Annotation of parser3/src/include/pa_string.h, revision 1.210
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:
1.210 ! moko 11: #define IDENT_PA_STRING_H "$Id: pa_string.h,v 1.209 2015/05/30 22:55:28 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
1.209 moko 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);
1.210 ! moko 58: unsigned long long int pa_atoul(const char *str, int base, const String* problem_source=0);
1.155 paf 59:
1.145 paf 60: /// this is result of pos functions which mean that substr were not found
61: #define STRING_NOT_FOUND ((size_t)-1)
62:
1.42 paf 63: /**
1.146 paf 64: String which knows the lang of all it's langs.
1.41 paf 65:
66: All pieces remember
67: - whether they are tainted or not,
1.146 paf 68: and the lang which should be used to detaint them
1.41 paf 69: */
1.176 misha 70:
71:
1.145 paf 72: class String: public PA_Object {
1.1 paf 73: public:
1.48 paf 74:
1.146 paf 75: /** piece is tainted or not. the lang to use when detaint
1.106 parser 76: remember to change String_Untaint_lang_name @ untaint.C along
1.148 paf 77:
78: WARNING WARNING WARNING WARNING WARNING WARNING
79:
80: pos function compares(<=) languages, that is used in searching
81: for table column separator being L_CLEAN or L_AS_IS.
82: they search for AS_IS, meaning AS_IS|CLEAN [doing <=L_AS_IS check].
83:
84: letters assigned for debugging, but it's important for no language-letter
85: come before L_AS_IS other then L_CLEAN
86:
87: WARNING WARNING WARNING WARNING WARNING WARNING
1.106 parser 88: */
1.145 paf 89: enum Language {
1.146 paf 90: L_UNSPECIFIED=0, ///< no real string has parts of this lange: it's just convinient to check when string's empty
1.145 paf 91: // these two must go before others, there are checks for >L_AS_IS
1.170 misha 92: L_CLEAN='0', ///< clean WARNING: read above warning before changing
93: L_AS_IS='A', ///< leave all characters intact WARNING: read above warning before changing
1.122 paf 94:
1.148 paf 95: L_PASS_APPENDED='P',
1.41 paf 96: /**<
1.146 paf 97: leave lang built into string being appended.
1.41 paf 98: just a flag, that value not stored
99: */
1.170 misha 100: L_TAINTED='T', ///< tainted, untaint lang as assigned later
1.146 paf 101: // untaint langs. assigned by ^untaint[lang]{...}
1.170 misha 102: L_FILE_SPEC='F', ///< file specification
103: L_HTTP_HEADER='h', ///< text in HTTP response header
104: L_MAIL_HEADER='m', ///< text in mail header
1.192 misha 105: L_URI='U', ///< text in uri
106: L_SQL='Q', ///< ^table:sql body
107: L_JS='J', ///< JavaScript code
1.198 misha 108: L_XML='X', ///< ^xdoc:create xml
1.192 misha 109: L_HTML='H', ///< HTML code
1.198 misha 110: L_REGEX='R', ///< RegExp
111: L_JSON='S', ///< JSON code
1.170 misha 112: L_HTTP_COOKIE='C', ///< cookies encoded as %uXXXX for compartibility with js functions encode/decode
1.192 misha 113: L_PARSER_CODE='p', ///< ^process body
1.148 paf 114: // READ WARNING ABOVE BEFORE ADDING ANYTHING
1.146 paf 115: L_OPTIMIZE_BIT = 0x80 ///< flag, requiring cstr whitespace optimization
1.27 paf 116: };
117:
1.157 paf 118: enum Trim_kind {
119: TRIM_BOTH,
120: TRIM_START,
121: TRIM_END
122: };
123:
1.187 misha 124: class Body;
1.176 misha 125:
1.147 paf 126: union Languages {
1.146 paf 127:
1.147 paf 128: struct {
1.160 paf 129: #ifdef PA_LITTLE_ENDIAN
1.147 paf 130: Language lang:8;
1.179 misha 131: size_t is_not_just_lang:sizeof(CORD)*8-8;
1.160 paf 132: #elif defined(PA_BIG_ENDIAN)
1.179 misha 133: size_t is_not_just_lang:sizeof(CORD)*8-8;
1.160 paf 134: Language lang:8;
135: #else
136: # error word endianness not determined for some obscure reason
137: #endif
1.147 paf 138: } opt;
139: CORD langs;
1.146 paf 140:
1.187 misha 141: CORD make_langs(const Body& current) const {
142: return opt.is_not_just_lang?langs:CORD_chars((char)opt.lang, current.length());
1.146 paf 143: }
144:
145: CORD make_langs(size_t aoffset, size_t alength) const {
1.147 paf 146: return opt.is_not_just_lang?
