Annotation of parser3/src/include/pa_string.h, revision 1.208
1.41 paf 1: /** @file
1.43 paf 2: Parser: string class decl.
3:
1.201 moko 4: Copyright (c) 2001-2012 Art. Lebedev Studio (http://www.artlebedev.com)
1.124 paf 5: Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)
1.1 paf 6: */
7:
8: #ifndef PA_STRING_H
9: #define PA_STRING_H
1.140 paf 10:
1.208 ! moko 11: #define IDENT_PA_STRING_H "$Id: pa_string.h,v 1.207 2015/04/08 17:36:19 moko Exp $"
1.145 paf 12:
13: // includes
1.4 paf 14: #include "pa_types.h"
1.145 paf 15: #include "pa_array.h"
16:
17: extern "C" { // cord's author forgot to do that
18: #define CORD_NO_IO
19: #include "cord.h"
1.186 misha 20:
21: #ifdef CORD_CAT_OPTIMIZATION
22: #define CORD_cat(x, y) CORD_cat_optimized(x, y)
23: #define CORD_cat_char_star(x, y, leny) CORD_cat_char_star_optimized(x, y, leny)
24: #endif
1.145 paf 25: };
1.4 paf 26:
1.164 paf 27: // defines
28:
1.183 misha 29: // cache hash code in String::Body for faster hash access
30: #define HASH_CODE_CACHING
31:
1.187 misha 32: // cache String::Body.length() for char* strings only, CORDs have own length
33: #define STRING_LENGTH_CACHING
34:
35:
1.146 paf 36: // cord extension
37: /* Returns true if x does contain */
38: /* char not_c at positions i..i+n. Value i,i+n must be < CORD_len(x). */
39: int CORD_range_contains_chr_greater_then(CORD x, size_t i, size_t n, int c);
1.148 paf 40: size_t CORD_block_count(CORD x);
1.146 paf 41:
1.145 paf 42: // forwards
1.9 paf 43:
1.145 paf 44: class Charset;
1.135 paf 45: class Table;
1.71 paf 46: class SQL_Connection;
1.101 parser 47: class Dictionary;
1.145 paf 48: class Request_charsets;
49: class String;
50: typedef Array<const String*> ArrayString;
1.178 misha 51: class VRegex;
1.145 paf 52:
1.155 paf 53: // generally useful
54:
55: int pa_atoi(const char* str, const String* problem_source=0);
56: double pa_atod(const char* str, const String* problem_source=0);
1.204 moko 57: unsigned int pa_atoui(const char *str, int base, const String* problem_source=0);
1.155 paf 58:
1.145 paf 59: /// this is result of pos functions which mean that substr were not found
60: #define STRING_NOT_FOUND ((size_t)-1)
61:
1.42 paf 62: /**
1.146 paf 63: String which knows the lang of all it's langs.
1.41 paf 64:
65: All pieces remember
66: - whether they are tainted or not,
1.146 paf 67: and the lang which should be used to detaint them
1.41 paf 68: */
1.176 misha 69:
70:
1.145 paf 71: class String: public PA_Object {
1.1 paf 72: public:
1.48 paf 73:
1.146 paf 74: /** piece is tainted or not. the lang to use when detaint
1.106 parser 75: remember to change String_Untaint_lang_name @ untaint.C along
1.148 paf 76:
77: WARNING WARNING WARNING WARNING WARNING WARNING
78:
79: pos function compares(<=) languages, that is used in searching
80: for table column separator being L_CLEAN or L_AS_IS.
81: they search for AS_IS, meaning AS_IS|CLEAN [doing <=L_AS_IS check].
82:
83: letters assigned for debugging, but it's important for no language-letter
84: come before L_AS_IS other then L_CLEAN
85:
86: WARNING WARNING WARNING WARNING WARNING WARNING
1.106 parser 87: */
1.145 paf 88: enum Language {
1.146 paf 89: L_UNSPECIFIED=0, ///< no real string has parts of this lange: it's just convinient to check when string's empty
1.145 paf 90: // these two must go before others, there are checks for >L_AS_IS
1.170 misha 91: L_CLEAN='0', ///< clean WARNING: read above warning before changing
92: L_AS_IS='A', ///< leave all characters intact WARNING: read above warning before changing
1.122 paf 93:
1.148 paf 94: L_PASS_APPENDED='P',
1.41 paf 95: /**<
1.146 paf 96: leave lang built into string being appended.
