin), self_length);
120:3): if(substr_begin==substr_end)
(paf 121:): return result;
1.52 paf 122:
1.172.2.21.2. (paf 123:): // first: letters themselves
1(paf 124:3): result.body=CORD_substr(body, substr_begin, substr_end-substr_begin);
1.46 paf 125:
1.172.2.21.2. (paf 126:): // next: their langs
3(paf 127:3): result.fragments.append_positions(fragments, substr_begin, substr_end);
1.53 paf 128:
1.60 paf 129: // SAPI::log(pool(), "piece of '%s' from %d to %d is '%s'",
1.172.2.21.2. 1(paf 130:3): //cstr(), substr_begin, substr_end, result.cstr());
1.53 paf 131: return result;
1.54 paf 132: }
133:
1.172.2.21.2. (paf 134:): size_t String::pos(CORD substr,
135:): size_t this_offset, Untaint_lang lang) const {
136:): // first: letters themselves
1(paf 137:3): size_t substr_begin=CORD_str(body, this_offset, substr);
138:3): if(substr_begin==CORD_NOT_FOUND)
(paf 139:): return STRING_NOT_FOUND;
140:):
141:): // next: check the lang when specified
142:):
143:): if(lang==UL_UNSPECIFIED) // ignore lang?
1(paf 144:3): return substr_begin;
(paf 145:):
146:): // substr must be in one fragment, and fragments' lang must = lang
1(paf 147:3): size_t substr_end=substr_begin+CORD_len(substr);
148:3): size_t fragment_begin=0;
(paf 149:): size_t fragment_end;
1(paf 150:3): for(Array_iterator<ArrayFragment::element_type> i(fragments); i.has_next(); fragment_begin=fragment_end) {
(paf 151:): const Fragment& fragment=i.next();
1(paf 152:3): fragment_end=fragment_begin+fragment.length;
(paf 153:):
1(paf 154:3): if(substr_begin<fragment_begin) // not reached fragments which may include 'result'?
(paf 155:): continue;
2(paf 156:3): if(substr_begin>=fragment_end) // begin of substr OUT of current fragment?
157:3): continue;
(paf 158:):
159:): if(substr_end>fragment_end) // end of substr OUT of current fragment?
2(paf 160:3): throw Exception(0, // (*) see below
161:3): this,
162:3): "searching for '%s' starting from %ud problem: found begin in one fragment, but end in another",
163:3): CORD_to_const_char_star(substr), this_offset);
(paf 164:):
2(paf 165:3): if(fragment.lang==lang)
166:3): return substr_begin;
167:3): else { // bad lang...
168:3): /// WARNING: this possibly skips assert (*), but it's fast
169:3): substr_begin=CORD_str(body, fragment_end/*...search AFTER for more*/, substr);
170:3): if(substr_begin==CORD_NOT_FOUND)
171:3): return STRING_NOT_FOUND;
172:3):
173:3): size_t substr_end=substr_begin+CORD_len(substr);
174:3): // and continuing with next fragment
175:3): }
(paf 176:): }
177:):
178:): return STRING_NOT_FOUND;
1.58 paf 179: }
180:
1.172.2.21.2. (paf 181:): size_t String::pos(const String& substr,
182:): size_t this_offset, Untaint_lang lang) const {
183:): return pos(substr.body, this_offset, lang);
1.60 paf 184: }
185:
1.172.2.14 paf 186: void String::split(ArrayString& result,
1.172.2.21.2. (paf 187:): size_t& pos_after,
188:): const char* delim,
1.172.2.10 paf 189: Untaint_lang lang, int limit) {
1.172.2.21.2. (paf 190:): size_t self_length=length();
191:): if(size_t delim_length=strlen(delim)) {
1.60 paf 192: int pos_before;
193: // while we have 'delim'...
1.172.2.21.2. (paf 194:): for(; (pos_before=pos(delim, pos_after, lang))>=0 && limit; limit--) {
195:): result+=&mid(pos_after, pos_before);
196:): pos_after=pos_before+delim_length;
1.60 paf 197: }
198: // last piece
1.172.2.21.2. (paf 199:): if(pos_after<self_length && limit) {
200:): result+=&mid(pos_after, self_length);
201:): pos_after=self_length;
1.60 paf 202: }
203: } else { // empty delim
1.172.2.21.2. (paf 204:): result+=this;
205:): pos_after+=self_length;
1.60 paf 206: }
207: }
208:
1.172.2.14 paf 209: void String::split(ArrayString& result,
1.172.2.21.2. (paf 210:): size_t& pos_after,
1.60 paf 211: const String& delim, Untaint_lang lang,
1.172.2.21.2. (paf 212:): int limit) const {
213:): if(delim) {
1.60 paf 214: int pos_before;
215: // while we have 'delim'...