1.194 misha 147: CORD_substr(langs, aoffset, alength, 0)
1.162 paf 148: :CORD_chars((char)opt.lang, alength);
1.145 paf 149: }
150:
1.157 paf 151: /// appending when 'langs' already contain something [simple cases handled elsewhere]
1.187 misha 152: void append(size_t current, const CORD to_nonempty_target_langs) {
1.146 paf 153: assert(langs);
1.187 misha 154: if(opt.is_not_just_lang)
155: langs=CORD_cat(langs, to_nonempty_target_langs);
156: else {
157: assert(current);
158: langs=CORD_cat(CORD_chars((char)opt.lang, current), to_nonempty_target_langs);
159: }
160: }
1.146 paf 161:
1.187 misha 162: void append(const Body& current, const CORD to_nonempty_target_langs) {
163: assert(langs);
1.147 paf 164: if(opt.is_not_just_lang)
1.146 paf 165: langs=CORD_cat(langs, to_nonempty_target_langs);
1.187 misha 166: else {
167: size_t current_size=current.length();
1.146 paf 168: assert(current_size);
1.187 misha 169: langs=CORD_cat(CORD_chars((char)opt.lang, current_size), to_nonempty_target_langs);
1.146 paf 170: }
1.145 paf 171: }
1.146 paf 172:
1.145 paf 173: public:
1.146 paf 174:
175: const char* v() const;
1.164 paf 176: void dump() const;
1.146 paf 177:
178: Languages(): langs(0) {}
1.147 paf 179: Languages(Language alang) {
180: opt.lang=alang;
181: opt.is_not_just_lang=0;
182: }
1.146 paf 183:
184: /// MUST be called exactly prior to modification of current [uses it's length]
1.187 misha 185: void append(size_t current, Language alang, size_t length) {
1.146 paf 186: assert(alang);
1.187 misha 187: assert(length);
1.146 paf 188:
1.208 moko 189: if(!opt.is_not_just_lang) {
1.147 paf 190: if(opt.lang) {
1.181 misha 191: if(opt.lang==alang) // same language? ignoring
1.146 paf 192: return;
193: } else {
1.147 paf 194: opt.lang=alang; // to uninitialized
1.146 paf 195: return;
1.145 paf 196: }
1.208 moko 197: }
1.146 paf 198:
1.187 misha 199: append(current, CORD_chars((char)alang, length));
1.146 paf 200: }
201:
1.187 misha 202: void append(const Body ¤t, Language alang, size_t length) {
1.176 misha 203: assert(alang);
1.187 misha 204: assert(length);
1.176 misha 205:
1.208 moko 206: if(!opt.is_not_just_lang) {
1.176 misha 207: if(opt.lang) {
1.183 misha 208: if(opt.lang==alang) // same language? ignoring
1.176 misha 209: return;
210: } else {
211: opt.lang=alang; // to uninitialized
212: return;
213: }
1.208 moko 214: }
1.176 misha 215:
1.187 misha 216: append(current, CORD_chars((char)alang, length));
1.176 misha 217: }
218:
1.187 misha 219: void appendHelper(const Body& current, Language alang, const Body &length_helper) {
220: assert(alang);
221:
1.208 moko 222: if(!opt.is_not_just_lang) {
1.187 misha 223: if(opt.lang) {
224: if(opt.lang==alang) // same language? ignoring
225: return;
226: } else {
227: opt.lang=alang; // to uninitialized
228: return;
229: }
1.208 moko 230: }
1.187 misha 231:
232: append(current, CORD_chars((char)alang, length_helper.length()));
233: }
234:
235: void appendHelper(const Body& current, const Languages &src, const Body& length_helper) {
1.186 misha 236: if(!langs){
1.176 misha 237: langs=src.langs; // to uninitialized
1.186 misha 238: #ifdef CORD_CAT_OPTIMIZATION
239: if(opt.is_not_just_lang && !CORD_IS_STRING(langs))
240: CORD_concatenation_protect(langs);
241: #endif
242: }
1.176 misha 243: else if(!src.opt.is_not_just_lang)
244: appendHelper(current, src.opt.lang, length_helper); // simplifying when simple source
245: else
246: append(current, src.make_langs(length_helper));
247: }
248:
1.146 paf 249: /// MUST be called exactly prior to modification of current [uses it's length]
1.187 misha 250: void append(const Body& current, const Languages src, size_t aoffset, size_t alength) {
1.146 paf 251: assert(alength);
252:
253: if(!langs) // to uninitialized?
