Annotation of parser3/src/include/pa_string.h, revision 1.144.2.28.2.17
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. 7(paf 11:3): static const char* IDENT_STRING_H="$Date: 2003/03/20 16:57:43 $";
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):
59:3): size_t length() const { return CORD_len(body); }
60:3):
61:3): bool is_empty() const { return body==CORD_EMPTY; }
62:3):
63:3): void append(const char *str, size_t length) {
64:3): CORD_cat_char_star(body, str, length);
65:3): }
66:3): void append(char c) { CORD_cat_char(body, c); }
67:3):
68:3): // could not figure out why this operator is needed [should do this chain: string->simple->==]
69:3): bool operator < (const StringBody& src) const { return CORD_cmp(body, src.body)<0; }
70:3): bool operator > (const StringBody& src) const { return CORD_cmp(body, src.body)>0; }
71:3): bool operator <= (const StringBody& src) const { return CORD_cmp(body, src.body)<=0; }
72:3): bool operator >= (const StringBody& src) const { return CORD_cmp(body, src.body)>=0; }
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):
76:3): // bool operator == (const char* y) const { return CORD_cmp(body, y)==0; }
77:3): // bool operator != (const char* y) const { return CORD_cmp(body, y)!=0; }
78:3):
79:3): int ncmp(size_t x_begin, const StringBody y, size_t y_begin, size_t size) const {
80:3): return CORD_ncmp(body, x_begin, y.body, y_begin, size);
81:3): }
82:3):
83:3): char fetch(size_t index) const { return CORD_fetch(body, index); }
84:3):
85:3): };
(paf 86:):
1.144.2.2 paf 87: /**
88: String which knows the language of all it's fragments.
1.41 paf 89:
90: All pieces remember
91: - whether they are tainted or not,
92: and the language which should be used to detaint them
93: */
1.144.2.28.2. (paf 94:): class String: public PA_Object {
95:):
7(paf 96:3): friend class StringBody;
97:3):
1.144.2.2 paf 98: public:
1.1 paf 99:
1.106 parser 100: /** piece is tainted or not. the language to use when detaint
101: remember to change String_Untaint_lang_name @ untaint.C along
102: */
1.144.2.28.2. 7(paf 103:3): enum Language {
104:3): L_UNSPECIFIED=0, ///< zero value handy for hash lookup @see untaint_lang_name2enum
105:3): // these two must go before others, there are checks for >L_AS_IS
106:3): L_CLEAN, ///< clean
107:3): L_AS_IS, ///< leave all characters intact
1.122 paf 108:
1.144.2.28.2. 7(paf 109:3): L_PASS_APPENDED,
1.41 paf 110: /**<
111: leave language built into string being appended.
112: just a flag, that value not stored
113: */
1.144.2.28.2. 7(paf 114:3): L_TAINTED, ///< tainted, untaint language as assigned later
1.122 paf 115: // untaint languages. assigned by ^untaint[lang]{...}
1.144.2.28.2. 7(paf 116:3): L_FILE_SPEC, ///< file specification
117:3): L_HTTP_HEADER, ///< text in HTTP response header
118:3): L_MAIL_HEADER, ///< text in mail header
119:3): L_URI, ///< text in uri
120:3): L_TABLE, ///< ^table:set body
121:3): L_SQL, ///< ^table:sql body
122:3): L_JS, ///< JavaScript code
123:3): L_XML, ///< ^dom:set xml
124:3): L_HTML, ///< HTML code (for editing)
125:3): L_OPTIMIZE_BIT = 0x8000 ///< flag, requiring cstr whitespace optimization
1.27 paf 126: };
127:
1.144.2.28.2. (paf 128:): struct Fragment {
7(paf 129:3): Language lang; ///< untaint flag, later untaint language
0(paf 130:3): size_t length; ///< length
7(paf 131:3): Fragment(Language alang, size_t asize): lang(alang), length(asize) {}
(paf 132:): };
133:):
134:): class ArrayFragment: public Array<Fragment> {
6(paf 135:3): void append(element_type src) {
(paf 136:): *static_cast<Array<Fragment> *>(this)+=src;
137:): }
6(paf 138:3): /// hiding from accidental USE, use append_positions
139:3): void append(const ArrayFragment& src, int offset, int limit) {
140:3): static_cast<Array<Fragment> *>(this)->append(src, offset, limit);
141:3): }
(paf 142:): public:
143:): ArrayFragment& operator += (element_type src) {
144:): if(count()) { // not empty?