1.41 paf 97: just a flag, that value not stored
98: */
1.170 misha 99: L_TAINTED='T', ///< tainted, untaint lang as assigned later
1.146 paf 100: // untaint langs. assigned by ^untaint[lang]{...}
1.170 misha 101: L_FILE_SPEC='F', ///< file specification
102: L_HTTP_HEADER='h', ///< text in HTTP response header
103: L_MAIL_HEADER='m', ///< text in mail header
1.192 misha 104: L_URI='U', ///< text in uri
105: L_SQL='Q', ///< ^table:sql body
106: L_JS='J', ///< JavaScript code
1.198 misha 107: L_XML='X', ///< ^xdoc:create xml
1.192 misha 108: L_HTML='H', ///< HTML code
1.198 misha 109: L_REGEX='R', ///< RegExp
110: L_JSON='S', ///< JSON code
1.170 misha 111: L_HTTP_COOKIE='C', ///< cookies encoded as %uXXXX for compartibility with js functions encode/decode
1.192 misha 112: L_PARSER_CODE='p', ///< ^process body
1.148 paf 113: // READ WARNING ABOVE BEFORE ADDING ANYTHING
1.146 paf 114: L_OPTIMIZE_BIT = 0x80 ///< flag, requiring cstr whitespace optimization
1.27 paf 115: };
116:
1.157 paf 117: enum Trim_kind {
118: TRIM_BOTH,
119: TRIM_START,
120: TRIM_END
121: };
122:
1.187 misha 123: class Body;
1.176 misha 124:
1.147 paf 125: union Languages {
1.146 paf 126:
1.147 paf 127: struct {
1.160 paf 128: #ifdef PA_LITTLE_ENDIAN
1.147 paf 129: Language lang:8;
1.179 misha 130: size_t is_not_just_lang:sizeof(CORD)*8-8;
1.160 paf 131: #elif defined(PA_BIG_ENDIAN)
1.179 misha 132: size_t is_not_just_lang:sizeof(CORD)*8-8;
1.160 paf 133: Language lang:8;
134: #else
135: # error word endianness not determined for some obscure reason
136: #endif
1.147 paf 137: } opt;
138: CORD langs;
1.146 paf 139:
1.187 misha 140: CORD make_langs(const Body& current) const {
141: return opt.is_not_just_lang?langs:CORD_chars((char)opt.lang, current.length());
1.146 paf 142: }
143:
144: CORD make_langs(size_t aoffset, size_t alength) const {
1.147 paf 145: return opt.is_not_just_lang?