1.147 paf 254: if(src.opt.is_not_just_lang)
1.194 misha 255: langs=CORD_substr(src.langs, aoffset, alength, 0); // to uninitialized complex
1.146 paf 256: else
1.147 paf 257: opt.lang=src.opt.lang; // to uninitialized simple
1.146 paf 258: else
1.147 paf 259: 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 260: return; // ignoring
261: else
262: append(current, src.make_langs(aoffset, alength));
263: }
264:
265: /// checks if we have lang<=alang all from aoffset to aoffset+alength
266: bool check_lang(Language alang, size_t aoffset, size_t alength) const {
267: if(alang==L_UNSPECIFIED) // ignore lang?
268: return true;
269:
1.147 paf 270: if(opt.is_not_just_lang)
1.146 paf 271: return CORD_range_contains_chr_greater_then(langs, aoffset, alength, (unsigned)alang)==0;
272: else
1.169 misha 273: return (unsigned)opt.lang<=(unsigned)alang;
1.146 paf 274: }
275:
1.148 paf 276: /// @returns count of blocks
277: /// @todo currently there can be adjucent blocks of same language. someday merge them
278: size_t count() const {
279: return opt.is_not_just_lang?
280: CORD_block_count(langs)
281: : opt.lang?
282: 1
283: : 0;
284: };
285:
1.187 misha 286: template<typename I>
287: void for_each(size_t current, int callback(char, size_t, I), I info) const {
288: if(opt.is_not_just_lang)
289: CORD_block_iter(langs, 0, (CORD_block_iter_fn)callback, info);
290: else
291: callback(opt.lang, current, info);
292: }
293:
294: template<typename I>
295: void for_each(const Body& current, int callback(char, size_t, I), I info) const {
1.147 paf 296: if(opt.is_not_just_lang)
1.146 paf 297: CORD_block_iter(langs, 0, (CORD_block_iter_fn)callback, info);
298: else
1.187 misha 299: callback(opt.lang, current.length(), info);
1.146 paf 300: }
301:
1.164 paf 302: bool invariant(size_t current_length) const {
1.146 paf 303: if(!langs)
304: return current_length==0;
1.147 paf 305: if(opt.is_not_just_lang)
1.146 paf 306: return CORD_len(langs)==current_length;
307: return true; // uncheckable, actually
308: }
309: };
310:
311: class Body {
312:
313: CORD body;
314:
1.183 misha 315: #ifdef HASH_CODE_CACHING
316: // cached hash code is not reseted on write operations as test shows
317: // that string body does not change after it is stored as a hash key
318: mutable uint hash_code;
319: #endif
320:
1.187 misha 321: #ifdef STRING_LENGTH_CACHING
322: // cached length is reseted on modification, used only for char*, not CORD
323: mutable size_t string_length;
324: #define INIT_LENGTH ,string_length(0)
325: #define ZERO_LENGTH string_length=0;
326: #else
327: #define INIT_LENGTH
328: #define ZERO_LENGTH
329: #endif
330:
1.146 paf 331: public:
332:
333: const char* v() const;
1.164 paf 334: void dump() const;
1.146 paf 335:
1.183 misha 336: #ifdef HASH_CODE_CACHING
1.187 misha 337: Body(): body(CORD_EMPTY), hash_code(0) INIT_LENGTH {}
338: Body(CORD abody, uint ahash_code): body(abody), hash_code(ahash_code) INIT_LENGTH {}
339: Body(CORD abody): body(abody), hash_code(0) INIT_LENGTH {
1.183 misha 340: #else
1.187 misha 341: Body(): body(CORD_EMPTY) INIT_LENGTH {}
342: Body(CORD abody): body(abody) INIT_LENGTH {
1.183 misha 343: #endif
1.186 misha 344: #ifdef CORD_CAT_OPTIMIZATION
345: assert(!body // no body
346: || *body // ordinary string
347: || body[1]==1 // CONCAT_HDR
348: || body[1]==3 // CONCAT_HDR_READ_ONLY
349: || body[1]==4 // FN_HDR
350: || body[1]==6 // SUBSTR_HDR
351: );
352: #else
1.146 paf 353: assert(!body // no body
354: || *body // ordinary string
355: || body[1]==1 // CONCAT_HDR
356: || body[1]==4 // FN_HDR
357: || body[1]==6 // SUBSTR_HDR