145:): // try to join with last
146:): Fragment& last=get(count()-1);
147:): if(last.lang==src.lang) {
0(paf 148:3): last.length+=src.length;
(paf 149:): return *this;
150:): }
151:): }
6(paf 152:3): append(src);
(paf 153:): return *this;
154:): }
6(paf 155:3): void append(const ArrayFragment& src) { append(src, 0, 0); }
156:3): void append_positions(const ArrayFragment& src, size_t substr_begin, size_t substr_end);
(paf 157:): };
158:):
159:): struct C {
160:): char *str;
0(paf 161:3): size_t length;
(paf 162:): operator char *() { return str; }
0(paf 163:3): C(char *astr, size_t asize): str(astr), length(asize) {}
(paf 164:): };
165:):
166:): private:
167:):
7(paf 168:3): StringBody body; ///< all characters of string
0(paf 169:3): ArrayFragment fragments; ///< fragment language+length info
(paf 170:):
1.8 paf 171: public:
172:
1.144.2.28.2. 1(paf 173:3): /// @todo check all places for length to be only HELPER length, and never be !=strlen(cstr)
0(paf 174:3): explicit String(const char* cstr=0, size_t helper_length=0, bool tainted=false);
1.14 paf 175: String(const String& src);
1.144.2.28.2. 7(paf 176:3): String(StringBody abody, Language alang): body(abody) {
177:3): fragments+=Fragment(alang, abody.length());
178:3): }
179:3):
180:3): /// for convinient hash lookup
181:3): operator StringBody() const { return body; }
182:3):
183:3): bool is_empty() const { return body.is_empty(); }
184:3): size_t length() const { return body.length(); }
(paf 185:):
2(paf 186:3): /// convert to CORD. if 'lang' known, forcing 'lang' to it
7(paf 187:3): CORD cstr_to_cord(Language lang=L_AS_IS,
188:3): SQL_Connection* connection=0,
(paf 189:): const Request_charsets *charsets=0) const;
190:):
2(paf 191:3): /// convert to constant C string. if 'lang' known, forcing 'lang' to it
7(paf 192:3): const char* cstr(Language lang=L_AS_IS,
193:3): SQL_Connection* connection=0,
1.144.2.14 paf 194: const Request_charsets *charsets=0) const {
1.144.2.28.2. (paf 195:): return CORD_to_const_char_star(cstr_to_cord(lang, connection, charsets));
196:): }
2(paf 197:3): /// convert to Modifiable C string. if 'lang' known, forcing 'lang' to it
7(paf 198:3): char *cstrm(Language lang=L_AS_IS,
199:3): SQL_Connection* connection=0,
(paf 200:): const Request_charsets *charsets=0) const {
201:): return CORD_to_char_star(cstr_to_cord(lang, connection, charsets));
1.144.2.9 paf 202: }
1.108 parser 203: /// puts pieces to buf
1.144.2.28.2. (paf 204:): C serialize(size_t prolog_size) const;
1.108 parser 205: /// appends pieces from buf to self
1.144.2.18 paf 206: bool deserialize(size_t prolog_size, void *buf, size_t buf_size, const char* file);
1.144.2.28.2. 6(paf 207:3): /** append fragment,
208:3): @c helper_length can be zero, it will be autocalced
209:3): when length<strlen(str) makes heap copy of @c str piece
1.46 paf 210: */
1.144.2.28.2. 7(paf 211:3): String& append(const char* str, size_t helper_length, Language lang);
(paf 212:):
7(paf 213:3): bool operator == (const char* y) const { return body==StringBody(y); }
(paf 214:):
215:): /// this starts with y
7(paf 216:3): bool starts_with(const char* y) const {
217:3): return body.ncmp(0/*x_begin*/, StringBody(y), 0/*y_begin*/, strlen(y))==0;
1.80 paf 218: }
1.144.2.28.2. (paf 219:): /// y starts with this