1.194 misha 146: CORD_substr(langs, aoffset, alength, 0)
1.162 paf 147: :CORD_chars((char)opt.lang, alength);
1.145 paf 148: }
149:
1.157 paf 150: /// appending when 'langs' already contain something [simple cases handled elsewhere]
1.187 misha 151: void append(size_t current, const CORD to_nonempty_target_langs) {
1.146 paf 152: assert(langs);
1.187 misha 153: if(opt.is_not_just_lang)
154: langs=CORD_cat(langs, to_nonempty_target_langs);
155: else {
156: assert(current);
157: langs=CORD_cat(CORD_chars((char)opt.lang, current), to_nonempty_target_langs);
158: }
159: }
1.146 paf 160:
1.187 misha 161: void append(const Body& current, const CORD to_nonempty_target_langs) {
162: assert(langs);
1.147 paf 163: if(opt.is_not_just_lang)
1.146 paf 164: langs=CORD_cat(langs, to_nonempty_target_langs);
1.187 misha 165: else {
166: size_t current_size=current.length();
1.146 paf 167: assert(current_size);
1.187 misha 168: langs=CORD_cat(CORD_chars((char)opt.lang, current_size), to_nonempty_target_langs);
1.146 paf 169: }
1.145 paf 170: }
1.146 paf 171:
1.145 paf 172: public:
1.146 paf 173:
174: const char* v() const;
1.164 paf 175: void dump() const;
1.146 paf 176:
177: Languages(): langs(0) {}
1.147 paf 178: Languages(Language alang) {
179: opt.lang=alang;
180: opt.is_not_just_lang=0;
181: }
1.146 paf 182:
183: /// MUST be called exactly prior to modification of current [uses it's length]
1.187 misha 184: void append(size_t current, Language alang, size_t length) {
1.146 paf 185: assert(alang);
1.187 misha 186: assert(length);
1.146 paf 187:
1.208 ! moko 188: if(!opt.is_not_just_lang) {
1.147 paf 189: if(opt.lang) {
1.181 misha 190: if(opt.lang==alang) // same language? ignoring
1.146 paf 191: return;
192: } else {
1.147 paf 193: opt.lang=alang; // to uninitialized
1.146 paf 194: return;
1.145 paf 195: }
1.208 ! moko 196: }
1.146 paf 197:
1.187 misha 198: append(current, CORD_chars((char)alang, length));
1.146 paf 199: }
200:
1.187 misha 201: void append(const Body ¤t, Language alang, size_t length) {
1.176 misha 202: assert(alang);
1.187 misha 203: assert(length);
1.176 misha 204:
1.208 ! moko 205: if(!opt.is_not_just_lang) {
1.176 misha 206: if(opt.lang) {
1.183 misha 207: if(opt.lang==alang) // same language? ignoring
1.176 misha 208: return;
209: } else {
210: opt.lang=alang; // to uninitialized
211: return;
212: }
1.208 ! moko 213: }
1.176 misha 214:
1.187 misha 215: append(current, CORD_chars((char)alang, length));
1.176 misha 216: }
217:
1.187 misha 218: void appendHelper(const Body& current, Language alang, const Body &length_helper) {
219: assert(alang);
220:
1.208 ! moko 221: if(!opt.is_not_just_lang) {
1.187 misha 222: if(opt.lang) {
223: if(opt.lang==alang) // same language? ignoring
224: return;
225: } else {
226: opt.lang=alang; // to uninitialized
227: return;
228: }
1.208 ! moko 229: }
1.187 misha 230:
231: append(current, CORD_chars((char)alang, length_helper.length()));
232: }
233:
234: void appendHelper(const Body& current, const Languages &src, const Body& length_helper) {
1.186 misha 235: if(!langs){
1.176 misha 236: langs=src.langs; // to uninitialized
1.186 misha 237: #ifdef CORD_CAT_OPTIMIZATION
238: if(opt.is_not_just_lang && !CORD_IS_STRING(langs))
239: CORD_concatenation_protect(langs);
240: #endif
241: }
1.176 misha 242: else if(!src.opt.is_not_just_lang)
243: appendHelper(current, src.opt.lang, length_helper); // simplifying when simple source
244: else
245: append(current, src.make_langs(length_helper));
246: }
247:
1.146 paf 248: /// MUST be called exactly prior to modification of current [uses it's length]
1.187 misha 249: void append(const Body& current, const Languages src, size_t aoffset, size_t alength) {
1.146 paf 250: assert(alength);
251:
252: if(!langs) // to uninitialized?
1.147 paf 253: if(src.opt.is_not_just_lang)
1.194 misha 254: langs=CORD_substr(src.langs, aoffset, alength, 0); // to uninitialized complex
1.146 paf 255: else
1.147 paf 256: opt.lang=src.opt.lang; // to uninitialized simple
1.146 paf 257: else
1.147 paf 258: if(!opt.is_not_just_lang && !src.opt.is_not_just_lang && opt.lang==src.opt.lang) // both simple & of same language?
1.146 paf 259: return; // ignoring
260: else
261: append(current, src.make_langs(aoffset, alength));
262: }
263:
264: /// checks if we have lang<=alang all from aoffset to aoffset+alength
265: bool check_lang(Language alang, size_t aoffset, size_t alength) const {
266: if(alang==L_UNSPECIFIED) // ignore lang?