358: );
1.186 misha 359: #endif
1.146 paf 360: }
1.181 misha 361:
1.146 paf 362: static Body Format(int value);
363:
1.187 misha 364: void clear() { ZERO_LENGTH body=CORD_EMPTY; }
1.146 paf 365:
366: bool operator! () const { return is_empty(); }
367:
1.183 misha 368: CORD get_cord() const { return body; }
369: uint get_hash_code() const;
1.146 paf 370:
1.197 misha 371: const char* cstr() const {
372: #ifdef STRING_LENGTH_CACHING
373: string_length = length();
374: if(string_length)
375: return const_cast<Body*>(this)->body=CORD_to_const_char_star(body, string_length);
376: #endif
377: return CORD_to_const_char_star(body, length());
378: }
379:
1.194 misha 380: char* cstrm() const { return CORD_to_char_star(body, length()); }
1.146 paf 381:
1.187 misha 382: #ifdef STRING_LENGTH_CACHING
383: void set_length(size_t alength){ string_length = alength; }
384: size_t length() const { return body ? CORD_IS_STRING(body) ? string_length ? string_length : (string_length=strlen(body)) : CORD_len(body) : 0; }
385: #else
1.146 paf 386: size_t length() const { return CORD_len(body); }
1.187 misha 387: #endif
1.146 paf 388:
389: bool is_empty() const { return body==CORD_EMPTY; }
390:
391: void append_know_length(const char *str, size_t known_length) {
1.187 misha 392: if(known_length){
393: if(body){
394: body = CORD_cat_char_star(body, str, known_length);
395: ZERO_LENGTH
396: } else {
397: body=str;
398: #ifdef STRING_LENGTH_CACHING
399: string_length=known_length;
400: #endif
401: }
402: }
1.146 paf 403: }
404: void append_strdup_know_length(const char* str, size_t known_length) {
405: if(known_length)
406: append_know_length(pa_strdup(str, known_length), known_length);
407: }
1.187 misha 408: void append(char c) { ZERO_LENGTH body=CORD_cat_char(body, c); }
409: Body& operator << (const Body src) { ZERO_LENGTH body=CORD_cat(body, src.body); return *this; }
410:
1.146 paf 411: Body& operator << (const char* str) { append_know_length(str, strlen(str)); return *this; }
412:
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; }
416: bool operator >= (const Body src) const { return CORD_cmp(body, src.body)>=0; }
1.207 moko 417:
1.146 paf 418: bool operator != (const Body src) const { return CORD_cmp(body, src.body)!=0; }
419: bool operator == (const Body src) const { return CORD_cmp(body, src.body)==0; }
420:
1.207 moko 421: bool operator != (const char *src) const { return CORD_cmp(body, src)!=0; }
422: bool operator == (const char *src) const { return CORD_cmp(body, src)==0; }
423:
1.146 paf 424: int ncmp(size_t x_begin, const Body y, size_t y_begin, size_t size) const {
425: return CORD_ncmp(body, x_begin, y.body, y_begin, size);
1.145 paf 426: }
427:
1.146 paf 428: char fetch(size_t index) const { return CORD_fetch(body, index); }
1.194 misha 429: Body mid(size_t aindex, size_t alength) const { return CORD_substr(body, aindex, alength, length()); }
430: size_t pos(const char* substr, size_t offset=0) const { return CORD_str(body, offset, substr, length()); }
1.146 paf 431: size_t pos(const Body substr, size_t offset=0) const {
1.180 misha 432: if(substr.is_empty())
1.146 paf 433: return STRING_NOT_FOUND; // in this case CORD_str returns 0 [parser users got used to -1]
1.150 paf 434:
435: // CORD_str checks for bad offset [CORD_chr does not]
1.194 misha 436: return CORD_str(body, offset, substr.body, length());
1.146 paf 437: }
1.207 moko 438: size_t pos(char c, size_t offset=0) const {
1.150 paf 439: if(offset>=length()) // CORD_chr does not check that [and ABORT's in that case]
440: return STRING_NOT_FOUND;