216: for(; (pos_before=pos(delim, pos_after, lang))>=0 && limit; limit--) {
1.172.2.21.2. (paf 217:): result+=&mid(pos_after, pos_before);
218:): pos_after=pos_before+delim.length();
1.60 paf 219: }
220: // last piece
1.172.2.21.2. (paf 221:): if(pos_after<length() && limit) {
222:): result+=&mid(pos_after, length());
223:): pos_after=length();
1.60 paf 224: }
225: } else { // empty delim
1.172.2.21.2. (paf 226:): result+=this;
227:): pos_after+=length();
1.60 paf 228: }
1.61 paf 229: }
230:
1.172.2.21.2. (paf 231:): static void regex_options(const String& options, int *result, bool& need_pre_post_match){
1.63 paf 232: struct Regex_option {
1.172.2.11 paf 233: const char* keyL;
234: const char* keyU;
1.63 paf 235: int clear, set;
236: int *result;
1.154 paf 237: bool *flag;
1.63 paf 238: } regex_option[]={
1.153 paf 239: {"i", "I", 0, PCRE_CASELESS, result}, // a=A
240: {"s", "S", 0, PCRE_DOTALL, result}, // \n\n$ [default]
241: {"x", "U", 0, PCRE_EXTENDED, result}, // whitespace in regex ignored
242: {"m", "M", PCRE_DOTALL, PCRE_MULTILINE, result}, // ^aaa\n$^bbb\n$
243: {"g", "G", 0, true, result+1}, // many rows
1.154 paf 244: {"'", 0, 0, 0, 0, &need_pre_post_match},
245: {0}
1.63 paf 246: };
1.171 paf 247: result[0]=PCRE_EXTRA | PCRE_DOTALL | PCRE_DOLLAR_ENDONLY;
1.63 paf 248: result[1]=0;
249:
250: if(options)
1.153 paf 251: for(Regex_option *o=regex_option; o->keyL; o++)
1.172.2.21.2. (paf 252:): if(options.pos(o->keyL)>=0
253:): || (o->keyU && options.pos(o->keyU)>=0)) {
1.154 paf 254: if(o->flag)
255: *o->flag=true;
256: else { // result
257: *o->result &= ~o->clear;
258: *o->result |= o->set;
259: }
1.63 paf 260: }
261: }
262:
1.172.2.21.2. (paf 263:): Table* String::match(Charset& source_charset,
1.172.2.4 paf 264: const String& regexp,
1.172.2.21.2. (paf 265:): const String& options,
1.172.2.4 paf 266: Row_action row_action, void *info,
1.172.2.20 paf 267: bool& just_matched) const {
1.172.2.21.2. (paf 268:): if(!regexp)
269:): throw Exception(0,
270:): 0,
1.73 paf 271: "regexp is empty");
1.154 paf 272:
1.172.2.21.2. (paf 273:): const char* pattern=regexp.cstr();
1.172.2.11 paf 274: const char* errptr;
1.62 paf 275: int erroffset;
1.172.2.21.2. (paf 276:): bool need_pre_post_match=false;
1.154 paf 277: int option_bits[2]; regex_options(options, option_bits, need_pre_post_match);
1.172.2.20 paf 278: bool global=option_bits[1]!=0;
1.172.2.21.2. (paf 279:): pcre *code=pcre_compile(pattern, option_bits[0],
1.62 paf 280: &errptr, &erroffset,
1.172.2.1 paf 281: source_charset.pcre_tables);
1.62 paf 282:
1.67 paf 283: if(!code)
1.172.2.21.2. (paf 284:): throw Exception(0,
285:): ®exp.mid(erroffset, regexp.length()),
1.74 paf 286: "regular expression syntax error - %s", errptr);
1.62 paf 287:
1.172.2.20 paf 288: int subpatterns=pcre_info(code, 0, 0);
289: if(subpatterns<0) {
1.100 parser 290: pcre_free(code);
1.149 paf 291: throw Exception(0,
1.172.2.21.2. (paf 292:): ®exp,
1.76 paf 293: "pcre_info error (%d)",
1.172.2.20 paf 294: subpatterns);
1.63 paf 295: }
296:
1.172.2.21.2. (paf 297:): const char* subject=cstr();
298:): size_t subject_length=strlen(subject);
299:): const int oveclength=(1/*match*/+MAX_MATCH_GROUPS)*3;
300:): int ovector[oveclength];
1.155 paf 301:
302: // create table
1.172.2.21.2. (paf 303:): Table* table=new Table(string_match_table_template);
1.63 paf 304:
1.64 paf 305: int exec_option_bits=0;
