Annotation of parser3/src/main/pa_string.C, revision 1.172.2.21.2.17
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. 7(paf 8:3): static const char* IDENT_STRING_C="$Date: 2003/03/21 15:00:36 $";
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:):
5(paf 81:3): // StringSimple methods
82:3):
83:3):
84:3): static int CORD_batched_iter_fn_generic_hash_code(const char* s, void * client_data) {
85:3): uint& result=*static_cast<uint*>(client_data);
86:3): generic_hash_code(result, s);
87:3): return 0;
88:3): };
7(paf 89:3): uint StringBody::hash_code() const {
5(paf 90:3): uint result=0;
91:3): CORD_iter5(body, 0, 0, CORD_batched_iter_fn_generic_hash_code, &result);
92:3): return result;
93:3): }
94:3):
(paf 95:): // String methods
1.172.2.2 paf 96:
1.172.2.21.2. (paf 97:): String::String(const char* cstr, size_t helper_length, bool tainted): body(CORD_EMPTY) {
5(paf 98:3): append(cstr, helper_length, tainted?L_TAINTED:L_CLEAN);
1.1 paf 99: }
1.140 paf 100:
1.172.2.21.2. (paf 101:): String::String(const String& src): body(src.body), fragments(src.fragments) {}
1.28 paf 102:
1.172.2.21.2. 6(paf 103:3): String& String::append(const char* str, size_t length, Language lang) {
3(paf 104:3): if(!str)
1.9 paf 105: return *this;
1.172.2.21.2. 6(paf 106:3): size_t use_length=length?length:strlen(str);
0(paf 107:3): if(!use_length)
1.9 paf 108: return *this;
1.122 paf 109:
1.172.2.21.2. 7(paf 110:3): body.append(str, use_length);
0(paf 111:3): fragments+=Fragment(lang, use_length);
1.1 paf 112:
113: return *this;
114: }
115:
1.172.2.21.2. 1(paf 116:3): /// @todo check in doc: whether it documents NOW bad situation "abc".mid(-1, 3) =were?="ab"
117:3): const String& String::mid(size_t substr_begin, size_t substr_end) const {
(paf 118:): String& result=*new String;
1.33 paf 119:
1.172.2.21.2. 1(paf 120:3): size_t self_length=length();
121:3): substr_begin=max(min(substr_begin, self_length), (size_t)0);
122:3): substr_end=min(max(substr_end, substr_begin), self_length);
123:3): if(substr_begin==substr_end)
(paf 124:): return result;
1.52 paf 125:
1.172.2.21.2. (paf 126:): // first: letters themselves
7(paf 127:3): result.body=body.mid(substr_begin, substr_end-substr_begin);
1.46 paf 128:
1.172.2.21.2. (paf 129:): // next: their langs
3(paf 130:3): result.fragments.append_positions(fragments, substr_begin, substr_end);
1.53 paf 131:
1.172.2.21.2. 5(paf 132:3): // SAPI::log("piece of '%s' from %d to %d is '%s'",
1(paf 133:3): //cstr(), substr_begin, substr_end, result.cstr());
1.53 paf 134: return result;
1.54 paf 135: }
136:
1.172.2.21.2. 7(paf 137:3): size_t String::pos(const StringBody substr,
5(paf 138:3): size_t this_offset, Language lang) const {
(paf 139:): // first: letters themselves
7(paf 140:3): size_t substr_begin=body.pos(this_offset, substr);
1(paf 141:3): if(substr_begin==CORD_NOT_FOUND)
(paf 142:): return STRING_NOT_FOUND;
143:):
144:): // next: check the lang when specified
145:):
5(paf 146:3): if(lang==L_UNSPECIFIED) // ignore lang?
1(paf 147:3): return substr_begin;
(paf 148:):
149:): // substr must be in one fragment, and fragments' lang must = lang
7(paf 150:3): size_t substr_end=substr_begin+substr.length();
1(paf 151:3): size_t fragment_begin=0;
(paf 152:): size_t fragment_end;
1(paf 153:3): for(Array_iterator<ArrayFragment::element_type> i(fragments); i.has_next(); fragment_begin=fragment_end) {
(paf 154:): const Fragment& fragment=i.next();
1(paf 155:3): fragment_end=fragment_begin+fragment.length;
(paf 156:):
1(paf 157:3): if(substr_begin<fragment_begin) // not reached fragments which may include 'result'?