441:
1.146 paf 442: return CORD_chr(body, offset, c);
1.145 paf 443: }
1.146 paf 444:
1.203 misha 445: size_t strrpbrk(const char* chars, size_t left, size_t right) const;
446:
447: size_t rskipchars(const char* chars, size_t left, size_t right) const;
448:
1.187 misha 449: template<typename I>
450: int for_each(int (*f)(char c, I), I info) const {
1.159 paf 451: return CORD_iter(body, (CORD_iter_fn)f, (void*)info);
452: }
453:
1.187 misha 454: template<typename I>
455: int for_each(int (*f1)(char c, I), int (*f2)(const char* s, I), I info) const {
1.159 paf 456: return CORD_iter5(body, 0, (CORD_iter_fn)f1, (CORD_batched_iter_fn)f2, info);
1.148 paf 457: }
1.146 paf 458:
459: void set_pos(CORD_pos& pos, size_t index) const { CORD_set_pos(pos, body, index); }
460:
1.157 paf 461: /// @returns this or 0 or mid. if returns this or 0 out_* are not filled
462: Body trim(Trim_kind kind=TRIM_BOTH, const char* chars=0,
1.188 misha 463: size_t* out_start=0, size_t* out_length=0, Charset* source_charset=0) const;
1.145 paf 464: };
465:
466: struct C {
467: const char *str;
468: size_t length;
469: operator const char *() { return str; }
1.158 paf 470: C(): str(0), length(0) {}
1.145 paf 471: C(const char *astr, size_t asize): str(astr), length(asize) {}
472: };
473:
474: struct Cm {
475: char *str;
476: size_t length;
477: //operator char *() { return str; }
1.158 paf 478: Cm(): str(0), length(0) {}
1.145 paf 479: Cm(char *astr, size_t asize): str(astr), length(asize) {}
480: };
481:
482: private:
483:
1.152 paf 484: Body body; ///< all characters of string
1.146 paf 485: Languages langs; ///< string characters lang
486:
487: const char* v() const;
1.164 paf 488: void dump() const;
489: #define ASSERT_STRING_INVARIANT(string) \
490: assert((string).langs.invariant((string).body.length()))
1.145 paf 491:
1.8 paf 492: public:
1.151 paf 493:
494: static const String Empty;
1.8 paf 495:
1.181 misha 496: explicit String(){};
1.182 misha 497: explicit String(const char* cstr, Language alang=L_CLEAN){
1.181 misha 498: if(cstr && *cstr){
499: body=cstr;
1.182 misha 500: langs=alang;
1.181 misha 501: }
502: }
1.187 misha 503: explicit String(const char* cstr, Language alang, size_t alength){
504: if(cstr && *cstr){
505: body=cstr;
506: #ifdef STRING_LENGTH_CACHING
507: body.set_length(alength);
508: #endif
1.182 misha 509: langs=alang;
1.181 misha 510: }
511: }
1.187 misha 512:
1.206 moko 513: String(int value, const char *format);
1.146 paf 514: String(Body abody, Language alang): body(abody), langs(alang) {
515: ASSERT_STRING_INVARIANT(*this);
516: }
517: String(const String& src): body(src.body), langs(src.langs) {
518: ASSERT_STRING_INVARIANT(*this);
1.145 paf 519: }
520:
521: /// for convinient hash lookup
1.183 misha 522: #ifdef HASH_CODE_CACHING
523: operator const Body&() const { return body; }
524: #else
1.146 paf 525: operator const Body() const { return body; }
1.183 misha 526: #endif
1.145 paf 527:
528: bool is_empty() const { return body.is_empty(); }
529: size_t length() const { return body.length(); }
1.171 misha 530: size_t length(Charset& charset) const;
1.145 paf 531:
1.191 misha 532: /// convert to CORD forcing lang tainting
533: Body cstr_to_string_body_taint(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const;
534: /// convert to CORD with tainting dirty to lang
535: Body cstr_to_string_body_untaint(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const;
1.145 paf 536:
1.189 misha 537: ///
538: const char* cstr() const {
539: return body.cstr();
540: }
541: ///
542: char* cstrm() const {