1.154 paf 306: int prestart=0;
307: int poststart=0;
1.172.2.21.2. (paf 308:): int postfinish=length();
1.63 paf 309: while(true) {
310: int exec_substrings=pcre_exec(code, 0,
1.172.2.21.2. (paf 311:): subject, subject_length, prestart,
312:): exec_option_bits, ovector, oveclength);
1.63 paf 313:
314: if(exec_substrings==PCRE_ERROR_NOMATCH) {
1.100 parser 315: pcre_free(code);
1.172.2.21.2. (paf 316:): row_action(table, 0/*last time, no raw*/, 0, 0, poststart, postfinish, info);
1.172.2.20 paf 317: if(global || subpatterns)
318: return table; // global or with subpatterns=true+result
319: else {
1.172.2.21.2. (paf 320:): just_matched=false; return 0; // not global=no result
1.172.2.20 paf 321: }
1.63 paf 322: }
323:
324: if(exec_substrings<0) {
1.100 parser 325: pcre_free(code);
1.172.2.21.2. (paf 326:): throw Exception(0,
327:): ®exp,
1.76 paf 328: "regular expression execute error (%d)",
1.63 paf 329: exec_substrings);
330: }
331:
1.154 paf 332: int prefinish=ovector[0];
333: poststart=ovector[1];
1.172.2.21.2. (paf 334:): ArrayString* row=new ArrayString;
1.172.2.4 paf 335: if(need_pre_post_match) {
1.172.2.21.2. (paf 336:): *row+=&mid(0, prefinish); // .prematch column value
337:): *row+=&mid(prefinish, poststart); // .match
338:): *row+=&mid(poststart, postfinish); // .postmatch
1.172.2.4 paf 339: } else {
1.172.2.21.2. (paf 340:): *row+=0; // .prematch column value
341:): *row+=0; // .match
342:): *row+=0; // .postmatch
1.172.2.4 paf 343: }
1.63 paf 344:
345: for(int i=1; i<exec_substrings; i++) {
1.69 paf 346: // -1:-1 case handled peacefully by mid() itself
1.172.2.21.2. (paf 347:): *row+=&mid(ovector[i*2+0], ovector[i*2+1]); // .i column value
1.63 paf 348: }
349:
1.172.2.20 paf 350: row_action(table, row, prestart, prefinish, poststart, postfinish, info);
1.63 paf 351:
1.172.2.20 paf 352: if(!global || prestart==poststart) { // not global | going to hang
1.100 parser 353: pcre_free(code);
1.172.2.21.2. (paf 354:): row_action(table, 0/*last time, no row*/, 0, 0, poststart, postfinish, info);
1.172.2.20 paf 355: return table;
1.63 paf 356: }
1.154 paf 357: prestart=poststart;
1.63 paf 358:
359: /*
360: if(option_bits[0] & PCRE_MULTILINE)
1.64 paf 361: exec_option_bits|=PCRE_NOTBOL; // start of subject+startoffset not BOL
1.63 paf 362: */
363: }
1.82 parser 364: }
365:
1.172.2.21.2. (paf 366:): String& String::change_case(Charset& source_charset, Change_case_kind kind) {
367:): String& result=*new String();
1.172.2.4 paf 368:
1.172.2.1 paf 369: const unsigned char *tables=source_charset.pcre_tables;
1.82 parser 370:
371: const unsigned char *a;
372: const unsigned char *b;
373: switch(kind) {
374: case CC_UPPER:
375: a=tables+lcc_offset;
376: b=tables+fcc_offset;
377: break;
378: case CC_LOWER:
379: a=tables+lcc_offset;
380: b=0;
381: break;
382: default:
1.172.2.21.2. (paf 383:): throw Exception(0,
384:): this,
1.82 parser 385: "unknown change case kind #%d",
386: static_cast<int>(kind)); // never
387: a=b=0; // calm, compiler
388: break; // never
389: }
390:
1.172.2.21.2. (paf 391:): char* new_cstr=cstrm();
392:): char *dest=new_cstr;
393:): unsigned char index;
394:): for(const char* current=new_cstr; index=(unsigned char)*current; current++) {
395:): unsigned char c=a[index];
396:): if(b)
397:): c=b[c];
1.82 parser 398:
1.172.2.21.2. (paf 399:): *dest++=(char)c;
400:): }
401:): result.body=new_cstr;
