Annotation of parser3/src/include/pa_string.h, revision 1.144.2.28.2.45
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. 5(paf 11:3): static const char* IDENT_STRING_H="$Date: 2003/04/02 08:35:43 $";
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) {
3(paf 44:3): assert(!body // no body
45:3): || *body // ordinary string
46:3): || body[1]==1 // CONCAT_HDR
47:3): || body[1]==4 // FN_HDR
48:3): || body[1]==6 // SUBSTR_HDR
49:3): );
5(paf 50:3): }
1(paf 51:3): /// WARNING: length is only HELPER length, str in ANY case should be zero-terminated
52:3): StringBody(const char* str, size_t helper_length): body(CORD_EMPTY) {
53:3): append_help_length(str, helper_length);
2(paf 54:3): }
4(paf 55:3): static StringBody Format(int value);
7(paf 56:3):
57:3): void clear() { body=CORD_EMPTY; }
58:3):
59:3): bool operator! () const { return is_empty(); }
60:3):
61:3): uint hash_code() const;
62:3):
0(paf 63:3): const char* cstr() const { return CORD_to_const_char_star(body); }
1(paf 64:3): char* cstrm() const { return CORD_to_char_star(body); }
0(paf 65:3):
7(paf 66:3): size_t length() const { return CORD_len(body); }
67:3):
68:3): bool is_empty() const { return body==CORD_EMPTY; }
69:3):
1(paf 70:3): void append_help_length(const char *str, size_t help_length) {
3(paf 71:3): #ifndef NDEBUG
72:3): bool were_empty=is_empty();
5(paf 73:3): #endif
1(paf 74:3): body=CORD_cat_char_star(body, str, help_length);
3(paf 75:3):
76:3): assert(!were_empty || body[0] /*append to empty must produce simple string*/);
7(paf 77:3): }
2(paf 78:3): void append_strdup(const char* str, size_t length) {
1(paf 79:3): append_help_length(pa_strdup(str, length), length);
80:3): }
2(paf 81:3): void append(char c) { body=CORD_cat_char(body, c); }
6(paf 82:3): StringBody& operator << (const StringBody src) { body=CORD_cat(body, src.body); return *this; }
1(paf 83:3): StringBody& operator << (const char* str) { body=CORD_cat_char_star(body, str, 0); return *this; }
7(paf 84:3):
85:3): // could not figure out why this operator is needed [should do this chain: string->simple->==]
8(paf 86:3): bool operator < (const StringBody src) const { return CORD_cmp(body, src.body)<0; }
87:3): bool operator > (const StringBody src) const { return CORD_cmp(body, src.body)>0; }
88:3): bool operator <= (const StringBody src) const { return CORD_cmp(body, src.body)<=0; }
89:3): bool operator >= (const StringBody src) const { return CORD_cmp(body, src.body)>=0; }
90:3): bool operator != (const StringBody src) const { return CORD_cmp(body, src.body)!=0; }
91:3): bool operator == (const StringBody src) const { return CORD_cmp(body, src.body)==0; }
7(paf 92:3):
93:3): int ncmp(size_t x_begin, const StringBody y, size_t y_begin, size_t size) const {
94:3): return CORD_ncmp(body, x_begin, y.body, y_begin, size);
95:3): }
96:3):
97:3): char fetch(size_t index) const { return CORD_fetch(body, index); }
1(paf 98:3): StringBody mid(size_t index, size_t length) const { return CORD_substr(body, index, length); }
3(paf 99:3): size_t pos(const char* substr, size_t offset=0) const { return CORD_str(body, offset, substr); }
1(paf 100:3): size_t pos(const StringBody substr, size_t offset=0) const {
101:3): if(!substr.length())
102:3): return STRING_NOT_FOUND; // in this case CORD_str returns 0 [parser users got used to -1]
103:3): return CORD_str(body, offset, substr.body);
104:3): }
2(paf 105:3): size_t pos(char c,
106:3): size_t offset=0) const {
107:3): return CORD_chr(body, offset, c);
108:3): }
1(paf 109:3):
110:3): template<typename I> void for_each(int (*callback)(const char* s, I), I info) const {
111:3): CORD_iter5(body, 0, 0, (CORD_batched_iter_fn)callback, info);
112:3): }
113:3):
114:3): void set_pos(CORD_pos& pos, size_t index) const { CORD_set_pos(pos, body, index); }
7(paf 115:3):
5(paf 116:3): StringBody normalize() const {
117:3): return StringBody(CORD_balance(body));
118:3): }
119:3):
120:3): void dump() const {
121:3): CORD_dump(body);
122:3): }
7(paf 123:3): };
(paf 124:):
1.144.2.2 paf 125: /**
126: String which knows the language of all it's fragments.