267: return true;
268:
1.147 paf 269: if(opt.is_not_just_lang)
1.146 paf 270: return CORD_range_contains_chr_greater_then(langs, aoffset, alength, (unsigned)alang)==0;
271: else
1.169 misha 272: return (unsigned)opt.lang<=(unsigned)alang;
1.146 paf 273: }
274:
1.148 paf 275: /// @returns count of blocks
276: /// @todo currently there can be adjucent blocks of same language. someday merge them
277: size_t count() const {
278: return opt.is_not_just_lang?
279: CORD_block_count(langs)
280: : opt.lang?
281: 1
282: : 0;
283: };
284:
1.187 misha 285: template<typename I>
286: void for_each(size_t current, int callback(char, size_t, I), I info) const {
287: if(opt.is_not_just_lang)
288: CORD_block_iter(langs, 0, (CORD_block_iter_fn)callback, info);
289: else
290: callback(opt.lang, current, info);
291: }
292:
293: template<typename I>
294: void for_each(const Body& current, int callback(char, size_t, I), I info) const {
1.147 paf 295: if(opt.is_not_just_lang)
1.146 paf 296: CORD_block_iter(langs, 0, (CORD_block_iter_fn)callback, info);
297: else
1.187 misha 298: callback(opt.lang, current.length(), info);
1.146 paf 299: }
300:
1.164 paf 301: bool invariant(size_t current_length) const {
1.146 paf 302: if(!langs)
303: return current_length==0;
1.147 paf 304: if(opt.is_not_just_lang)
1.146 paf 305: return CORD_len(langs)==current_length;
306: return true; // uncheckable, actually
307: }
308: };
309:
310: class Body {
311:
312: CORD body;
313:
1.183 misha 314: #ifdef HASH_CODE_CACHING
315: // cached hash code is not reseted on write operations as test shows
316: // that string body does not change after it is stored as a hash key
317: mutable uint hash_code;
318: #endif
319:
1.187 misha 320: #ifdef STRING_LENGTH_CACHING
321: // cached length is reseted on modification, used only for char*, not CORD
322: mutable size_t string_length;
323: #define INIT_LENGTH ,string_length(0)
324: #define ZERO_LENGTH string_length=0;
325: #else
326: #define INIT_LENGTH
327: #define ZERO_LENGTH
328: #endif
329:
1.146 paf 330: public:
331:
332: const char* v() const;
1.164 paf 333: void dump() const;
1.146 paf 334:
1.183 misha 335: #ifdef HASH_CODE_CACHING
1.187 misha 336: Body(): body(CORD_EMPTY), hash_code(0) INIT_LENGTH {}
337: Body(CORD abody, uint ahash_code): body(abody), hash_code(ahash_code) INIT_LENGTH {}
338: Body(CORD abody): body(abody), hash_code(0) INIT_LENGTH {
1.183 misha 339: #else
1.187 misha 340: Body(): body(CORD_EMPTY) INIT_LENGTH {}
341: Body(CORD abody): body(abody) INIT_LENGTH {
1.183 misha 342: #endif
1.186 misha 343: #ifdef CORD_CAT_OPTIMIZATION
344: assert(!body // no body
345: || *body // ordinary string
346: || body[1]==1 // CONCAT_HDR
347: || body[1]==3 // CONCAT_HDR_READ_ONLY
348: || body[1]==4 // FN_HDR
349: || body[1]==6 // SUBSTR_HDR
350: );
351: #else
1.146 paf 352: assert(!body // no body
353: || *body // ordinary string
354: || body[1]==1 // CONCAT_HDR
355: || body[1]==4 // FN_HDR
356: || body[1]==6 // SUBSTR_HDR
357: );
1.186 misha 358: #endif
1.146 paf 359: }
1.181 misha 360:
1.146 paf 361: static Body Format(int value);
362:
1.187 misha 363: void clear() { ZERO_LENGTH body=CORD_EMPTY; }
1.146 paf 364:
365: bool operator! () const { return is_empty(); }
366:
1.183 misha 367: CORD get_cord() const { return body; }
368: uint get_hash_code() const;
1.146 paf 369:
1.197 misha 370: const char* cstr() const {
371: #ifdef STRING_LENGTH_CACHING
372: string_length = length();
373: if(string_length)
374: return const_cast<Body*>(this)->body=CORD_to_const_char_star(body, string_length);
375: #endif