7(paf 220:3): bool this_starts(const char* y) const {
221:3): return body.ncmp(0/*x_begin*/, StringBody(y), 0/*y_begin*/, strlen(y))==0;
(paf 222:): }
1.80 paf 223:
1.144.2.28.2. 7(paf 224:3): String& append_to(String& dest, Language lang, bool forced) const;
225:3): String& append(const String& src, Language lang, bool forced=false) {
1.144.2.11 paf 226: return src.append_to(*this, lang, forced);
227: }
1.144.2.28.2. 7(paf 228:3): String& operator << (const String& src) { return append(src, L_PASS_APPENDED); }
229:3): String& operator << (const char* src) { return append(src, 0, L_AS_IS); }
1.100 parser 230:
1.142 paf 231: /// extracts first char of a string, if any
232: char first_char() const {
1.144.2.28.2. 7(paf 233:3): return is_empty()?0:body.fetch(0);
1.142 paf 234: }
1.54 paf 235:
236: /// extracts [start, finish) piece of string
1.144.2.28.2. 4(paf 237:3): const String& mid(size_t substr_begin, size_t substr_end) const;
(paf 238:):
239:): /**
7(paf 240:3): ignore lang if it's L_UNSPECIFIED
(paf 241:): but when specified: look for substring that lies in ONE fragment in THAT lang
242:): @return position of substr in string, -1 means "not found" [const char* version]
243:): */
244:): size_t pos(CORD substr,
7(paf 245:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
246:3): /// String version of @see pos(const char*, int, Language)
(paf 247:): size_t pos(const String& substr,
7(paf 248:3): size_t this_offset=0, Language lang=L_UNSPECIFIED) const;
1.62 paf 249:
1.144.2.22 paf 250: void split(ArrayString& result,
1.144.2.28.2. 5(paf 251:3): size_t& pos_after,
(paf 252:): const char* delim,
7(paf 253:3): Language lang=L_UNSPECIFIED, int limit=-1);
1.144.2.22 paf 254: void split(ArrayString& result,
1.144.2.28.2. 5(paf 255:3): size_t& pos_after,
1.62 paf 256: const String& delim,
1.144.2.28.2. 7(paf 257:3): Language lang=L_UNSPECIFIED, int limit=-1) const;
1.62 paf 258:
1.144.2.28.2. (paf 259:): typedef void (*Row_action)(Table* table, ArrayString* row,
1.136 paf 260: int prestart, int prefinish,
261: int poststart, int postfinish,
1.144.2.9 paf 262: void *info);
1.87 parser 263: /**
1.144.2.5 paf 264: @return table of found items, if any.
1.87 parser 265: table format is defined and fixed[can be used by others]:
266: @verbatim
267: prematch/match/postmatch/1/2/3/...
268: @endverbatim
269: */
1.144.2.28.2. (paf 270:): Table* match(Charset& source_charset,
1.64 paf 271: const String& regexp,
1.144.2.28.2. (paf 272:): const String& options,
1.99 parser 273: Row_action row_action, void *info,
1.144.2.27 paf 274: bool& just_matched) const;
1.87 parser 275: enum Change_case_kind {
276: CC_UPPER,
277: CC_LOWER
278: };
1.144.2.28.2. (paf 279:): String& change_case(Charset& source_charset,
1.144.2.17 paf 280: Change_case_kind kind);
1.144.2.28.2. (paf 281:): const String& replace(const Dictionary& dict) const;
282:): double as_double() const;
283:): int as_int() const;
1.137 paf 284:
1.7 paf 285: private: //disabled
286:
1.12 paf 287: String& operator = (const String&) { return *this; }
1.7 paf 288:
1.1 paf 289: };
1.144.2.7 paf 290:
291: /// simple hash code of string. used by Hash
1.144.2.28.2. 7(paf 292:3): inline uint hash_code(const StringBody& self) {
1.144.2.7 paf 293: return self.hash_code();
294: }
1.1 paf 295:
296: #endif
E-mail: