Annotation of parser3/src/main/pa_string.C, revision 1.172.2.21.2.34
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: 2003/03/27 10:53:01 $";
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) {
3(paf 49:3): const element_type fragment=i.next();
1(paf 50:3): fragment_end=fragment_begin+fragment.length;
(paf 51:):
5(paf 52:3): if(substr_begin<fragment_begin) // not reached fragments which may include 'substr'?
(paf 53:): continue;
54:):
6(paf 55:3): if(substr_begin>fragment_end) // not reached fragments which may include 'substr'?
56:3): continue;
57:3):
(paf 58:): // found first fragment including piece of 'substr'
59:): size_t piece_end=min(substr_end, fragment_end);
1(paf 60:3): *this+=Fragment(fragment.lang, piece_end-substr_begin);
(paf 61:):
62:): while(substr_end>fragment_end) { // are there more fragments including pieces of 'substr'?
1(paf 63:3): fragment_begin=fragment_end;
64:3):
3(paf 65:3): const element_type fragment=i.next();
1(paf 66:3): fragment_end=fragment_begin+fragment.length;
(paf 67:):
68:): if(substr_end>fragment_end) // are there still more?
3(paf 69:3): append(Fragment(fragment.lang, fragment.length)); // appending whole fragment
(paf 70:): else { // no, it was last
3(paf 71:3): append(Fragment(fragment.lang, substr_end-fragment_begin));
72:3): return;
(paf 73:): }
74:): }
75:):
76:): break;
77:): }
78:): }
79:):
8(paf 80:3): // StringBody methods
5(paf 81:3):
9(paf 82:3): StringBody StringBody::Format(int value) {
8(paf 83:3): char local[MAX_NUMBER];
9(paf 84:3): size_t length=snprintf(local, MAX_NUMBER, "%d", value);
85:3): return StringBody(pa_strdup(local, length), length);
8(paf 86:3): }
5(paf 87:3):
3(paf 88:3): static int CORD_batched_iter_fn_generic_hash_code(char c, void * client_data) {
89:3): uint& result=*static_cast<uint*>(client_data);
90:3): generic_hash_code(result, c);
91:3): return 0;
92:3): }
5(paf 93:3): static int CORD_batched_iter_fn_generic_hash_code(const char* s, void * client_data) {
94:3): uint& result=*static_cast<uint*>(client_data);
95:3): generic_hash_code(result, s);
96:3): return 0;
97:3): };
7(paf 98:3): uint StringBody::hash_code() const {
5(paf 99:3): uint result=0;
3(paf 100:3): CORD_iter5(body, 0,
101:3): CORD_batched_iter_fn_generic_hash_code,
102:3): CORD_batched_iter_fn_generic_hash_code, &result);
5(paf 103:3): return result;
104:3): }
105:3):
(paf 106:): // String methods
1.172.2.2 paf 107:
1.172.2.21.2. 4(paf 108:3): String::String(const char* cstr, size_t help_length, bool tainted): body(CORD_EMPTY) {
109:3): append_help_length(cstr, help_length, tainted?L_TAINTED:L_CLEAN);
1.1 paf 110: }
1.140 paf 111:
1.172.2.21.2. 0(paf 112:3): String::String(const String& src): body(src.body) {
113:3): fragments.append(src.fragments);
114:3): }
1.28 paf 115:
1.172.2.21.2. 4(paf 116:3): String& String::append_help_length(const char* str, size_t helper_length, Language lang) {
3(paf 117:3): if(!str)
1.9 paf 118: return *this;
1.172.2.21.2. 4(paf 119:3): size_t use_length=helper_length?helper_length:strlen(str);
0(paf 120:3): if(!use_length)