376: return CORD_to_const_char_star(body, length());
377: }
378:
1.194 misha 379: char* cstrm() const { return CORD_to_char_star(body, length()); }
1.146 paf 380:
1.187 misha 381: #ifdef STRING_LENGTH_CACHING
382: void set_length(size_t alength){ string_length = alength; }
383: size_t length() const { return body ? CORD_IS_STRING(body) ? string_length ? string_length : (string_length=strlen(body)) : CORD_len(body) : 0; }
384: #else
1.146 paf 385: size_t length() const { return CORD_len(body); }
1.187 misha 386: #endif
1.146 paf 387:
388: bool is_empty() const { return body==CORD_EMPTY; }
389:
390: void append_know_length(const char *str, size_t known_length) {
1.187 misha 391: if(known_length){
392: if(body){
393: body = CORD_cat_char_star(body, str, known_length);
394: ZERO_LENGTH
395: } else {
396: body=str;
397: #ifdef STRING_LENGTH_CACHING
398: string_length=known_length;
399: #endif
400: }
401: }
1.146 paf 402: }
403: void append_strdup_know_length(const char* str, size_t known_length) {
404: if(known_length)
405: append_know_length(pa_strdup(str, known_length), known_length);
406: }
1.187 misha 407: void append(char c) { ZERO_LENGTH body=CORD_cat_char(body, c); }
408: Body& operator << (const Body src) { ZERO_LENGTH body=CORD_cat(body, src.body); return *this; }
409:
1.146 paf 410: Body& operator << (const char* str) { append_know_length(str, strlen(str)); return *this; }
411:
412: bool operator < (const Body src) const { return CORD_cmp(body, src.body)<0; }
413: bool operator > (const Body src) const { return CORD_cmp(body, src.body)>0; }
414: bool operator <= (const Body src) const { return CORD_cmp(body, src.body)<=0; }
415: bool operator >= (const Body src) const { return CORD_cmp(body, src.body)>=0; }
1.207 moko 416:
1.146 paf 417: bool operator != (const Body src) const { return CORD_cmp(body, src.body)!=0; }
418: bool operator == (const Body src) const { return CORD_cmp(body, src.body)==0; }
419:
1.207 moko 420: bool operator != (const char *src) const { return CORD_cmp(body, src)!=0; }
421: bool operator == (const char *src) const { return CORD_cmp(body, src)==0; }
422:
1.146 paf 423: int ncmp(size_t x_begin, const Body y, size_t y_begin, size_t size) const {
424: return CORD_ncmp(body, x_begin, y.body, y_begin, size);
1.145 paf 425: }
426:
1.146 paf 427: char fetch(size_t index) const { return CORD_fetch(body, index); }
1.194 misha 428: Body mid(size_t aindex, size_t alength) const { return CORD_substr(body, aindex, alength, length()); }
429: size_t pos(const char* substr, size_t offset=0) const { return CORD_str(body, offset, substr, length()); }
1.146 paf 430: size_t pos(const Body substr, size_t offset=0) const {
1.180 misha 431: if(substr.is_empty())
1.146 paf 432: return STRING_NOT_FOUND; // in this case CORD_str returns 0 [parser users got used to -1]
1.150 paf 433:
434: // CORD_str checks for bad offset [CORD_chr does not]
1.194 misha 435: return CORD_str(body, offset, substr.body, length());
1.146 paf 436: }
1.207 moko 437: size_t pos(char c, size_t offset=0) const {
1.150 paf 438: if(offset>=length()) // CORD_chr does not check that [and ABORT's in that case]
439: return STRING_NOT_FOUND;
440:
1.146 paf 441: return CORD_chr(body, offset, c);
1.145 paf 442: }
1.146 paf 443:
1.203 misha 444: size_t strrpbrk(const char* chars, size_t left, size_t right) const;
445:
446: size_t rskipchars(const char* chars, size_t left, size_t right) const;
447:
1.187 misha 448: template<typename I>
449: int for_each(int (*f)(char c, I), I info) const {
1.159 paf 450: return CORD_iter(body, (CORD_iter_fn)f, (void*)info);
