Annotation of parser3/src/include/pa_string.h, revision 1.144.2.28.2.21
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: 2003/03/21 16:16:43 $";
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
19:): #include "cord.h"
20:): };
21:):
1.144.2.12 paf 22: // forwards
1.144.2.10 paf 23:
1.144.2.28 paf 24: class Charset;
1.144.2.28.2. (paf 25:): class Table;
1.144.2.5 paf 26: class SQL_Connection;
27: class Dictionary;
1.144.2.15 paf 28: class Request_charsets;
1.144.2.28.2. (paf 29:): class String;
7(paf 30:3): typedef Array<const String*> ArrayString;
1.144.2.5 paf 31:
32: // helpers
1.144.2.28.2. 0(paf 33:3): /*
(paf 34:): /// appends clean piece to String @see String::append
7(paf 35:3): #define APPEND_CLEAN(src, length) append(src, length, String::L_CLEAN)
(paf 36:): /// appends piece to String as-is @see String::append
7(paf 37:3): #define APPEND_AS_IS(src, length) append(src, length, String::L_AS_IS)
(paf 38:): /// appends tainted piece to String @see String::append
7(paf 39:3): #define APPEND_TAINTED(src, length) append(src, length, String::L_TAINTED)
(paf 40:): /// handy: appends const char* piece to String @see String::append
41:): #define APPEND_CONST(src) APPEND_AS_IS(src, 0)
0(paf 42:3): */
(paf 43:): /// this is result of pos functions which mean that substr were not found
44:): #define STRING_NOT_FOUND ((size_t)-1)
45:):
7(paf 46:3): class StringBody {
47:3):
48:3): CORD body;
49:3):
50:3): public:
51:3):
52:3): StringBody(): body(CORD_EMPTY) {}
53:3): StringBody(CORD abody): body(abody) {}
54:3):
55:3): void clear() { body=CORD_EMPTY; }
56:3):
57:3): uint hash_code() const;
58:3):
0(paf 59:3): const char* cstr() const { return CORD_to_const_char_star(body); }
1(paf 60:3): char* cstrm() const { return CORD_to_char_star(body); }
0(paf 61:3):
7(paf 62:3): size_t length() const { return CORD_len(body); }
63:3):
64:3): bool is_empty() const { return body==CORD_EMPTY; }
65:3):
66:3): void append(const char *str, size_t length) {
67:3): CORD_cat_char_star(body, str, length);
68:3): }
69:3): void append(char c) { CORD_cat_char(body, c); }
9(paf 70:3): void append(StringBody src) { CORD_cat(body, src.body); }
7(paf 71:3):
72:3): // could not figure out why this operator is needed [should do this chain: string->simple->==]
73:3): bool operator < (const StringBody& src) const { return CORD_cmp(body, src.body)<0; }
74:3): bool operator > (const StringBody& src) const { return CORD_cmp(body, src.body)>0; }
75:3): bool operator <= (const StringBody& src) const { return CORD_cmp(body, src.body)<=0; }
76:3): bool operator >= (const StringBody& src) const { return CORD_cmp(body, src.body)>=0; }
77:3): bool operator != (const StringBody& src) const { return CORD_cmp(body, src.body)!=0; }
78:3): bool operator == (const StringBody& src) const { return CORD_cmp(body, src.body)==0; }
79:3):
80:3): int ncmp(size_t x_begin, const StringBody y, size_t y_begin, size_t size) const {
81:3): return CORD_ncmp(body, x_begin, y.body, y_begin, size);
82:3): }
83:3):
84:3): char fetch(size_t index) const { return CORD_fetch(body, index); }
1(paf 85:3): StringBody mid(size_t index, size_t length) const { return CORD_substr(body, index, length); }
86:3): size_t pos(size_t offset, const char* substr) const { return CORD_str(body, offset, substr); }
87:3): size_t pos(size_t offset, const StringBody substr) const { return CORD_str(body, offset, substr.body); }
88:3):
89:3): template<typename I> void for_each(int (*callback)(const char* s, I), I info) const {
90:3): CORD_iter5(body, 0, 0, (CORD_batched_iter_fn)callback, info);
91:3): }
92:3):
93:3): void set_pos(CORD_pos& pos, size_t index) const { CORD_set_pos(pos, body, index); }
7(paf 94:3):
95:3): };
(paf 96:):
1.144.2.2 paf 97: /**
98: String which knows the language of all it's fragments.