1.9 paf 121: return *this;
1.122 paf 122:
1.172.2.21.2. 4(paf 123:3): body.append_help_length(str, use_length);
0(paf 124:3): fragments+=Fragment(lang, use_length);
1.1 paf 125:
126: return *this;
127: }
1.172.2.21.2. 4(paf 128:3): String& String::append_strdup(const char* str, size_t length, Language lang) {
129:3): body.append_strdup(str, length);
130:3): fragments+=Fragment(lang, length);
131:3):
132:3): return *this;
133:3): }
1.1 paf 134:
1.172.2.21.2. 1(paf 135:3): /// @todo check in doc: whether it documents NOW bad situation "abc".mid(-1, 3) =were?="ab"
8(paf 136:3): String& String::mid(size_t substr_begin, size_t substr_end) const {
(paf 137:): String& result=*new String;
1.33 paf 138:
1.172.2.21.2. 1(paf 139:3): size_t self_length=length();
140:3): substr_begin=max(min(substr_begin, self_length), (size_t)0);
141:3): substr_end=min(max(substr_end, substr_begin), self_length);
142:3): if(substr_begin==substr_end)
(paf 143:): return result;
1.52 paf 144:
1.172.2.21.2. (paf 145:): // first: letters themselves
7(paf 146:3): result.body=body.mid(substr_begin, substr_end-substr_begin);
1.46 paf 147:
1.172.2.21.2. (paf 148:): // next: their langs
3(paf 149:3): result.fragments.append_positions(fragments, substr_begin, substr_end);
1.53 paf 150:
1.172.2.21.2. 5(paf 151:3): // SAPI::log("piece of '%s' from %d to %d is '%s'",
1(paf 152:3): //cstr(), substr_begin, substr_end, result.cstr());
1.53 paf 153: return result;
1.54 paf 154: }
155:
1.172.2.21.2. 7(paf 156:3): size_t String::pos(const StringBody substr,
5(paf 157:3): size_t this_offset, Language lang) const {
(paf 158:): // first: letters themselves
2(paf 159:3): size_t substr_begin=body.pos(substr, this_offset);
1(paf 160:3): if(substr_begin==CORD_NOT_FOUND)
(paf 161:): return STRING_NOT_FOUND;
162:):
163:): // next: check the lang when specified
164:):
5(paf 165:3): if(lang==L_UNSPECIFIED) // ignore lang?
1(paf 166:3): return substr_begin;
(paf 167:):
168:): // substr must be in one fragment, and fragments' lang must = lang
7(paf 169:3): size_t substr_end=substr_begin+substr.length();
1(paf 170:3): size_t fragment_begin=0;
(paf 171:): size_t fragment_end;
1(paf 172:3): for(Array_iterator<ArrayFragment::element_type> i(fragments); i.has_next(); fragment_begin=fragment_end) {
3(paf 173:3): const Fragment fragment=i.next();
1(paf 174:3): fragment_end=fragment_begin+fragment.length;
(paf 175:):
1(paf 176:3): if(substr_begin<fragment_begin) // not reached fragments which may include 'result'?
(paf 177:): continue;
2(paf 178:3): if(substr_begin>=fragment_end) // begin of substr OUT of current fragment?
179:3): continue;
(paf 180:):
181:): if(substr_end>fragment_end) // end of substr OUT of current fragment?
2(paf 182:3): throw Exception(0, // (*) see below
183:3): this,
184:3): "searching for '%s' starting from %ud problem: found begin in one fragment, but end in another",
7(paf 185:3): substr.cstr(), this_offset);
(paf 186:):
4(paf 187:3): if(fragment.lang<=lang)
2(paf 188:3): return substr_begin;
189:3): else { // bad lang...