(paf 158:): continue;
2(paf 159:3): if(substr_begin>=fragment_end) // begin of substr OUT of current fragment?
160:3): continue;
(paf 161:):
162:): if(substr_end>fragment_end) // end of substr OUT of current fragment?
2(paf 163:3): throw Exception(0, // (*) see below
164:3): this,
165:3): "searching for '%s' starting from %ud problem: found begin in one fragment, but end in another",
7(paf 166:3): substr.cstr(), this_offset);
(paf 167:):
2(paf 168:3): if(fragment.lang==lang)
169:3): return substr_begin;
170:3): else { // bad lang...
171:3): /// WARNING: this possibly skips assert (*), but it's fast
7(paf 172:3): substr_begin=body.pos(fragment_end/*...search AFTER for more*/, substr);
2(paf 173:3): if(substr_begin==CORD_NOT_FOUND)
174:3): return STRING_NOT_FOUND;
175:3):
7(paf 176:3): size_t substr_end=substr_begin+substr.length();
2(paf 177:3): // and continuing with next fragment
178:3): }
(paf 179:): }
180:):
181:): return STRING_NOT_FOUND;
1.58 paf 182: }
183:
1.172.2.21.2. (paf 184:): size_t String::pos(const String& substr,
5(paf 185:3): size_t this_offset, Language lang) const {
(paf 186:): return pos(substr.body, this_offset, lang);
1.60 paf 187: }
188:
1.172.2.14 paf 189: void String::split(ArrayString& result,
1.172.2.21.2. (paf 190:): size_t& pos_after,
191:): const char* delim,
5(paf 192:3): Language lang, int limit) {
(paf 193:): size_t self_length=length();
194:): if(size_t delim_length=strlen(delim)) {
1.60 paf 195: int pos_before;
196: // while we have 'delim'...
1.172.2.21.2. (paf 197:): for(; (pos_before=pos(delim, pos_after, lang))>=0 && limit; limit--) {
198:): result+=&mid(pos_after, pos_before);
199:): pos_after=pos_before+delim_length;
1.60 paf 200: }
201: // last piece
1.172.2.21.2. (paf 202:): if(pos_after<self_length && limit) {
203:): result+=&mid(pos_after, self_length);
204:): pos_after=self_length;
1.60 paf 205: }
206: } else { // empty delim
1.172.2.21.2. (paf 207:): result+=this;
208:): pos_after+=self_length;
1.60 paf 209: }
210: }
211:
1.172.2.14 paf 212: void String::split(ArrayString& result,
1.172.2.21.2. (paf 213:): size_t& pos_after,
5(paf 214:3): const String& delim, Language lang,
(paf 215:): int limit) const {
5(paf 216:3): if(!delim.is_empty()) {
1.60 paf 217: int pos_before;
218: // while we have 'delim'...