451: }
452:
1.187 misha 453: template<typename I>
454: int for_each(int (*f1)(char c, I), int (*f2)(const char* s, I), I info) const {
1.159 paf 455: return CORD_iter5(body, 0, (CORD_iter_fn)f1, (CORD_batched_iter_fn)f2, info);
1.148 paf 456: }
1.146 paf 457:
458: void set_pos(CORD_pos& pos, size_t index) const { CORD_set_pos(pos, body, index); }
459:
1.157 paf 460: /// @returns this or 0 or mid. if returns this or 0 out_* are not filled
461: Body trim(Trim_kind kind=TRIM_BOTH, const char* chars=0,
1.188 misha 462: size_t* out_start=0, size_t* out_length=0, Charset* source_charset=0) const;
1.145 paf 463: };
464:
465: struct C {
466: const char *str;
467: size_t length;
468: operator const char *() { return str; }
1.158 paf 469: C(): str(0), length(0) {}
1.145 paf 470: C(const char *astr, size_t asize): str(astr), length(asize) {}
471: };
472:
473: struct Cm {
474: char *str;
475: size_t length;
476: //operator char *() { return str; }
1.158 paf 477: Cm(): str(0), length(0) {}
1.145 paf 478: Cm(char *astr, size_t asize): str(astr), length(asize) {}
479: };
480:
481: private:
482:
1.152 paf 483: Body body; ///< all characters of string
1.146 paf 484: Languages langs; ///< string characters lang
485:
486: const char* v() const;
1.164 paf 487: void dump() const;
488: #define ASSERT_STRING_INVARIANT(string) \
489: assert((string).langs.invariant((string).body.length()))
1.145 paf 490:
1.8 paf 491: public:
1.151 paf 492:
493: static const String Empty;
1.8 paf 494:
1.181 misha 495: explicit String(){};
1.182 misha 496: explicit String(const char* cstr, Language alang=L_CLEAN){
1.181 misha 497: if(cstr && *cstr){
498: body=cstr;
1.182 misha 499: langs=alang;
1.181 misha 500: }
501: }
1.187 misha 502: explicit String(const char* cstr, Language alang, size_t alength){
503: if(cstr && *cstr){
504: body=cstr;
505: #ifdef STRING_LENGTH_CACHING
506: body.set_length(alength);
507: #endif
1.182 misha 508: langs=alang;
1.181 misha 509: }
510: }
1.187 misha 511:
1.206 moko 512: String(int value, const char *format);
1.146 paf 513: String(Body abody, Language alang): body(abody), langs(alang) {
514: ASSERT_STRING_INVARIANT(*this);
515: }
516: String(const String& src): body(src.body), langs(src.langs) {
517: ASSERT_STRING_INVARIANT(*this);
1.145 paf 518: }
519:
520: /// for convinient hash lookup
1.183 misha 521: #ifdef HASH_CODE_CACHING
522: operator const Body&() const { return body; }
523: #else
1.146 paf 524: operator const Body() const { return body; }
1.183 misha 525: #endif
1.145 paf 526:
527: bool is_empty() const { return body.is_empty(); }
528: size_t length() const { return body.length(); }
1.171 misha 529: size_t length(Charset& charset) const;
1.145 paf 530:
1.191 misha 531: /// convert to CORD forcing lang tainting
532: Body cstr_to_string_body_taint(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const;
533: /// convert to CORD with tainting dirty to lang
534: Body cstr_to_string_body_untaint(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const;
1.145 paf 535:
1.189 misha 536: ///
537: const char* cstr() const {
538: return body.cstr();
539: }
540: ///
541: char* cstrm() const {
542: return body.cstrm();
543: }
544:
545: /// convert to constant C string forcing lang tainting
1.190 misha 546: const char* taint_cstr(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 547: return cstr_to_string_body_taint(lang, connection, charsets).cstr();
548: }
1.190 misha 549: char *taint_cstrm(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 550: return cstr_to_string_body_taint(lang, connection, charsets).cstrm();