190:3): /// WARNING: this possibly skips assert (*), but it's fast
191:3): substr_begin=body.pos(substr, fragment_end/*...search AFTER for more*/);
192:3): if(substr_begin==CORD_NOT_FOUND)
193:3): return STRING_NOT_FOUND;
194:3):
7(paf 195:3): size_t substr_end=substr_begin+substr.length();
2(paf 196:3): // and continuing with next fragment
197:3): }
(paf 198:): }
199:):
200:): return STRING_NOT_FOUND;
1.58 paf 201: }
202:
1.172.2.21.2. (paf 203:): size_t String::pos(const String& substr,
5(paf 204:3): size_t this_offset, Language lang) const {
(paf 205:): return pos(substr.body, this_offset, lang);
1.60 paf 206: }
207:
1.172.2.14 paf 208: void String::split(ArrayString& result,
1.172.2.21.2. (paf 209:): size_t& pos_after,
210:): const char* delim,
2(paf 211:3): Language lang, int limit) const {
(paf 212:): size_t self_length=length();
213:): if(size_t delim_length=strlen(delim)) {
1.60 paf 214: int pos_before;
215: // while we have 'delim'...
1.172.2.21.2. (paf 216:): for(; (pos_before=pos(delim, pos_after, lang))>=0 && limit; limit--) {
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<self_length && limit) {
222:): result+=&mid(pos_after, self_length);
223:): pos_after=self_length;
1.60 paf 224: }
225: } else { // empty delim
1.172.2.21.2. (paf 226:): result+=this;
227:): pos_after+=self_length;
1.60 paf 228: }
229: }
230:
1.172.2.14 paf 231: void String::split(ArrayString& result,
1.172.2.21.2. (paf 232:): size_t& pos_after,
5(paf 233:3): const String& delim, Language lang,
(paf 234:): int limit) const {
5(paf 235:3): if(!delim.is_empty()) {
1.60 paf 236: int pos_before;
237: // while we have 'delim'...
238: for(; (pos_before=pos(delim, pos_after, lang))>=0 && limit; limit--) {
1.172.2.21.2. (paf 239:): result+=&mid(pos_after, pos_before);
240:): pos_after=pos_before+delim.length();
1.60 paf 241: }
242: // last piece
1.172.2.21.2. (paf 243:): if(pos_after<length() && limit) {
244:): result+=&mid(pos_after, length());
245:): pos_after=length();
1.60 paf 246: }
247: } else { // empty delim
1.172.2.21.2. (paf 248:): result+=this;
249:): pos_after+=length();
1.60 paf 250: }
1.61 paf 251: }
252:
1.172.2.21.2. 2(paf 253:3): static void regex_options(const String* options, int *result, bool& need_pre_post_match){
1.63 paf 254: struct Regex_option {
1.172.2.11 paf 255: const char* keyL;
256: const char* keyU;
1.63 paf 257: int clear, set;
258: int *result;
1.154 paf 259: bool *flag;
1.63 paf 260: } regex_option[]={
1.153 paf 261: {"i", "I", 0, PCRE_CASELESS, result}, // a=A
262: {"s", "S", 0, PCRE_DOTALL, result}, // \n\n$ [default]
263: {"x", "U", 0, PCRE_EXTENDED, result}, // whitespace in regex ignored
264: {"m", "M", PCRE_DOTALL, PCRE_MULTILINE, result}, // ^aaa\n$^bbb\n$
265: {"g", "G", 0, true, result+1}, // many rows
1.154 paf 266: {"'", 0, 0, 0, 0, &need_pre_post_match},
267: {0}
1.63 paf 268: };
1.171 paf 269: result[0]=PCRE_EXTRA | PCRE_DOTALL | PCRE_DOLLAR_ENDONLY;
1.63 paf 270: result[1]=0;
271:
1.172.2.21.2. 2(paf 272:3): if(options && !options->is_empty())
1.153 paf 273: for(Regex_option *o=regex_option; o->keyL; o++)
1.172.2.21.2. 2(paf 274:3): if(options->pos(o->keyL)>=0
275:3): || (o->keyU && options->pos(o->keyU)>=0)) {