1.41 paf 127:
128: All pieces remember
129: - whether they are tainted or not,
130: and the language which should be used to detaint them
131: */
1.144.2.28.2. (paf 132:): class String: public PA_Object {
133:):
2(paf 134:3): // friend class StringBody;
7(paf 135:3):
1.144.2.2 paf 136: public:
1.1 paf 137:
1.106 parser 138: /** piece is tainted or not. the language to use when detaint
139: remember to change String_Untaint_lang_name @ untaint.C along
140: */
1.144.2.28.2. 7(paf 141:3): enum Language {
142:3): L_UNSPECIFIED=0, ///< zero value handy for hash lookup @see untaint_lang_name2enum
143:3): // these two must go before others, there are checks for >L_AS_IS
144:3): L_CLEAN, ///< clean
145:3): L_AS_IS, ///< leave all characters intact
1.122 paf 146:
1.144.2.28.2. 7(paf 147:3): L_PASS_APPENDED,
1.41 paf 148: /**<
149: leave language built into string being appended.
150: just a flag, that value not stored
151: */
1.144.2.28.2. 7(paf 152:3): L_TAINTED, ///< tainted, untaint language as assigned later
1.122 paf 153: // untaint languages. assigned by ^untaint[lang]{...}
1.144.2.28.2. 7(paf 154:3): L_FILE_SPEC, ///< file specification
155:3): L_HTTP_HEADER, ///< text in HTTP response header
156:3): L_MAIL_HEADER, ///< text in mail header
157:3): L_URI, ///< text in uri
158:3): L_TABLE, ///< ^table:set body
159:3): L_SQL, ///< ^table:sql body
160:3): L_JS, ///< JavaScript code
161:3): L_XML, ///< ^dom:set xml
162:3): L_HTML, ///< HTML code (for editing)
163:3): L_OPTIMIZE_BIT = 0x8000 ///< flag, requiring cstr whitespace optimization
1.27 paf 164: };
165:
1.144.2.28.2. (paf 166:): struct Fragment {
7(paf 167:3): Language lang; ///< untaint flag, later untaint language
0(paf 168:3): size_t length; ///< length
9(paf 169:3): Fragment(Language alang, size_t asize): lang(alang), length(asize) {
3(paf 170:3): assert(alang!=L_UNSPECIFIED);
171:3): assert(asize!=(size_t)-1);
9(paf 172:3): }
(paf 173:): };
174:):
175:): class ArrayFragment: public Array<Fragment> {
6(paf 176:3): void append(element_type src) {
(paf 177:): *static_cast<Array<Fragment> *>(this)+=src;
178:): }
6(paf 179:3): /// hiding from accidental USE, use append_positions
180:3): void append(const ArrayFragment& src, int offset, int limit) {
181:3): static_cast<Array<Fragment> *>(this)->append(src, offset, limit);
182:3): }
(paf 183:): public:
184:): ArrayFragment& operator += (element_type src) {
8(paf 185:3): if(size_t lcount=count()) { // not empty?