1.41 paf 99:
100: All pieces remember
101: - whether they are tainted or not,
102: and the language which should be used to detaint them
103: */
1.144.2.28.2. (paf 104:): class String: public PA_Object {
105:):
7(paf 106:3): friend class StringBody;
107:3):
1.144.2.2 paf 108: public:
1.1 paf 109:
1.106 parser 110: /** piece is tainted or not. the language to use when detaint
111: remember to change String_Untaint_lang_name @ untaint.C along
112: */
1.144.2.28.2. 7(paf 113:3): enum Language {
114:3): L_UNSPECIFIED=0, ///< zero value handy for hash lookup @see untaint_lang_name2enum
115:3): // these two must go before others, there are checks for >L_AS_IS
116:3): L_CLEAN, ///< clean
117:3): L_AS_IS, ///< leave all characters intact
1.122 paf 118:
1.144.2.28.2. 7(paf 119:3): L_PASS_APPENDED,
1.41 paf 120: /**<
121: leave language built into string being appended.
122: just a flag, that value not stored
123: */
1.144.2.28.2. 7(paf 124:3): L_TAINTED, ///< tainted, untaint language as assigned later
1.122 paf 125: // untaint languages. assigned by ^untaint[lang]{...}
1.144.2.28.2. 7(paf 126:3): L_FILE_SPEC, ///< file specification
127:3): L_HTTP_HEADER, ///< text in HTTP response header
128:3): L_MAIL_HEADER, ///< text in mail header
129:3): L_URI, ///< text in uri
130:3): L_TABLE, ///< ^table:set body
131:3): L_SQL, ///< ^table:sql body
132:3): L_JS, ///< JavaScript code
133:3): L_XML, ///< ^dom:set xml
134:3): L_HTML, ///< HTML code (for editing)
135:3): L_OPTIMIZE_BIT = 0x8000 ///< flag, requiring cstr whitespace optimization
1.27 paf 136: };
137:
1.144.2.28.2. (paf 138:): struct Fragment {
7(paf 139:3): Language lang; ///< untaint flag, later untaint language
0(paf 140:3): size_t length; ///< length
7(paf 141:3): Fragment(Language alang, size_t asize): lang(alang), length(asize) {}
(paf 142:): };
143:):
144:): class ArrayFragment: public Array<Fragment> {
6(paf 145:3): void append(element_type src) {
(paf 146:): *static_cast<Array<Fragment> *>(this)+=src;
147:): }
6(paf 148:3): /// hiding from accidental USE, use append_positions
149:3): void append(const ArrayFragment& src, int offset, int limit) {
150:3): static_cast<Array<Fragment> *>(this)->append(src, offset, limit);
151:3): }
(paf 152:): public:
153:): ArrayFragment& operator += (element_type src) {
154:): if(count()) { // not empty?