1.154 paf 276: if(o->flag)
277: *o->flag=true;
278: else { // result
279: *o->result &= ~o->clear;
280: *o->result |= o->set;
281: }
1.63 paf 282: }
283: }
284:
1.172.2.21.2. (paf 285:): Table* String::match(Charset& source_charset,
1.172.2.4 paf 286: const String& regexp,
1.172.2.21.2. 2(paf 287:3): const String* options,
1.172.2.4 paf 288: Row_action row_action, void *info,
1.172.2.20 paf 289: bool& just_matched) const {
1.172.2.21.2. 5(paf 290:3): if(regexp.is_empty())
(paf 291:): throw Exception(0,
292:): 0,
1.73 paf 293: "regexp is empty");
1.154 paf 294:
1.172.2.21.2. (paf 295:): const char* pattern=regexp.cstr();
1.172.2.11 paf 296: const char* errptr;
1.62 paf 297: int erroffset;
1.172.2.21.2. (paf 298:): bool need_pre_post_match=false;
7(paf 299:3): int option_bits[2]={0}; regex_options(options, option_bits, need_pre_post_match);
1.172.2.20 paf 300: bool global=option_bits[1]!=0;
1.172.2.21.2. (paf 301:): pcre *code=pcre_compile(pattern, option_bits[0],
1.62 paf 302: &errptr, &erroffset,
1.172.2.1 paf 303: source_charset.pcre_tables);
1.62 paf 304:
1.67 paf 305: if(!code)
1.172.2.21.2. (paf 306:): throw Exception(0,
307:): ®exp.mid(erroffset, regexp.length()),
1.74 paf 308: "regular expression syntax error - %s", errptr);
1.62 paf 309:
1.172.2.20 paf 310: int subpatterns=pcre_info(code, 0, 0);
311: if(subpatterns<0) {
1.100 parser 312: pcre_free(code);
1.149 paf 313: throw Exception(0,
1.172.2.21.2. (paf 314:): ®exp,
1.76 paf 315: "pcre_info error (%d)",
1.172.2.20 paf 316: subpatterns);
1.63 paf 317: }
318:
1.172.2.21.2. (paf 319:): const char* subject=cstr();
320:): size_t subject_length=strlen(subject);
321:): const int oveclength=(1/*match*/+MAX_MATCH_GROUPS)*3;
322:): int ovector[oveclength];
1.155 paf 323:
324: // create table
1.172.2.21.2. 1(paf 325:3): Table& table=*new Table(string_match_table_template);
1.63 paf 326:
1.64 paf 327: int exec_option_bits=0;
1.154 paf 328: int prestart=0;
329: int poststart=0;
1.172.2.21.2. (paf 330:): int postfinish=length();
1.63 paf 331: while(true) {
332: int exec_substrings=pcre_exec(code, 0,
1.172.2.21.2. (paf 333:): subject, subject_length, prestart,
334:): exec_option_bits, ovector, oveclength);
1.63 paf 335:
336: if(exec_substrings==PCRE_ERROR_NOMATCH) {
1.100 parser 337: pcre_free(code);
1.172.2.21.2. (paf 338:): row_action(table, 0/*last time, no raw*/, 0, 0, poststart, postfinish, info);
1.172.2.20 paf 339: if(global || subpatterns)
1.172.2.21.2. 2(paf 340:3): return &table; // global or with subpatterns=true+result
1.172.2.20 paf 341: else {
1.172.2.21.2. (paf 342:): just_matched=false; return 0; // not global=no result
1.172.2.20 paf 343: }
1.63 paf 344: }
345:
346: if(exec_substrings<0) {
1.100 parser 347: pcre_free(code);
1.172.2.21.2. (paf 348:): throw Exception(0,
349:): ®exp,
1.76 paf 350: "regular expression execute error (%d)",
1.63 paf 351: exec_substrings);
352: }
353:
1.154 paf 354: int prefinish=ovector[0];
355: poststart=ovector[1];
1.172.2.21.2. (paf 356:): ArrayString* row=new ArrayString;
1.172.2.4 paf 357: if(need_pre_post_match) {
1.172.2.21.2. (paf 358:): *row+=&mid(0, prefinish); // .prematch column value
359:): *row+=&mid(prefinish, poststart); // .match
