Annotation of parser3/src/include/pa_string.h, revision 1.144.2.28.2.41
1.41 paf 1: /** @file
1.43 paf 2: Parser: string class decl.
3:
1.144.2.18 paf 4: Copyright (c) 2001-2003 ArtLebedev Group (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:
1.144.2.28.2. 1(paf 11:3): static const char* IDENT_STRING_H="$Date: 2003/03/27 10:53:01 $";
1.144.2.10 paf 12:
13: // includes
1.1 paf 14:
1.144.2.12 paf 15: #include "pa_types.h"
16: #include "pa_array.h"
1.4 paf 17:
1.144.2.28.2. (paf 18:): extern "C" { // cord's author forgot to do that
7(paf 19:3): #define CORD_NO_IO
(paf 20:): #include "cord.h"
21:): };
22:):
1.144.2.12 paf 23: // forwards
1.144.2.10 paf 24:
1.144.2.28 paf 25: class Charset;
1.144.2.28.2. (paf 26:): class Table;
1.144.2.5 paf 27: class SQL_Connection;
28: class Dictionary;
1.144.2.15 paf 29: class Request_charsets;
1.144.2.28.2. (paf 30:): class String;
7(paf 31:3): typedef Array<const String*> ArrayString;
1.144.2.5 paf 32:
1.144.2.28.2. (paf 33:): /// this is result of pos functions which mean that substr were not found
34:): #define STRING_NOT_FOUND ((size_t)-1)
35:):
7(paf 36:3): class StringBody {
37:3):
38:3): CORD body;
39:3):
40:3): public:
41:3):
42:3): StringBody(): body(CORD_EMPTY) {}
5(paf 43:3): StringBody(CORD abody): body(abody) {
44:3): #ifdef _DEBUG
45:3): if(body && !*body)
46:3): switch(*(char*)(body+1)) {
47:3): case 1: case 4: case 6: break;
48:3): default: _asm int 3; break;
49:3): }
50:3): #endif
51:3): }
1(paf 52:3): /// WARNING: length is only HELPER length, str in ANY case should be zero-terminated
53:3): StringBody(const char* str, size_t helper_length): body(CORD_EMPTY) {
54:3): append_help_length(str, helper_length);
2(paf 55:3): }
4(paf 56:3): static StringBody Format(int value);
7(paf 57:3):
58:3): void clear() { body=CORD_EMPTY; }
59:3):
60:3): bool operator! () const { return is_empty(); }
61:3):
62:3): uint hash_code() const;
63:3):
0(paf 64:3): const char* cstr() const { return CORD_to_const_char_star(body); }
1(paf 65:3): char* cstrm() const { return CORD_to_char_star(body); }
0(paf 66:3):
7(paf 67:3): size_t length() const { return CORD_len(body); }
68:3):
69:3): bool is_empty() const { return body==CORD_EMPTY; }
70:3):
1(paf 71:3): void append_help_length(const char *str, size_t help_length) {
5(paf 72:3): #ifdef _DEBUG
73:3): if(!*str)
74:3): _asm int 3;
75:3): bool were_empty=cstr==CORD_EMPTY;
76:3): #endif
1(paf 77:3): body=CORD_cat_char_star(body, str, help_length);
5(paf 78:3): #ifdef _DEBUG
79:3): if(were_empty && !*str)
80:3): _asm int 3;
81:3): #endif
7(paf 82:3): }
1(paf 83:3): String& append_strdup(const char* str, size_t length) {
84:3): append_help_length(pa_strdup(str, length), length);
85:3): }
2(paf 86:3): void append(char c) { body=CORD_cat_char(body, c); }
6(paf 87:3): StringBody& operator << (const StringBody src) { body=CORD_cat(body, src.body); return *this; }
1(paf 88:3): StringBody& operator << (const char* str) { body=CORD_cat_char_star(body, str, 0); return *this; }
7(paf 89:3):
90:3): // could not figure out why this operator is needed [should do this chain: string->simple->==]
8(paf 91:3): bool operator < (const StringBody src) const { return CORD_cmp(body, src.body)<0; }
92:3): bool operator > (const StringBody src) const { return CORD_cmp(body, src.body)>0; }
93:3): bool operator <= (const StringBody src) const { return CORD_cmp(body, src.body)<=0; }
94:3): bool operator >= (const StringBody src) const { return CORD_cmp(body, src.body)>=0; }
95:3): bool operator != (const StringBody src) const { return CORD_cmp(body, src.body)!=0; }
96:3): bool operator == (const StringBody src) const { return CORD_cmp(body, src.body)==0; }
7(paf 97:3):
98:3): int ncmp(size_t x_begin, const StringBody y, size_t y_begin, size_t size) const {