219: for(; (pos_before=pos(delim, pos_after, lang))>=0 && limit; limit--) {
1.172.2.21.2. (paf 220:): result+=&mid(pos_after, pos_before);
221:): pos_after=pos_before+delim.length();
1.60 paf 222: }
223: // last piece
1.172.2.21.2. (paf 224:): if(pos_after<length() && limit) {
225:): result+=&mid(pos_after, length());
226:): pos_after=length();
1.60 paf 227: }
228: } else { // empty delim
1.172.2.21.2. (paf 229:): result+=this;
230:): pos_after+=length();
1.60 paf 231: }
1.61 paf 232: }
233:
1.172.2.21.2. (paf 234:): static void regex_options(const String& options, int *result, bool& need_pre_post_match){
1.63 paf 235: struct Regex_option {
1.172.2.11 paf 236: const char* keyL;
237: const char* keyU;
1.63 paf 238: int clear, set;
239: int *result;
1.154 paf 240: bool *flag;
1.63 paf 241: } regex_option[]={
1.153 paf 242: {"i", "I", 0, PCRE_CASELESS, result}, // a=A
243: {"s", "S", 0, PCRE_DOTALL, result}, // \n\n$ [default]
244: {"x", "U", 0, PCRE_EXTENDED, result}, // whitespace in regex ignored
245: {"m", "M", PCRE_DOTALL, PCRE_MULTILINE, result}, // ^aaa\n$^bbb\n$
246: {"g", "G", 0, true, result+1}, // many rows
1.154 paf 247: {"'", 0, 0, 0, 0, &need_pre_post_match},
248: {0}
1.63 paf 249: };
1.171 paf 250: result[0]=PCRE_EXTRA | PCRE_DOTALL | PCRE_DOLLAR_ENDONLY;
1.63 paf 251: result[1]=0;
252:
1.172.2.21.2. 5(paf 253:3): if(!options.is_empty())
1.153 paf 254: for(Regex_option *o=regex_option; o->keyL; o++)
1.172.2.21.2. (paf 255:): if(options.pos(o->keyL)>=0
256:): || (o->keyU && options.pos(o->keyU)>=0)) {
1.154 paf 257: if(o->flag)
258: *o->flag=true;
259: else { // result
260: *o->result &= ~o->clear;
261: *o->result |= o->set;
262: }
1.63 paf 263: }
264: }
265:
1.172.2.21.2. (paf 266:): Table* String::match(Charset& source_charset,
1.172.2.4 paf 267: const String& regexp,
1.172.2.21.2. (paf 268:): const String& options,
1.172.2.4 paf 269: Row_action row_action, void *info,
1.172.2.20 paf 270: bool& just_matched) const {
1.172.2.21.2. 5(paf 271:3): if(regexp.is_empty())
(paf 272:): throw Exception(0,
273:): 0,
1.73 paf 274: "regexp is empty");
1.154 paf 275:
1.172.2.21.2. (paf 276:): const char* pattern=regexp.cstr();
1.172.2.11 paf 277: const char* errptr;
1.62 paf 278: int erroffset;
1.172.2.21.2. (paf 279:): bool need_pre_post_match=false;
1.154 paf 280: int option_bits[2]; regex_options(options, option_bits, need_pre_post_match);
1.172.2.20 paf 281: bool global=option_bits[1]!=0;
1.172.2.21.2. (paf 282:): pcre *code=pcre_compile(pattern, option_bits[0],
1.62 paf 283: &errptr, &erroffset,
1.172.2.1 paf 284: source_charset.pcre_tables);
1.62 paf 285:
1.67 paf 286: if(!code)
1.172.2.21.2. (paf 287:): throw Exception(0,
288:): ®exp.mid(erroffset, regexp.length()),
1.74 paf 289: "regular expression syntax error - %s", errptr);
1.62 paf 290:
1.172.2.20 paf 291: int subpatterns=pcre_info(code, 0, 0);
292: if(subpatterns<0) {
1.100 parser 293: pcre_free(code);
1.149 paf 294: throw Exception(0,
1.172.2.21.2. (paf 295:): ®exp,
1.76 paf 296: "pcre_info error (%d)",
1.172.2.20 paf 297: subpatterns);
1.63 paf 298: }
299:
1.172.2.21.2. (paf 300:): const char* subject=cstr();
301:): size_t subject_length=strlen(subject);
302:): const int oveclength=(1/*match*/+MAX_MATCH_GROUPS)*3;
303:): int ovector[oveclength];
1.155 paf 304:
305: // create table
1.172.2.21.2. (paf 306:): Table* table=new Table(string_match_table_template);
1.63 paf 307:
1.64 paf 308: int exec_option_bits=0;
1.154 paf 309: int prestart=0;
310: int poststart=0;
1.172.2.21.2. (paf 311:): int postfinish=length();
1.63 paf 312: while(true) {
313: int exec_substrings=pcre_exec(code, 0,
1.172.2.21.2. (paf 314:): subject, subject_length, prestart,
315:): exec_option_bits, ovector, oveclength);
1.63 paf 316:
317: if(exec_substrings==PCRE_ERROR_NOMATCH) {
1.100 parser 318: pcre_free(code);
1.172.2.21.2. (paf 319:): row_action(table, 0/*last time, no raw*/, 0, 0, poststart, postfinish, info);