360:): *row+=&mid(poststart, postfinish); // .postmatch
1.172.2.4 paf 361: } else {
1.172.2.21.2. (paf 362:): *row+=0; // .prematch column value
363:): *row+=0; // .match
364:): *row+=0; // .postmatch
1.172.2.4 paf 365: }
1.63 paf 366:
367: for(int i=1; i<exec_substrings; i++) {
1.69 paf 368: // -1:-1 case handled peacefully by mid() itself
1.172.2.21.2. (paf 369:): *row+=&mid(ovector[i*2+0], ovector[i*2+1]); // .i column value
1.63 paf 370: }
371:
1.172.2.20 paf 372: row_action(table, row, prestart, prefinish, poststart, postfinish, info);
1.63 paf 373:
1.172.2.20 paf 374: if(!global || prestart==poststart) { // not global | going to hang
1.100 parser 375: pcre_free(code);
1.172.2.21.2. (paf 376:): row_action(table, 0/*last time, no row*/, 0, 0, poststart, postfinish, info);
2(paf 377:3): return &table;
1.63 paf 378: }
1.154 paf 379: prestart=poststart;
1.63 paf 380:
381: /*
382: if(option_bits[0] & PCRE_MULTILINE)
1.64 paf 383: exec_option_bits|=PCRE_NOTBOL; // start of subject+startoffset not BOL
1.63 paf 384: */
385: }
1.82 parser 386: }
387:
1.172.2.21.2. 0(paf 388:3): String& String::change_case(Charset& source_charset, Change_case_kind kind) const {
(paf 389:): String& result=*new String();
0(paf 390:3): if(is_empty())
391:3): return result;
1.172.2.4 paf 392:
1.172.2.1 paf 393: const unsigned char *tables=source_charset.pcre_tables;
1.82 parser 394:
395: const unsigned char *a;
396: const unsigned char *b;
397: switch(kind) {
398: case CC_UPPER:
399: a=tables+lcc_offset;
400: b=tables+fcc_offset;
401: break;
402: case CC_LOWER:
403: a=tables+lcc_offset;
404: b=0;
405: break;
406: default:
1.172.2.21.2. (paf 407:): throw Exception(0,
408:): this,
1.82 parser 409: "unknown change case kind #%d",
410: static_cast<int>(kind)); // never
411: a=b=0; // calm, compiler
412: break; // never
413: }
414:
1.172.2.21.2. (paf 415:): char* new_cstr=cstrm();
416:): char *dest=new_cstr;
417:): unsigned char index;
418:): for(const char* current=new_cstr; index=(unsigned char)*current; current++) {
419:): unsigned char c=a[index];
420:): if(b)
421:): c=b[c];
1.82 parser 422:
1.172.2.21.2. (paf 423:): *dest++=(char)c;
424:): }
425:): result.body=new_cstr;
3(paf 426:3): result.fragments.append(fragments);
1.89 parser 427:
1.101 parser 428: return result;
429: }
430:
1.172.2.21.2. (paf 431:): const String& String::replace(const Dictionary& dict) const {
432:): String& result=*new String();
433:): const char* old_cstr=cstr();
1(paf 434:3): const char* prematch_begin=old_cstr;
(paf 435:):
1(paf 436:3): const char* current=old_cstr;
437:3): while(*current) {
438:3): if(Table::element_type row=dict.first_that_begins(current)) {
(paf 439:): // prematch
1(paf 440:3): if(size_t prematch_length=current-prematch_begin) {
4(paf 441:3): result.body.append_strdup(prematch_begin, prematch_length);
442:3): result.fragments.append_positions(fragments, prematch_begin-old_cstr, current-old_cstr);
1.101 parser 443: }
444:
1.172.2.21.2. (paf 445:): // match
446:):
447:): const String* a=row->get(0);
1(paf 448:3): // skip 'a' in 'current'; move prematch_begin
449:3): current+=a->length(); prematch_begin=current;
1.170 paf 450:
1.172.2.21.2. (paf 451:): if(row->count()>1) { // are there any b?