543: return body.cstrm();
544: }
545:
546: /// convert to constant C string forcing lang tainting
1.190 misha 547: const char* taint_cstr(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 548: return cstr_to_string_body_taint(lang, connection, charsets).cstr();
549: }
1.190 misha 550: char *taint_cstrm(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 551: return cstr_to_string_body_taint(lang, connection, charsets).cstrm();
552: }
553:
554: /// convert to constant C string with tainting dirty to lang
1.190 misha 555: const char* untaint_cstr(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 556: return cstr_to_string_body_untaint(lang, connection, charsets).cstr();
557: }
1.190 misha 558: char *untaint_cstrm(Language lang, SQL_Connection* connection=0, const Request_charsets *charsets=0) const {
1.189 misha 559: return cstr_to_string_body_untaint(lang, connection, charsets).cstrm();
560: }
1.195 misha 561:
1.196 misha 562: const char* untaint_and_transcode_cstr(Language lang, const Request_charsets *charsets) const;
1.195 misha 563:
1.202 moko 564: bool is_not_just_lang() const {
565: return langs.opt.is_not_just_lang !=0;
566: }
567:
568: Language just_lang() const {
569: return langs.opt.lang;
570: }
571:
1.108 parser 572: /// puts pieces to buf
1.145 paf 573: Cm serialize(size_t prolog_size) const;
1.108 parser 574: /// appends pieces from buf to self
1.145 paf 575: bool deserialize(size_t prolog_size, void *buf, size_t buf_size);
1.146 paf 576: /// @see Body::append_know_length
1.145 paf 577: String& append_know_length(const char* str, size_t known_length, Language lang);
1.146 paf 578: /// @see Body::append_help_length
1.145 paf 579: String& append_help_length(const char* str, size_t helper_length, Language lang);
580: String& append_strdup(const char* str, size_t helper_length, Language lang);
581:
1.146 paf 582: bool operator == (const char* y) const { return body==Body(y); }
583: bool operator != (const char* y) const { return body!=Body(y); }
1.145 paf 584:
585: /// this starts with y
586: bool starts_with(const char* y) const {
1.146 paf 587: return body.ncmp(0/*x_begin*/, Body(y), 0/*y_begin*/, strlen(y))==0;
1.145 paf 588: }
589: /// x starts with this
590: bool this_starts(const char* x) const {
1.146 paf 591: return Body(x).ncmp(0/*x_begin*/, body, 0/*y_begin*/, length())==0;
1.26 paf 592: }
593:
1.187 misha 594: String& append_to(String& dest, Language lang, bool forced=false) const;
1.145 paf 595: String& append(const String& src, Language lang, bool forced=false) {
596: return src.append_to(*this, lang, forced);
597: }
1.199 misha 598: String& append_quoted(const String* src, Language lang=L_JSON){
599: *this << "\"";
600: if(src)
601: this->append(*src, lang, true/*forced lang*/);
602: *this << "\"";
603: return *this;
604: }
605:
1.145 paf 606: String& operator << (const String& src) { return append(src, L_PASS_APPENDED); }
607: String& operator << (const char* src) { return append_help_length(src, 0, L_AS_IS); }
1.187 misha 608: String& operator << (const Body& src){
609: langs.appendHelper(body, L_AS_IS, src);
610: body<<src;
611: return *this;
612: }
1.100 parser 613:
1.142 paf 614: char first_char() const {
1.145 paf 615: return is_empty()?0:body.fetch(0);
1.142 paf 616: }
1.54 paf 617:
1.205 moko 618: char last_char() const {
619: return is_empty()?0:body.fetch(body.length()-1);
620: }
621:
1.145 paf 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: bool operator == (const String& src) const { return body==src.body; }
628:
1.54 paf 629: /// extracts [start, finish) piece of string
1.145 paf 630: String& mid(size_t substr_begin, size_t substr_end) const;
1.172 misha 631: String& mid(Charset& charset, size_t from, size_t to, size_t helper_length=0) const;
1.145 paf 632:
633: /**
634: ignore lang if it's L_UNSPECIFIED
635: but when specified: look for substring that lies in ONE fragment in THAT lang
636: @return position of substr in string, -1 means "not found" [const char* version]
637: */
1.146 paf 638: size_t pos(const Body substr,
1.145 paf 639: size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
640: /// String version of @see pos(const char*, int, Language)
641: size_t pos(const String& substr,
642: size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
643: size_t pos(char c,
644: size_t this_offset=0) const {
645: return body.pos(c, this_offset);
646: }
1.173 misha 647: size_t pos(Charset& charset,
1.171 misha 648: const String& substr,
649: size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
1.55 paf 650:
1.203 misha 651: size_t strrpbrk(const char* chars, size_t left=0) const {
652: return (length()) ? body.strrpbrk(chars, left, length()-1) : STRING_NOT_FOUND;
653: }
654: size_t strrpbrk(const char* chars, size_t left, size_t right) const {
655: return body.strrpbrk(chars, left, right);
656: }
657:
658: size_t rskipchars(const char* chars, size_t left=0) const {
659: return (length()) ? body.rskipchars(chars, left, length()-1) : STRING_NOT_FOUND;
660: }
661: size_t rskipchars(const char* chars, size_t left, size_t right) const {
662: return body.rskipchars(chars, left, right);
663: }
664:
1.145 paf 665: void split(ArrayString& result,
666: size_t& pos_after,
667: const char* delim,
668: Language lang=L_UNSPECIFIED, int limit=-1) const;
669: void split(ArrayString& result,
670: size_t& pos_after,
1.62 paf 671: const String& delim,
1.145 paf 672: Language lang=L_UNSPECIFIED, int limit=-1) const;
1.62 paf 673:
1.145 paf 674: typedef void (*Row_action)(Table& table, ArrayString* row,
1.136 paf 675: int prestart, int prefinish,
676: int poststart, int postfinish,
1.68 paf 677: void *info);
1.87 parser 678: /**
1.145 paf 679: @return table of found items, if any.
1.87 parser 680: table format is defined and fixed[can be used by others]:
681: @verbatim
682: prematch/match/postmatch/1/2/3/...
683: @endverbatim
684: */
1.178 misha 685: Table* match(VRegex* vregex,
1.99 parser 686: Row_action row_action, void *info,
1.168 misha 687: int& matches_count) const;
1.87 parser 688: enum Change_case_kind {
689: CC_UPPER,
690: CC_LOWER
691: };
1.145 paf 692: String& change_case(Charset& source_charset,
1.87 parser 693: Change_case_kind kind) const;
1.145 paf 694: const String& replace(const Dictionary& dict) const;
1.188 misha 695: const String& trim(Trim_kind kind=TRIM_BOTH, const char* chars=0, Charset* source_charset=0) const;
1.155 paf 696: double as_double() const { return pa_atod(cstr(), this); }
697: int as_int() const { return pa_atoi(cstr(), this); }
1.167 misha 698: bool as_bool() const { return as_int()!=0; }
1.174 misha 699: const String& escape(Charset& source_charset) const;
1.137 paf 700:
1.7 paf 701: private: //disabled
702:
1.12 paf 703: String& operator = (const String&) { return *this; }
1.7 paf 704:
1.1 paf 705: };
1.119 paf 706:
1.183 misha 707: #ifndef HASH_CODE_CACHING
1.145 paf 708: /// simple hash code of string. used by Hash
1.146 paf 709: inline uint hash_code(const String::Body self) {
1.184 misha 710: return self.get_hash_code();
1.119 paf 711: }
1.183 misha 712: #endif
1.147 paf 713:
714:
1.1 paf 715: #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