551: }
552:
553: /// convert to constant C string with tainting dirty to lang
1.190 misha 554: const char* untaint_cstr(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 555: return cstr_to_string_body_untaint(lang, connection, charsets).cstr();
556: }
1.190 misha 557: char *untaint_cstrm(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 558: return cstr_to_string_body_untaint(lang, connection, charsets).cstrm();
559: }
1.195 misha 560:
1.196 misha 561: const char* untaint_and_transcode_cstr(Language lang, const Request_charsets *charsets) const;
1.195 misha 562:
1.202 moko 563: bool is_not_just_lang() const {
564: return langs.opt.is_not_just_lang !=0;
565: }
566:
567: Language just_lang() const {
568: return langs.opt.lang;
569: }
570:
1.108 parser 571: /// puts pieces to buf
1.145 paf 572: Cm serialize(size_t prolog_size) const;
1.108 parser 573: /// appends pieces from buf to self
1.145 paf 574: bool deserialize(size_t prolog_size, void *buf, size_t buf_size);
1.146 paf 575: /// @see Body::append_know_length
1.145 paf 576: String& append_know_length(const char* str, size_t known_length, Language lang);
1.146 paf 577: /// @see Body::append_help_length
1.145 paf 578: String& append_help_length(const char* str, size_t helper_length, Language lang);
579: String& append_strdup(const char* str, size_t helper_length, Language lang);
580:
1.146 paf 581: bool operator == (const char* y) const { return body==Body(y); }
582: bool operator != (const char* y) const { return body!=Body(y); }
1.145 paf 583:
584: /// this starts with y
585: bool starts_with(const char* y) const {
1.146 paf 586: return body.ncmp(0/*x_begin*/, Body(y), 0/*y_begin*/, strlen(y))==0;
1.145 paf 587: }
588: /// x starts with this
589: bool this_starts(const char* x) const {
1.146 paf 590: return Body(x).ncmp(0/*x_begin*/, body, 0/*y_begin*/, length())==0;
1.26 paf 591: }
592:
1.187 misha 593: String& append_to(String& dest, Language lang, bool forced=false) const;
1.145 paf 594: String& append(const String& src, Language lang, bool forced=false) {
595: return src.append_to(*this, lang, forced);
596: }
1.199 misha 597: String& append_quoted(const String* src, Language lang=L_JSON){
598: *this << "\"";
599: if(src)
600: this->append(*src, lang, true/*forced lang*/);
601: *this << "\"";
602: return *this;
603: }
604:
1.145 paf 605: String& operator << (const String& src) { return append(src, L_PASS_APPENDED); }
606: String& operator << (const char* src) { return append_help_length(src, 0, L_AS_IS); }
1.187 misha 607: String& operator << (const Body& src){
608: langs.appendHelper(body, L_AS_IS, src);
609: body<<src;
610: return *this;
611: }
1.100 parser 612:
1.142 paf 613: char first_char() const {
1.145 paf 614: return is_empty()?0:body.fetch(0);
1.142 paf 615: }
1.54 paf 616:
1.205 moko 617: char last_char() const {
618: return is_empty()?0:body.fetch(body.length()-1);
619: }
620:
1.145 paf 621: bool operator < (const String& src) const { return body<src.body; }
622: bool operator > (const String& src) const { return body>src.body; }
623: bool operator <= (const String& src) const { return body<=src.body; }
624: bool operator >= (const String& src) const { return body>=src.body; }
625: bool operator != (const String& src) const { return body!=src.body; }
626: bool operator == (const String& src) const { return body==src.body; }
627:
1.54 paf 628: /// extracts [start, finish) piece of string
1.145 paf 629: String& mid(size_t substr_begin, size_t substr_end) const;
1.172 misha 630: String& mid(Charset& charset, size_t from, size_t to, size_t helper_length=0) const;