3(paf 402:3): result.fragments.append(fragments);
1.89 parser 403:
1.101 parser 404: return result;
405: }
406:
1.172.2.21.2. (paf 407:): const String& String::replace(const Dictionary& dict) const {
408:): String& result=*new String();
409:): const char* old_cstr=cstr();
1(paf 410:3): const char* prematch_begin=old_cstr;
(paf 411:):
412:): for(const char* current=old_cstr; *current; ) {
1(paf 413:3): if(Table::element_type row=dict.first_that_begins(current)) {
(paf 414:): // prematch
1(paf 415:3): if(size_t prematch_length=current-prematch_begin) {
416:3): result.body=CORD_cat_char_star(result.body, prematch_begin, prematch_length);
4(paf 417:3): result.fragments.append_positions(fragments, prematch_begin-old_cstr, current-old_cstr);
1.101 parser 418: }
419:
1.172.2.21.2. (paf 420:): // match
421:):
422:): const String* a=row->get(0);
1(paf 423:3): // skip 'a' in 'current'; move prematch_begin
424:3): current+=a->length(); prematch_begin=current;
1.170 paf 425:
1.172.2.21.2. (paf 426:): if(row->count()>1) { // are there any b?
427:): const String* b=row->get(1);
428:): if(size_t b_length=b->length()) {
429:): result.body=CORD_cat(result.body, b->cstr_to_cord());
3(paf 430:3): result.fragments.append(b->fragments);
(paf 431:): }
432:): }
433:): } else // simply advance
434:): current++;
435:): }
1.156 paf 436:
1.172.2.21.2. (paf 437:): // postmatch
3(paf 438:3): if(size_t postmatch_length=current-prematch_begin) {
439:3): result.body=CORD_cat_char_star(result.body, prematch_begin, postmatch_length);
4(paf 440:3): result.fragments.append_positions(fragments, prematch_begin-old_cstr, current-old_cstr);
(paf 441:): }
1.156 paf 442:
1.89 parser 443: return result;
444: }
445:
1.172.2.21.2. (paf 446:): double String::as_double() const {
1.89 parser 447: double result;
1.172.2.21.2. (paf 448:): const char *str=cstr();
449:):
450:): while(*str && isspace(*str))
451:): str++;
452:): if(!*str)
1.162 paf 453: return 0;
1.161 paf 454:
1.102 parser 455: char *error_pos;
1.89 parser 456: // 0xABC
1.172.2.21.2. (paf 457:): if(str[0]=='0')
458:): if(str[1]=='x' || str[1]=='X')
459:): result=(double)(unsigned long)strtol(str, &error_pos, 0);
1.99 parser 460: else
1.172.2.21.2. (paf 461:): result=(double)strtod(str+1/*skip leading 0*/, &error_pos);
1.89 parser 462: else
1.172.2.21.2. (paf 463:): result=(double)strtod(str, &error_pos);
1.89 parser 464:
1.159 paf 465: while(char c=*error_pos++)
466: if(!isspace(c))
467: throw Exception("number.format",
1.172.2.21.2. (paf 468:): this,
1.159 paf 469: "invalid number (double)");
1.89 parser 470:
471: return result;
472: }
1.172.2.21.2. (paf 473:): int String::as_int() const {
1.89 parser 474: int result;
1.172.2.21.2. (paf 475:): const char *str=cstr();
476:):
477:): while(*str && isspace(*str))
478:): str++;
479:): if(!*str)
1.162 paf 480: return 0;
1.161 paf 481:
1.102 parser 482: char *error_pos;
1.89 parser 483: // 0xABC
1.172.2.21.2. (paf 484:): if(str[0]=='0')
485:): if(str[1]=='x' || str[1]=='X')
486:): result=(int)(unsigned long)strtol(str, &error_pos, 0);
1.99 parser 487: else
1.172.2.21.2. (paf 488:): result=(int)strtol(str+1/*skip leading 0*/, &error_pos, 0);
1.89 parser 489: else
1.172.2.21.2. (paf 490:): result=(int)strtol(str, &error_pos, 0);
1.89 parser 491:
1.159 paf 492: while(char c=*error_pos++)
493: if(!isspace(c))
494: throw Exception("number.format",
1.172.2.21.2. (paf 495:): this,