452:): const String* b=row->get(1);
453:): if(size_t b_length=b->length()) {
1(paf 454:3): result.body<<b->cstr_to_string_body();
3(paf 455:3): result.fragments.append(b->fragments);
(paf 456:): }
457:): }
458:): } else // simply advance
459:): current++;
460:): }
1.156 paf 461:
1.172.2.21.2. (paf 462:): // postmatch
3(paf 463:3): if(size_t postmatch_length=current-prematch_begin) {
4(paf 464:3): result.body.append_strdup(prematch_begin, postmatch_length);
465:3): result.fragments.append_positions(fragments, prematch_begin-old_cstr, current-old_cstr);
(paf 466:): }
1.156 paf 467:
1.89 parser 468: return result;
469: }
470:
1.172.2.21.2. (paf 471:): double String::as_double() const {
1.89 parser 472: double result;
1.172.2.21.2. (paf 473:): const char *str=cstr();
474:):
475:): while(*str && isspace(*str))
476:): str++;
477:): if(!*str)
1.162 paf 478: return 0;
1.161 paf 479:
1.102 parser 480: char *error_pos;
1.89 parser 481: // 0xABC
1.172.2.21.2. (paf 482:): if(str[0]=='0')
483:): if(str[1]=='x' || str[1]=='X')
484:): result=(double)(unsigned long)strtol(str, &error_pos, 0);
1.99 parser 485: else
1.172.2.21.2. (paf 486:): result=(double)strtod(str+1/*skip leading 0*/, &error_pos);
1.89 parser 487: else
1.172.2.21.2. (paf 488:): result=(double)strtod(str, &error_pos);
1.89 parser 489:
1.159 paf 490: while(char c=*error_pos++)
491: if(!isspace(c))
492: throw Exception("number.format",
1.172.2.21.2. (paf 493:): this,
1.159 paf 494: "invalid number (double)");
1.89 parser 495:
496: return result;
497: }
1.172.2.21.2. (paf 498:): int String::as_int() const {
1.89 parser 499: int result;
1.172.2.21.2. (paf 500:): const char *str=cstr();
501:):
502:): while(*str && isspace(*str))
503:): str++;
504:): if(!*str)
1.162 paf 505: return 0;
1.161 paf 506:
1.102 parser 507: char *error_pos;
1.89 parser 508: // 0xABC
1.172.2.21.2. (paf 509:): if(str[0]=='0')
510:): if(str[1]=='x' || str[1]=='X')
511:): result=(int)(unsigned long)strtol(str, &error_pos, 0);
1.99 parser 512: else
1.172.2.21.2. (paf 513:): result=(int)strtol(str+1/*skip leading 0*/, &error_pos, 0);
1.89 parser 514: else
1.172.2.21.2. (paf 515:): result=(int)strtol(str, &error_pos, 0);
1.89 parser 516:
1.159 paf 517: while(char c=*error_pos++)
518: if(!isspace(c))
519: throw Exception("number.format",
1.172.2.21.2. (paf 520:): this,
1.159 paf 521: "invalid number (int)");
1.82 parser 522:
523: return result;
1.61 paf 524: }
1.113 parser 525:
1.172.2.4 paf 526: inline void uint2uchars(uint word, uchar *bytes) {
527: bytes[0]=word&0xFF;
528: bytes[1]=(word>>8)&0xFF;
529: bytes[2]=(word>>16)&0xFF;
530: bytes[3]=(word>>24)&0xFF;
531: }
532: inline uint uchars2uint(uchar *bytes) {
533: return bytes[3]<<24
534: | bytes[2]<<16
535: | bytes[1]<<8
536: | bytes[0];
537: }
538:
1.172.2.21.2. 7(paf 539:3): static int serialize_body_piece(const char* s, char** cur) {