1.172.2.20 paf 320: if(global || subpatterns)
321: return table; // global or with subpatterns=true+result
322: else {
1.172.2.21.2. (paf 323:): just_matched=false; return 0; // not global=no result
1.172.2.20 paf 324: }
1.63 paf 325: }
326:
327: if(exec_substrings<0) {
1.100 parser 328: pcre_free(code);
1.172.2.21.2. (paf 329:): throw Exception(0,
330:): ®exp,
1.76 paf 331: "regular expression execute error (%d)",
1.63 paf 332: exec_substrings);
333: }
334:
1.154 paf 335: int prefinish=ovector[0];
336: poststart=ovector[1];
1.172.2.21.2. (paf 337:): ArrayString* row=new ArrayString;
1.172.2.4 paf 338: if(need_pre_post_match) {
1.172.2.21.2. (paf 339:): *row+=&mid(0, prefinish); // .prematch column value
340:): *row+=&mid(prefinish, poststart); // .match
341:): *row+=&mid(poststart, postfinish); // .postmatch
1.172.2.4 paf 342: } else {
1.172.2.21.2. (paf 343:): *row+=0; // .prematch column value
344:): *row+=0; // .match
345:): *row+=0; // .postmatch
1.172.2.4 paf 346: }
1.63 paf 347:
348: for(int i=1; i<exec_substrings; i++) {
1.69 paf 349: // -1:-1 case handled peacefully by mid() itself
1.172.2.21.2. (paf 350:): *row+=&mid(ovector[i*2+0], ovector[i*2+1]); // .i column value
1.63 paf 351: }
352:
1.172.2.20 paf 353: row_action(table, row, prestart, prefinish, poststart, postfinish, info);
1.63 paf 354:
1.172.2.20 paf 355: if(!global || prestart==poststart) { // not global | going to hang
1.100 parser 356: pcre_free(code);
1.172.2.21.2. (paf 357:): row_action(table, 0/*last time, no row*/, 0, 0, poststart, postfinish, info);
1.172.2.20 paf 358: return table;
1.63 paf 359: }
1.154 paf 360: prestart=poststart;
1.63 paf 361:
362: /*
363: if(option_bits[0] & PCRE_MULTILINE)
1.64 paf 364: exec_option_bits|=PCRE_NOTBOL; // start of subject+startoffset not BOL
1.63 paf 365: */
366: }
1.82 parser 367: }
368:
1.172.2.21.2. (paf 369:): String& String::change_case(Charset& source_charset, Change_case_kind kind) {
370:): String& result=*new String();
1.172.2.4 paf 371:
1.172.2.1 paf 372: const unsigned char *tables=source_charset.pcre_tables;
1.82 parser 373:
374: const unsigned char *a;
375: const unsigned char *b;
376: switch(kind) {
377: case CC_UPPER:
378: a=tables+lcc_offset;
379: b=tables+fcc_offset;
380: break;
381: case CC_LOWER:
382: a=tables+lcc_offset;
383: b=0;
384: break;
385: default:
1.172.2.21.2. (paf 386:): throw Exception(0,
387:): this,
1.82 parser 388: "unknown change case kind #%d",
389: static_cast<int>(kind)); // never
390: a=b=0; // calm, compiler
391: break; // never
392: }
393:
1.172.2.21.2. (paf 394:): char* new_cstr=cstrm();
395:): char *dest=new_cstr;
396:): unsigned char index;
397:): for(const char* current=new_cstr; index=(unsigned char)*current; current++) {
398:): unsigned char c=a[index];
399:): if(b)
400:): c=b[c];
1.82 parser 401:
1.172.2.21.2. (paf 402:): *dest++=(char)c;
403:): }
404:): result.body=new_cstr;
3(paf 405:3): result.fragments.append(fragments);
1.89 parser 406:
1.101 parser 407: return result;
408: }
409:
1.172.2.21.2. (paf 410:): const String& String::replace(const Dictionary& dict) const {
411:): String& result=*new String();
412:): const char* old_cstr=cstr();
1(paf 413:3): const char* prematch_begin=old_cstr;
(paf 414:):
415:): for(const char* current=old_cstr; *current; ) {
1(paf 416:3): if(Table::element_type row=dict.first_that_begins(current)) {
(paf 417:): // prematch
1(paf 418:3): if(size_t prematch_length=current-prematch_begin) {
7(paf 419:3): result.body.append(prematch_begin, prematch_length);
4(paf 420:3): result.fragments.append_positions(fragments, prematch_begin-old_cstr, current-old_cstr);
1.101 parser 421: }
422:
1.172.2.21.2. (paf 423:): // match
424:):
425:): const String* a=row->get(0);
1(paf 426:3): // skip 'a' in 'current'; move prematch_begin
427:3): current+=a->length(); prematch_begin=current;
1.170 paf 428:
1.172.2.21.2. (paf 429:): if(row->count()>1) { // are there any b?