(paf 186:): // try to join with last
4(paf 187:3): Fragment& last=get_ref(lcount-1);
(paf 188:): if(last.lang==src.lang) {
0(paf 189:3): last.length+=src.length;
(paf 190:): return *this;
191:): }
192:): }
6(paf 193:3): append(src);
(paf 194:): return *this;
195:): }
6(paf 196:3): void append(const ArrayFragment& src) { append(src, 0, 0); }
197:3): void append_positions(const ArrayFragment& src, size_t substr_begin, size_t substr_end);
(paf 198:): };
199:):
200:): struct C {
201:): char *str;
0(paf 202:3): size_t length;
(paf 203:): operator char *() { return str; }
0(paf 204:3): C(char *astr, size_t asize): str(astr), length(asize) {}
(paf 205:): };
206:):
207:): private:
208:):
7(paf 209:3): StringBody body; ///< all characters of string
0(paf 210:3): ArrayFragment fragments; ///< fragment language+length info
(paf 211:):
1.8 paf 212: public:
213:
1.144.2.28.2. 1(paf 214:3): explicit String(const char* cstr=0, size_t help_length=0, bool tainted=false);
1.14 paf 215: String(const String& src);
1.144.2.28.2. 7(paf 216:3): String(StringBody abody, Language alang): body(abody) {
217:3): fragments+=Fragment(alang, abody.length());
218:3): }
219:3):
220:3): /// for convinient hash lookup
2(paf 221:3): operator const StringBody() const { return body; }
7(paf 222:3):
223:3): bool is_empty() const { return body.is_empty(); }
224:3): size_t length() const { return body.length(); }
(paf 225:):
2(paf 226:3): /// convert to CORD. if 'lang' known, forcing 'lang' to it
1(paf 227:3): StringBody cstr_to_string_body(Language lang=L_AS_IS,
7(paf 228:3): SQL_Connection* connection=0,
(paf 229:): const Request_charsets *charsets=0) const;
230:):
2(paf 231:3): /// convert to constant C string. if 'lang' known, forcing 'lang' to it
7(paf 232:3): const char* cstr(Language lang=L_AS_IS,
233:3): SQL_Connection* connection=0,
1.144.2.14 paf 234: const Request_charsets *charsets=0) const {
1.144.2.28.2. 1(paf 235:3): return cstr_to_string_body(lang, connection, charsets).cstr();
(paf 236:): }
2(paf 237:3): /// convert to Modifiable C string. if 'lang' known, forcing 'lang' to it
7(paf 238:3): char *cstrm(Language lang=L_AS_IS,
239:3): SQL_Connection* connection=0,
(paf 240:): const Request_charsets *charsets=0) const {
1(paf 241:3): return cstr_to_string_body(lang, connection, charsets).cstrm();
1.144.2.9 paf 242: }
1.108 parser 243: /// puts pieces to buf
1.144.2.28.2. (paf 244:): C serialize(size_t prolog_size) const;
1.108 parser 245: /// appends pieces from buf to self
1.144.2.28.2. 4(paf 246:3): bool deserialize(size_t prolog_size, void *buf, size_t buf_size);
1(paf 247:3): /// @see StringBody::append_help_length
248:3): String& append_help_length(const char* str, size_t helper_length, Language lang);
249:3): String& append_strdup(const char* str, size_t length, Language lang);
(paf 250:):
7(paf 251:3): bool operator == (const char* y) const { return body==StringBody(y); }
0(paf 252:3): bool operator != (const char* y) const { return body!=StringBody(y); }
(paf 253:):
254:): /// this starts with y
7(paf 255:3): bool starts_with(const char* y) const {
256:3): return body.ncmp(0/*x_begin*/, StringBody(y), 0/*y_begin*/, strlen(y))==0;
1.80 paf 257: }
1.144.2.28.2. 9(paf 258:3): /// x starts with this
259:3): bool this_starts(const char* x) const {
260:3): return StringBody(x).ncmp(0/*x_begin*/, body, 0/*y_begin*/, length())==0;
(paf 261:): }
1.80 paf 262:
1.144.2.28.2. 7(paf 263:3): String& append_to(String& dest, Language lang, bool forced) const;
264:3): String& append(const String& src, Language lang, bool forced=false) {