99:3): return CORD_ncmp(body, x_begin, y.body, y_begin, size);
100:3): }
101:3):
102:3): char fetch(size_t index) const { return CORD_fetch(body, index); }
1(paf 103:3): StringBody mid(size_t index, size_t length) const { return CORD_substr(body, index, length); }
3(paf 104:3): size_t pos(const char* substr, size_t offset=0) const { return CORD_str(body, offset, substr); }
1(paf 105:3): size_t pos(const StringBody substr, size_t offset=0) const {
106:3): if(!substr.length())
107:3): return STRING_NOT_FOUND; // in this case CORD_str returns 0 [parser users got used to -1]
108:3): return CORD_str(body, offset, substr.body);
109:3): }
2(paf 110:3): size_t pos(char c,
111:3): size_t offset=0) const {
112:3): return CORD_chr(body, offset, c);
113:3): }
1(paf 114:3):
115:3): template<typename I> void for_each(int (*callback)(const char* s, I), I info) const {
116:3): CORD_iter5(body, 0, 0, (CORD_batched_iter_fn)callback, info);
117:3): }
118:3):
119:3): void set_pos(CORD_pos& pos, size_t index) const { CORD_set_pos(pos, body, index); }
7(paf 120:3):
121:3): };
(paf 122:):
1.144.2.2 paf 123: /**
124: String which knows the language of all it's fragments.
1.41 paf 125:
126: All pieces remember
127: - whether they are tainted or not,
128: and the language which should be used to detaint them
129: */
1.144.2.28.2. (paf 130:): class String: public PA_Object {
131:):
2(paf 132:3): // friend class StringBody;
7(paf 133:3):
1.144.2.2 paf 134: public:
1.1 paf 135:
1.106 parser 136: /** piece is tainted or not. the language to use when detaint
137: remember to change String_Untaint_lang_name @ untaint.C along
138: */
1.144.2.28.2. 7(paf 139:3): enum Language {
140:3): L_UNSPECIFIED=0, ///< zero value handy for hash lookup @see untaint_lang_name2enum
141:3): // these two must go before others, there are checks for >L_AS_IS
142:3): L_CLEAN, ///< clean
143:3): L_AS_IS, ///< leave all characters intact
1.122 paf 144:
1.144.2.28.2. 7(paf 145:3): L_PASS_APPENDED,
1.41 paf 146: /**<
147: leave language built into string being appended.
148: just a flag, that value not stored
149: */
1.144.2.28.2. 7(paf 150:3): L_TAINTED, ///< tainted, untaint language as assigned later
1.122 paf 151: // untaint languages. assigned by ^untaint[lang]{...}
1.144.2.28.2. 7(paf 152:3): L_FILE_SPEC, ///< file specification
153:3): L_HTTP_HEADER, ///< text in HTTP response header
154:3): L_MAIL_HEADER, ///< text in mail header
155:3): L_URI, ///< text in uri
156:3): L_TABLE, ///< ^table:set body
157:3): L_SQL, ///< ^table:sql body
158:3): L_JS, ///< JavaScript code
159:3): L_XML, ///< ^dom:set xml
160:3): L_HTML, ///< HTML code (for editing)
161:3): L_OPTIMIZE_BIT = 0x8000 ///< flag, requiring cstr whitespace optimization
1.27 paf 162: };
163:
1.144.2.28.2. (paf 164:): struct Fragment {
7(paf 165:3): Language lang; ///< untaint flag, later untaint language
0(paf 166:3): size_t length; ///< length
9(paf 167:3): Fragment(Language alang, size_t asize): lang(alang), length(asize) {
0(paf 168:3): #ifdef _DEBUG
9(paf 169:3): if(alang==L_UNSPECIFIED)
0(paf 170:3): _asm int 3; // MUST NOT
171:3): if(asize==(size_t)-1)
172:3): _asm int 3; // MUST NOT
173:3): #endif
9(paf 174:3): }
(paf 175:): };
176:):
177:): class ArrayFragment: public Array<Fragment> {
6(paf 178:3): void append(element_type src) {
(paf 179:): *static_cast<Array<Fragment> *>(this)+=src;
180:): }
6(paf 181:3): /// hiding from accidental USE, use append_positions
182:3): void append(const ArrayFragment& src, int offset, int limit) {
183:3): static_cast<Array<Fragment> *>(this)->append(src, offset, limit);
184:3): }
(paf 185:): public:
186:): ArrayFragment& operator += (element_type src) {
8(paf 187:3): if(size_t lcount=count()) { // not empty?
(paf 188:): // try to join with last
0(paf 189:3): Fragment& last=unchecked_get_ref(lcount-1);
(paf 190:): if(last.lang==src.lang) {
0(paf 191:3): last.length+=src.length;
(paf 192:): return *this;
193:): }
194:): }
6(paf 195:3): append(src);
(paf 196:): return *this;
197:): }
6(paf 198:3): void append(const ArrayFragment& src) { append(src, 0, 0); }
199:3): void append_positions(const ArrayFragment& src, size_t substr_begin, size_t substr_end);
(paf 200:): };
201:):
202:): struct C {
203:): char *str;
0(paf 204:3): size_t length;
(paf 205:): operator char *() { return str; }
0(paf 206:3): C(char *astr, size_t asize): str(astr), length(asize) {}
(paf 207:): };
208:):
209:): private:
210:):
7(paf 211:3): StringBody body; ///< all characters of string
0(paf 212:3): ArrayFragment fragments; ///< fragment language+length info
(paf 213:):
1.8 paf 214: public:
215:
1.144.2.28.2. 1(paf 216:3): explicit String(const char* cstr=0, size_t help_length=0, bool tainted=false);
1.14 paf 217: String(const String& src);
1.144.2.28.2. 7(paf 218:3): String(StringBody abody, Language alang): body(abody) {
219:3): fragments+=Fragment(alang, abody.length());
220:3): }
221:3):
222:3): /// for convinient hash lookup
2(paf 223:3): operator const StringBody() const { return body; }
7(paf 224:3):
225:3): bool is_empty() const { return body.is_empty(); }
226:3): size_t length() const { return body.length(); }
(paf 227:):
2(paf 228:3): /// convert to CORD. if 'lang' known, forcing 'lang' to it
1(paf 229:3): StringBody cstr_to_string_body(Language lang=L_AS_IS,
7(paf 230:3): SQL_Connection* connection=0,
(paf 231:): const Request_charsets *charsets=0) const;
232:):
2(paf 233:3): /// convert to constant C string. if 'lang' known, forcing 'lang' to it
7(paf 234:3): const char* cstr(Language lang=L_AS_IS,
235:3): SQL_Connection* connection=0,
1.144.2.14 paf 236: const Request_charsets *charsets=0) const {
1.144.2.28.2. 1(paf 237:3): return cstr_to_string_body(lang, connection, charsets).cstr();
(paf 238:): }
2(paf 239:3): /// convert to Modifiable C string. if 'lang' known, forcing 'lang' to it
7(paf 240:3): char *cstrm(Language lang=L_AS_IS,
241:3): SQL_Connection* connection=0,
(paf 242:): const Request_charsets *charsets=0) const {
1(paf 243:3): return cstr_to_string_body(lang, connection, charsets).cstrm();
1.144.2.9 paf 244: }
1.108 parser 245: /// puts pieces to buf
1.144.2.28.2. (paf 246:): C serialize(size_t prolog_size) const;
1.108 parser 247: /// appends pieces from buf to self
1.144.2.28.2. 4(paf 248:3): bool deserialize(size_t prolog_size, void *buf, size_t buf_size);
1(paf 249:3): /// @see StringBody::append_help_length
250:3): String& append_help_length(const char* str, size_t helper_length, Language lang);
251:3): String& append_strdup(const char* str, size_t length, Language lang);
(paf 252:):
7(paf 253:3): bool operator == (const char* y) const { return body==StringBody(y); }
0(paf 254:3): bool operator != (const char* y) const { return body!=StringBody(y); }
(paf 255:):
256:): /// this starts with y
7(paf 257:3): bool starts_with(const char* y) const {
258:3): return body.ncmp(0/*x_begin*/, StringBody(y), 0/*y_begin*/, strlen(y))==0;
1.80 paf 259: }
1.144.2.28.2. 9(paf 260:3): /// x starts with this
261:3): bool this_starts(const char* x) const {
262:3): return StringBody(x).ncmp(0/*x_begin*/, body, 0/*y_begin*/, length())==0;
(paf 263:): }
1.80 paf 264:
1.144.2.28.2. 7(paf 265:3): String& append_to(String& dest, Language lang, bool forced) const;
266:3): String& append(const String& src, Language lang, bool forced=false) {
1.144.2.11 paf 267: return src.append_to(*this, lang, forced);
268: }