430:): const String* b=row->get(1);
431:): if(size_t b_length=b->length()) {
7(paf 432:3): result.body.append(b->cstr_to_string_body());
3(paf 433:3): result.fragments.append(b->fragments);
(paf 434:): }
435:): }
436:): } else // simply advance
437:): current++;
438:): }
1.156 paf 439:
1.172.2.21.2. (paf 440:): // postmatch
3(paf 441:3): if(size_t postmatch_length=current-prematch_begin) {
7(paf 442:3): result.body.append(prematch_begin, postmatch_length);
4(paf 443:3): result.fragments.append_positions(fragments, prematch_begin-old_cstr, current-old_cstr);
(paf 444:): }
1.156 paf 445:
1.89 parser 446: return result;
447: }
448:
1.172.2.21.2. (paf 449:): double String::as_double() const {
1.89 parser 450: double result;
1.172.2.21.2. (paf 451:): const char *str=cstr();
452:):
453:): while(*str && isspace(*str))
454:): str++;
455:): if(!*str)
1.162 paf 456: return 0;
1.161 paf 457:
1.102 parser 458: char *error_pos;
1.89 parser 459: // 0xABC
1.172.2.21.2. (paf 460:): if(str[0]=='0')
461:): if(str[1]=='x' || str[1]=='X')
462:): result=(double)(unsigned long)strtol(str, &error_pos, 0);
1.99 parser 463: else
1.172.2.21.2. (paf 464:): result=(double)strtod(str+1/*skip leading 0*/, &error_pos);
1.89 parser 465: else
1.172.2.21.2. (paf 466:): result=(double)strtod(str, &error_pos);
1.89 parser 467:
1.159 paf 468: while(char c=*error_pos++)
469: if(!isspace(c))
470: throw Exception("number.format",
1.172.2.21.2. (paf 471:): this,
1.159 paf 472: "invalid number (double)");
1.89 parser 473:
474: return result;
475: }
1.172.2.21.2. (paf 476:): int String::as_int() const {
1.89 parser 477: int result;
1.172.2.21.2. (paf 478:): const char *str=cstr();
479:):
480:): while(*str && isspace(*str))
481:): str++;
482:): if(!*str)
1.162 paf 483: return 0;
1.161 paf 484:
1.102 parser 485: char *error_pos;
1.89 parser 486: // 0xABC
1.172.2.21.2. (paf 487:): if(str[0]=='0')
488:): if(str[1]=='x' || str[1]=='X')
489:): result=(int)(unsigned long)strtol(str, &error_pos, 0);
1.99 parser 490: else
1.172.2.21.2. (paf 491:): result=(int)strtol(str+1/*skip leading 0*/, &error_pos, 0);
1.89 parser 492: else
1.172.2.21.2. (paf 493:): result=(int)strtol(str, &error_pos, 0);
1.89 parser 494:
1.159 paf 495: while(char c=*error_pos++)
496: if(!isspace(c))
497: throw Exception("number.format",
1.172.2.21.2. (paf 498:): this,
1.159 paf 499: "invalid number (int)");
1.82 parser 500:
501: return result;
1.61 paf 502: }
1.113 parser 503:
1.172.2.4 paf 504: inline void uint2uchars(uint word, uchar *bytes) {
505: bytes[0]=word&0xFF;
506: bytes[1]=(word>>8)&0xFF;
507: bytes[2]=(word>>16)&0xFF;
508: bytes[3]=(word>>24)&0xFF;
509: }
510: inline uint uchars2uint(uchar *bytes) {
511: return bytes[3]<<24
512: | bytes[2]<<16
513: | bytes[1]<<8
514: | bytes[0];
515: }
516:
1.172.2.21.2. 7(paf 517:3): static int serialize_body_piece(const char* s, char** cur) {