1.159 paf 496: "invalid number (int)");
1.82 parser 497:
498: return result;
1.61 paf 499: }
1.113 parser 500:
1.172.2.4 paf 501: inline void uint2uchars(uint word, uchar *bytes) {
502: bytes[0]=word&0xFF;
503: bytes[1]=(word>>8)&0xFF;
504: bytes[2]=(word>>16)&0xFF;
505: bytes[3]=(word>>24)&0xFF;
506: }
507: inline uint uchars2uint(uchar *bytes) {
508: return bytes[3]<<24
509: | bytes[2]<<16
510: | bytes[1]<<8
511: | bytes[0];
512: }
513:
1.172.2.21.2. (paf 514:): static int CORD_batched_iter_fn_append(const char* s, void* client_data) {
515:): char*& cur=*static_cast<char**>(client_data);
516:):
517:): size_t length=strlen(s);
518:): memcpy(cur, s, length); cur+=length;
519:): return 0;
520:): };
521:): String::C String::serialize(size_t prolog_length) const {
522:): size_t buf_length=
523:): prolog_length
524:): +fragments.count()*(sizeof(Untaint_lang)+sizeof(size_t))
525:): +length();
526:): C result(new(PointerFreeGC) char[buf_length], buf_length);
527:):
528:): // 1: prolog
529:): char *cur=result.str+prolog_length;
530:):
531:):
532:): // 2: fragments.count
533:): size_t fragments_count=fragments.count();
534:): memcpy(cur, &fragments_count, sizeof(fragments_count)); cur+=sizeof(fragments_count);
1.113 parser 535:
1.172.2.21.2. (paf 536:): // 3: lang info
537:): for(Array_iterator<ArrayFragment::element_type> i(fragments); i.has_next(); ) {
538:): const Fragment& fragment=i.next();
1.123 paf 539: // lang
1.172.2.21.2. (paf 540:): memcpy(cur, &fragment.lang, sizeof(fragment.lang)); cur+=sizeof(fragment.lang);
541:): // length
542:): memcpy(cur, &fragment.length, sizeof(fragment.length)); cur+=sizeof(fragment.length);
543:): }
544:):
545:): // 4: letters
546:): CORD_iter5(body, 0, 0, CORD_batched_iter_fn_append, &cur);
547:):
548:): return result;
1.113 parser 549: }
1.172.2.21.2. (paf 550:): bool String::deserialize(size_t prolog_length, void *buf, size_t buf_length, const char* file) {
551:): if(buf_length<=prolog_length)
1.148 paf 552: return false;
1.172.2.21.2. (paf 553:): buf_length-=prolog_length;
1.113 parser 554:
1.172.2.21.2. (paf 555:): // 1: prolog
556:): const char* cur=(const char* )buf+prolog_length;
1.148 paf 557:
1.172.2.21.2. (paf 558:): // 2: fragments.count
559:): if(buf_length<sizeof(size_t)) // fragments.count don't fit?
560:): return false;
561:): size_t fragments_count=*reinterpret_cast<const size_t*>(cur); cur+=sizeof(size_t);
562:): buf_length-=sizeof(size_t);
1.128 paf 563:
1.172.2.21.2. (paf 564:): // 3: lang info
565:): size_t total_length=0;
566:): for(size_t f=0; f<fragments_count; f++) {
567:): size_t piece_length=sizeof(Untaint_lang)+sizeof(size_t);
568:): if(buf_length<piece_length) // lang+length
1.148 paf 569: return false;
570:
1.172.2.21.2. (paf 571:): Untaint_lang lang=*reinterpret_cast<const Untaint_lang *>(cur); cur+=sizeof(Untaint_lang);
572:): size_t fragment_length=*reinterpret_cast<const size_t*>(cur); cur+=sizeof(size_t);
573:): fragments+=Fragment(lang, fragment_length);
574:): total_length+=fragment_length;
1.113 parser 575:
1.172.2.21.2. (paf 576:): buf_length-=piece_length;
1.113 parser 577: }
1.172.2.21.2. (paf 578:):
579:): // 4: letters
580:): if(buf_length!=total_length)
581:): return false;
582:):
583:): body=CORD_cat_char_star(CORD_EMPTY, cur, buf_length);
584:):
1.148 paf 585: return true;
1.113 parser 586: }
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