155:): // try to join with last
156:): Fragment& last=get(count()-1);
157:): if(last.lang==src.lang) {
0(paf 158:3): last.length+=src.length;
(paf 159:): return *this;
160:): }
161:): }
6(paf 162:3): append(src);
(paf 163:): return *this;
164:): }
6(paf 165:3): void append(const ArrayFragment& src) { append(src, 0, 0); }
166:3): void append_positions(const ArrayFragment& src, size_t substr_begin, size_t substr_end);
(paf 167:): };
168:):
169:): struct C {
170:): char *str;
0(paf 171:3): size_t length;
(paf 172:): operator char *() { return str; }
0(paf 173:3): C(char *astr, size_t asize): str(astr), length(asize) {}
(paf 174:): };
175:):
176:): private:
177:):
7(paf 178:3): StringBody body; ///< all characters of string
0(paf 179:3): ArrayFragment fragments; ///< fragment language+length info
(paf 180:):
1.8 paf 181: public:
182:
1.144.2.28.2. 1(paf 183:3): /// @todo check all places for length to be only HELPER length, and never be !=strlen(cstr)
0(paf 184:3): explicit String(const char* cstr=0, size_t helper_length=0, bool tainted=false);
1.14 paf 185: String(const String& src);
1.144.2.28.2. 7(paf 186:3): String(StringBody abody, Language alang): body(abody) {
187:3): fragments+=Fragment(alang, abody.length());
188:3): }
189:3):
190:3): /// for convinient hash lookup
191:3): operator StringBody() const { return body; }
192:3):
193:3): bool is_empty() const { return body.is_empty(); }
194:3): size_t length() const { return body.length(); }
(paf 195:):
2(paf 196:3): /// convert to CORD. if 'lang' known, forcing 'lang' to it
1(paf 197:3): StringBody cstr_to_string_body(Language lang=L_AS_IS,
7(paf 198:3): SQL_Connection* connection=0,
(paf 199:): const Request_charsets *charsets=0) const;
200:):
2(paf 201:3): /// convert to constant C string. if 'lang' known, forcing 'lang' to it
7(paf 202:3): const char* cstr(Language lang=L_AS_IS,
203:3): SQL_Connection* connection=0,
1.144.2.14 paf 204: const Request_charsets *charsets=0) const {
1.144.2.28.2. 1(paf 205:3): return cstr_to_string_body(lang, connection, charsets).cstr();
(paf 206:): }
2(paf 207:3): /// convert to Modifiable C string. if 'lang' known, forcing 'lang' to it
7(paf 208:3): char *cstrm(Language lang=L_AS_IS,
209:3): SQL_Connection* connection=0,
(paf 210:): const Request_charsets *charsets=0) const {
1(paf 211:3): return cstr_to_string_body(lang, connection, charsets).cstrm();
1.144.2.9 paf 212: }
1.108 parser 213: /// puts pieces to buf
1.144.2.28.2. (paf 214:): C serialize(size_t prolog_size) const;
1.108 parser 215: /// appends pieces from buf to self
1.144.2.18 paf 216: bool deserialize(size_t prolog_size, void *buf, size_t buf_size, const char* file);
1.144.2.28.2. 6(paf 217:3): /** append fragment,
9(paf 218:3): @c length can be zero, it will be autocalced
6(paf 219:3): when length<strlen(str) makes heap copy of @c str piece
1.46 paf 220: */
1.144.2.28.2. 9(paf 221:3): String& append(const char* str, size_t length, Language lang);
(paf 222:):
7(paf 223:3): bool operator == (const char* y) const { return body==StringBody(y); }
0(paf 224:3): bool operator != (const char* y) const { return body!=StringBody(y); }
(paf 225:):
226:): /// this starts with y
7(paf 227:3): bool starts_with(const char* y) const {
228:3): return body.ncmp(0/*x_begin*/, StringBody(y), 0/*y_begin*/, strlen(y))==0;
1.80 paf 229: }
1.144.2.28.2. (paf 230:): /// y starts with this
7(paf 231:3): bool this_starts(const char* y) const {
232:3): return body.ncmp(0/*x_begin*/, StringBody(y), 0/*y_begin*/, strlen(y))==0;
(paf 233:): }
1.80 paf 234:
1.144.2.28.2. 7(paf 235:3): String& append_to(String& dest, Language lang, bool forced) const;
236:3): String& append(const String& src, Language lang, bool forced=false) {