1.144.2.11 paf 265: return src.append_to(*this, lang, forced);
266: }
1.144.2.28.2. 7(paf 267:3): String& operator << (const String& src) { return append(src, L_PASS_APPENDED); }
1(paf 268:3): String& operator << (const char* src) { return append_help_length(src, 0, L_AS_IS); }
8(paf 269:3): String& operator << (const StringBody src) {
6(paf 270:3): body<<src;
9(paf 271:3): fragments+=Fragment(L_AS_IS, src.length());
272:3): return *this;
273:3): }
1.100 parser 274:
1.142 paf 275: /// extracts first char of a string, if any
276: char first_char() const {
1.144.2.28.2. 7(paf 277:3): return is_empty()?0:body.fetch(0);
1.142 paf 278: }
1.144.2.28.2. 8(paf 279:3):
280:3): bool operator < (const String& src) const { return body<src.body; }
281:3): bool operator > (const String& src) const { return body>src.body; }
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; }
1.54 paf 286:
287: /// extracts [start, finish) piece of string
1.144.2.28.2. 2(paf 288:3): String& mid(size_t substr_begin, size_t substr_end) const;
(paf 289:):
290:): /**
7(paf 291:3): ignore lang if it's L_UNSPECIFIED
(paf 292:): but when specified: look for substring that lies in ONE fragment in THAT lang
293:): @return position of substr in string, -1 means "not found" [const char* version]
294:): */
1(paf 295:3): size_t pos(const StringBody substr,
7(paf 296:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
297:3): /// String version of @see pos(const char*, int, Language)
(paf 298:): size_t pos(const String& substr,
7(paf 299:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
2(paf 300:3): size_t pos(char c,
301:3): size_t this_offset=0) const {
302:3): return body.pos(c, this_offset);
303:3): }
1.62 paf 304:
1.144.2.22 paf 305: void split(ArrayString& result,
1.144.2.28.2. 2(paf 306:3): size_t& pos_after,
(paf 307:): const char* delim,
2(paf 308:3): Language lang=L_UNSPECIFIED, int limit=-1) const;
1.144.2.22 paf 309: void split(ArrayString& result,
1.144.2.28.2. 5(paf 310:3): size_t& pos_after,
1.62 paf 311: const String& delim,
1.144.2.28.2. 7(paf 312:3): Language lang=L_UNSPECIFIED, int limit=-1) const;
1.62 paf 313:
1.144.2.28.2. 6(paf 314:3): typedef void (*Row_action)(Table& table, ArrayString* row,
1.136 paf 315: int prestart, int prefinish,
316: int poststart, int postfinish,
1.144.2.9 paf 317: void *info);
1.87 parser 318: /**
1.144.2.5 paf 319: @return table of found items, if any.
1.87 parser 320: table format is defined and fixed[can be used by others]:
321: @verbatim
322: prematch/match/postmatch/1/2/3/...
323: @endverbatim
324: */
1.144.2.28.2. (paf 325:): Table* match(Charset& source_charset,
1.64 paf 326: const String& regexp,
1.144.2.28.2. 5(paf 327:3): const String* options,
1.99 parser 328: Row_action row_action, void *info,
1.144.2.27 paf 329: bool& just_matched) const;
1.87 parser 330: enum Change_case_kind {
331: CC_UPPER,
332: CC_LOWER
333: };
1.144.2.28.2. (paf 334:): String& change_case(Charset& source_charset,
5(paf 335:3): Change_case_kind kind) const;
(paf 336:): const String& replace(const Dictionary& dict) const;
337:): double as_double() const;
338:): int as_int() const;
1.137 paf 339:
1.7 paf 340: private: //disabled
341:
1.12 paf 342: String& operator = (const String&) { return *this; }
1.7 paf 343:
1.1 paf 344: };
1.144.2.7 paf 345:
346: /// simple hash code of string. used by Hash
1.144.2.28.2. 8(paf 347:3): inline uint hash_code(const StringBody self) {
1.144.2.7 paf 348: return self.hash_code();
349: }
1.1 paf 350:
351: #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