(paf 518:): size_t length=strlen(s);
7(paf 519:3): memcpy(*cur, s, length); *cur+=length;
(paf 520:): return 0;
521:): };
522:): String::C String::serialize(size_t prolog_length) const {
523:): size_t buf_length=
524:): prolog_length
5(paf 525:3): +fragments.count()*(sizeof(Language)+sizeof(size_t))
(paf 526:): +length();
527:): C result(new(PointerFreeGC) char[buf_length], buf_length);
528:):
529:): // 1: prolog
530:): char *cur=result.str+prolog_length;
531:):
532:):
533:): // 2: fragments.count
534:): size_t fragments_count=fragments.count();
535:): memcpy(cur, &fragments_count, sizeof(fragments_count)); cur+=sizeof(fragments_count);
1.113 parser 536:
1.172.2.21.2. (paf 537:): // 3: lang info
538:): for(Array_iterator<ArrayFragment::element_type> i(fragments); i.has_next(); ) {
539:): const Fragment& fragment=i.next();
1.123 paf 540: // lang
1.172.2.21.2. (paf 541:): memcpy(cur, &fragment.lang, sizeof(fragment.lang)); cur+=sizeof(fragment.lang);
542:): // length
543:): memcpy(cur, &fragment.length, sizeof(fragment.length)); cur+=sizeof(fragment.length);
544:): }
545:):
546:): // 4: letters
7(paf 547:3): body.for_each(serialize_body_piece, &cur);
(paf 548:):
549:): return result;
1.113 parser 550: }
1.172.2.21.2. (paf 551:): bool String::deserialize(size_t prolog_length, void *buf, size_t buf_length, const char* file) {
552:): if(buf_length<=prolog_length)
1.148 paf 553: return false;
1.172.2.21.2. (paf 554:): buf_length-=prolog_length;
1.113 parser 555:
1.172.2.21.2. (paf 556:): // 1: prolog
557:): const char* cur=(const char* )buf+prolog_length;
1.148 paf 558:
1.172.2.21.2. (paf 559:): // 2: fragments.count
560:): if(buf_length<sizeof(size_t)) // fragments.count don't fit?
561:): return false;
562:): size_t fragments_count=*reinterpret_cast<const size_t*>(cur); cur+=sizeof(size_t);
563:): buf_length-=sizeof(size_t);
1.128 paf 564:
1.172.2.21.2. (paf 565:): // 3: lang info
566:): size_t total_length=0;
567:): for(size_t f=0; f<fragments_count; f++) {
5(paf 568:3): size_t piece_length=sizeof(Language)+sizeof(size_t);
(paf 569:): if(buf_length<piece_length) // lang+length
1.148 paf 570: return false;
571:
1.172.2.21.2. 5(paf 572:3): Language lang=*reinterpret_cast<const Language *>(cur); cur+=sizeof(Language);
(paf 573:): size_t fragment_length=*reinterpret_cast<const size_t*>(cur); cur+=sizeof(size_t);
574:): fragments+=Fragment(lang, fragment_length);
575:): total_length+=fragment_length;
1.113 parser 576:
1.172.2.21.2. (paf 577:): buf_length-=piece_length;
1.113 parser 578: }
1.172.2.21.2. (paf 579:):
580:): // 4: letters
581:): if(buf_length!=total_length)
582:): return false;
583:):
584:): body=CORD_cat_char_star(CORD_EMPTY, cur, buf_length);
585:):
1.148 paf 586: return true;
1.113 parser 587: }
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