1.144.2.11 paf 237: return src.append_to(*this, lang, forced);
238: }
1.144.2.28.2. 7(paf 239:3): String& operator << (const String& src) { return append(src, L_PASS_APPENDED); }
240:3): String& operator << (const char* src) { return append(src, 0, L_AS_IS); }
9(paf 241:3): String& operator << (const StringBody& src) {
242:3): body.append(src);
243:3): fragments+=Fragment(L_AS_IS, src.length());
244:3): return *this;
245:3): }
1.100 parser 246:
1.142 paf 247: /// extracts first char of a string, if any
248: char first_char() const {
1.144.2.28.2. 7(paf 249:3): return is_empty()?0:body.fetch(0);
1.142 paf 250: }
1.144.2.28.2. 8(paf 251:3):
252:3): bool operator < (const String& src) const { return body<src.body; }
253:3): bool operator > (const String& src) const { return body>src.body; }
254:3): bool operator <= (const String& src) const { return body<=src.body; }
255:3): bool operator >= (const String& src) const { return body>=src.body; }
256:3): bool operator != (const String& src) const { return body!=src.body; }
257:3): bool operator == (const String& src) const { return body==src.body; }
1.54 paf 258:
259: /// extracts [start, finish) piece of string
1.144.2.28.2. 4(paf 260:3): const String& mid(size_t substr_begin, size_t substr_end) const;
(paf 261:):
262:): /**
7(paf 263:3): ignore lang if it's L_UNSPECIFIED
(paf 264:): but when specified: look for substring that lies in ONE fragment in THAT lang
265:): @return position of substr in string, -1 means "not found" [const char* version]
266:): */
1(paf 267:3): size_t pos(const StringBody substr,
7(paf 268:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
269:3): /// String version of @see pos(const char*, int, Language)
(paf 270:): size_t pos(const String& substr,
7(paf 271:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
1.62 paf 272:
1.144.2.22 paf 273: void split(ArrayString& result,
1.144.2.28.2. 5(paf 274:3): size_t& pos_after,
(paf 275:): const char* delim,
7(paf 276:3): Language lang=L_UNSPECIFIED, int limit=-1);
1.144.2.22 paf 277: void split(ArrayString& result,
1.144.2.28.2. 5(paf 278:3): size_t& pos_after,
1.62 paf 279: const String& delim,
1.144.2.28.2. 7(paf 280:3): Language lang=L_UNSPECIFIED, int limit=-1) const;
1.62 paf 281:
1.144.2.28.2. (paf 282:): typedef void (*Row_action)(Table* table, ArrayString* row,
1.136 paf 283: int prestart, int prefinish,
284: int poststart, int postfinish,
1.144.2.9 paf 285: void *info);
1.87 parser 286: /**
1.144.2.5 paf 287: @return table of found items, if any.
1.87 parser 288: table format is defined and fixed[can be used by others]:
289: @verbatim
290: prematch/match/postmatch/1/2/3/...
291: @endverbatim
292: */
1.144.2.28.2. (paf 293:): Table* match(Charset& source_charset,
1.64 paf 294: const String& regexp,
1.144.2.28.2. (paf 295:): const String& options,
1.99 parser 296: Row_action row_action, void *info,
1.144.2.27 paf 297: bool& just_matched) const;
1.87 parser 298: enum Change_case_kind {
299: CC_UPPER,
300: CC_LOWER
301: };
1.144.2.28.2. (paf 302:): String& change_case(Charset& source_charset,
1.144.2.17 paf 303: Change_case_kind kind);
1.144.2.28.2. (paf 304:): const String& replace(const Dictionary& dict) const;
305:): double as_double() const;
306:): int as_int() const;
1.137 paf 307:
1.7 paf 308: private: //disabled
309:
1.12 paf 310: String& operator = (const String&) { return *this; }
1.7 paf 311:
1.1 paf 312: };
1.144.2.7 paf 313:
314: /// simple hash code of string. used by Hash
1.144.2.28.2. 7(paf 315:3): inline uint hash_code(const StringBody& self) {
1.144.2.7 paf 316: return self.hash_code();
317: }
1.1 paf 318:
319: #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