1.144.2.28.2. 7(paf 269:3): String& operator << (const String& src) { return append(src, L_PASS_APPENDED); }
1(paf 270:3): String& operator << (const char* src) { return append_help_length(src, 0, L_AS_IS); }
8(paf 271:3): String& operator << (const StringBody src) {
6(paf 272:3): body<<src;
9(paf 273:3): fragments+=Fragment(L_AS_IS, src.length());
274:3): return *this;
275:3): }
1.100 parser 276:
1.142 paf 277: /// extracts first char of a string, if any
278: char first_char() const {
1.144.2.28.2. 7(paf 279:3): return is_empty()?0:body.fetch(0);
1.142 paf 280: }
1.144.2.28.2. 8(paf 281:3):
282:3): bool operator < (const String& src) const { return body<src.body; }
283:3): bool operator > (const String& src) const { return body>src.body; }
284:3): bool operator <= (const String& src) const { return body<=src.body; }
285:3): bool operator >= (const String& src) const { return body>=src.body; }
286:3): bool operator != (const String& src) const { return body!=src.body; }
287:3): bool operator == (const String& src) const { return body==src.body; }
1.54 paf 288:
289: /// extracts [start, finish) piece of string
1.144.2.28.2. 2(paf 290:3): String& mid(size_t substr_begin, size_t substr_end) const;
(paf 291:):
292:): /**
7(paf 293:3): ignore lang if it's L_UNSPECIFIED
(paf 294:): but when specified: look for substring that lies in ONE fragment in THAT lang
295:): @return position of substr in string, -1 means "not found" [const char* version]
296:): */
1(paf 297:3): size_t pos(const StringBody substr,
7(paf 298:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
299:3): /// String version of @see pos(const char*, int, Language)
(paf 300:): size_t pos(const String& substr,
7(paf 301:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
2(paf 302:3): size_t pos(char c,
303:3): size_t this_offset=0) const {
304:3): return body.pos(c, this_offset);
305:3): }
1.62 paf 306:
1.144.2.22 paf 307: void split(ArrayString& result,
1.144.2.28.2. 2(paf 308:3): size_t& pos_after,
(paf 309:): const char* delim,
2(paf 310:3): Language lang=L_UNSPECIFIED, int limit=-1) const;
1.144.2.22 paf 311: void split(ArrayString& result,
1.144.2.28.2. 5(paf 312:3): size_t& pos_after,
1.62 paf 313: const String& delim,
1.144.2.28.2. 7(paf 314:3): Language lang=L_UNSPECIFIED, int limit=-1) const;
1.62 paf 315:
1.144.2.28.2. 6(paf 316:3): typedef void (*Row_action)(Table& table, ArrayString* row,
1.136 paf 317: int prestart, int prefinish,
318: int poststart, int postfinish,
1.144.2.9 paf 319: void *info);
1.87 parser 320: /**
1.144.2.5 paf 321: @return table of found items, if any.
1.87 parser 322: table format is defined and fixed[can be used by others]:
323: @verbatim
324: prematch/match/postmatch/1/2/3/...
325: @endverbatim
326: */
1.144.2.28.2. (paf 327:): Table* match(Charset& source_charset,
1.64 paf 328: const String& regexp,
1.144.2.28.2. 5(paf 329:3): const String* options,
1.99 parser 330: Row_action row_action, void *info,
1.144.2.27 paf 331: bool& just_matched) const;
1.87 parser 332: enum Change_case_kind {
333: CC_UPPER,
334: CC_LOWER
335: };
1.144.2.28.2. (paf 336:): String& change_case(Charset& source_charset,
5(paf 337:3): Change_case_kind kind) const;
(paf 338:): const String& replace(const Dictionary& dict) const;
339:): double as_double() const;
340:): int as_int() const;
1.137 paf 341:
1.7 paf 342: private: //disabled
343:
1.12 paf 344: String& operator = (const String&) { return *this; }
1.7 paf 345:
1.1 paf 346: };
1.144.2.7 paf 347:
348: /// simple hash code of string. used by Hash
1.144.2.28.2. 8(paf 349:3): inline uint hash_code(const StringBody self) {
1.144.2.7 paf 350: return self.hash_code();
351: }
1.1 paf 352:
353: #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