1.145 paf 631:
632: /**
633: ignore lang if it's L_UNSPECIFIED
634: but when specified: look for substring that lies in ONE fragment in THAT lang
635: @return position of substr in string, -1 means "not found" [const char* version]
636: */
1.146 paf 637: size_t pos(const Body substr,
1.145 paf 638: size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
639: /// String version of @see pos(const char*, int, Language)
640: size_t pos(const String& substr,
641: size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
642: size_t pos(char c,
643: size_t this_offset=0) const {
644: return body.pos(c, this_offset);
645: }
1.173 misha 646: size_t pos(Charset& charset,
1.171 misha 647: const String& substr,
648: size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
1.55 paf 649:
1.203 misha 650: size_t strrpbrk(const char* chars, size_t left=0) const {
651: return (length()) ? body.strrpbrk(chars, left, length()-1) : STRING_NOT_FOUND;
652: }
653: size_t strrpbrk(const char* chars, size_t left, size_t right) const {
654: return body.strrpbrk(chars, left, right);
655: }
656:
657: size_t rskipchars(const char* chars, size_t left=0) const {
658: return (length()) ? body.rskipchars(chars, left, length()-1) : STRING_NOT_FOUND;
659: }
660: size_t rskipchars(const char* chars, size_t left, size_t right) const {
661: return body.rskipchars(chars, left, right);
662: }
663:
1.145 paf 664: void split(ArrayString& result,
665: size_t& pos_after,
666: const char* delim,
667: Language lang=L_UNSPECIFIED, int limit=-1) const;
668: void split(ArrayString& result,
669: size_t& pos_after,
1.62 paf 670: const String& delim,
1.145 paf 671: Language lang=L_UNSPECIFIED, int limit=-1) const;
1.62 paf 672:
1.145 paf 673: typedef void (*Row_action)(Table& table, ArrayString* row,
1.136 paf 674: int prestart, int prefinish,
675: int poststart, int postfinish,
1.68 paf 676: void *info);
1.87 parser 677: /**
1.145 paf 678: @return table of found items, if any.
1.87 parser 679: table format is defined and fixed[can be used by others]:
680: @verbatim
681: prematch/match/postmatch/1/2/3/...
682: @endverbatim
683: */
1.178 misha 684: Table* match(VRegex* vregex,
1.99 parser 685: Row_action row_action, void *info,
1.168 misha 686: int& matches_count) const;
1.87 parser 687: enum Change_case_kind {
688: CC_UPPER,
689: CC_LOWER
690: };
1.145 paf 691: String& change_case(Charset& source_charset,
1.87 parser 692: Change_case_kind kind) const;
1.145 paf 693: const String& replace(const Dictionary& dict) const;
1.188 misha 694: const String& trim(Trim_kind kind=TRIM_BOTH, const char* chars=0, Charset* source_charset=0) const;
1.155 paf 695: double as_double() const { return pa_atod(cstr(), this); }
696: int as_int() const { return pa_atoi(cstr(), this); }
1.167 misha 697: bool as_bool() const { return as_int()!=0; }
1.174 misha 698: const String& escape(Charset& source_charset) const;
1.137 paf 699:
1.7 paf 700: private: //disabled
701:
1.12 paf 702: String& operator = (const String&) { return *this; }
1.7 paf 703:
1.1 paf 704: };
1.119 paf 705:
1.183 misha 706: #ifndef HASH_CODE_CACHING
1.145 paf 707: /// simple hash code of string. used by Hash
1.146 paf 708: inline uint hash_code(const String::Body self) {
1.184 misha 709: return self.get_hash_code();
1.119 paf 710: }
1.183 misha 711: #endif
1.147 paf 712:
713:
1.1 paf 714: #endif
E-mail:
![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to pa_string.h CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [parser3project] / parser3 / src / include