(paf 540:): size_t length=strlen(s);
7(paf 541:3): memcpy(*cur, s, length); *cur+=length;
(paf 542:): return 0;
543:): };
544:): String::C String::serialize(size_t prolog_length) const {
545:): size_t buf_length=
546:): prolog_length
5(paf 547:3): +fragments.count()*(sizeof(Language)+sizeof(size_t))
(paf 548:): +length();
549:): C result(new(PointerFreeGC) char[buf_length], buf_length);
550:):
551:): // 1: prolog
552:): char *cur=result.str+prolog_length;
553:):
554:):
555:): // 2: fragments.count
556:): size_t fragments_count=fragments.count();
557:): memcpy(cur, &fragments_count, sizeof(fragments_count)); cur+=sizeof(fragments_count);
1.113 parser 558:
1.172.2.21.2. (paf 559:): // 3: lang info
560:): for(Array_iterator<ArrayFragment::element_type> i(fragments); i.has_next(); ) {
3(paf 561:3): const Fragment fragment=i.next();
1.123 paf 562: // lang
1.172.2.21.2. (paf 563:): memcpy(cur, &fragment.lang, sizeof(fragment.lang)); cur+=sizeof(fragment.lang);
564:): // length
565:): memcpy(cur, &fragment.length, sizeof(fragment.length)); cur+=sizeof(fragment.length);
566:): }
567:):
568:): // 4: letters
7(paf 569:3): body.for_each(serialize_body_piece, &cur);
(paf 570:):
571:): return result;
1.113 parser 572: }
1.172.2.21.2. 9(paf 573:3): bool String::deserialize(size_t prolog_length, void *buf, size_t buf_length) {
(paf 574:): if(buf_length<=prolog_length)
1.148 paf 575: return false;
1.172.2.21.2. (paf 576:): buf_length-=prolog_length;
1.113 parser 577:
1.172.2.21.2. (paf 578:): // 1: prolog
579:): const char* cur=(const char* )buf+prolog_length;
1.148 paf 580:
1.172.2.21.2. (paf 581:): // 2: fragments.count
582:): if(buf_length<sizeof(size_t)) // fragments.count don't fit?
583:): return false;
584:): size_t fragments_count=*reinterpret_cast<const size_t*>(cur); cur+=sizeof(size_t);
585:): buf_length-=sizeof(size_t);
1.128 paf 586:
1.172.2.21.2. (paf 587:): // 3: lang info
588:): size_t total_length=0;
589:): for(size_t f=0; f<fragments_count; f++) {
5(paf 590:3): size_t piece_length=sizeof(Language)+sizeof(size_t);
(paf 591:): if(buf_length<piece_length) // lang+length
1.148 paf 592: return false;
593:
1.172.2.21.2. 5(paf 594:3): Language lang=*reinterpret_cast<const Language *>(cur); cur+=sizeof(Language);
(paf 595:): size_t fragment_length=*reinterpret_cast<const size_t*>(cur); cur+=sizeof(size_t);
596:): fragments+=Fragment(lang, fragment_length);
597:): total_length+=fragment_length;
1.113 parser 598:
1.172.2.21.2. (paf 599:): buf_length-=piece_length;
1.113 parser 600: }
1.172.2.21.2. (paf 601:):
602:): // 4: letters
603:): if(buf_length!=total_length)
604:): return false;
605:):
606:): body=CORD_cat_char_star(CORD_EMPTY, cur, buf_length);
607:):
1.148 paf 608: return true;
1.113 parser 609: }
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