--- parser3/src/include/pa_string.h 2004/04/06 07:53:57 1.159 +++ parser3/src/include/pa_string.h 2008/07/21 07:37:02 1.174 @@ -1,17 +1,16 @@ /** @file Parser: string class decl. - Copyright (c) 2001-2004 ArtLebedev Group (http://www.artlebedev.com) + Copyright (c) 2001-2005 ArtLebedev Group (http://www.artlebedev.com) Author: Alexandr Petrosian (http://paf.design.ru) */ #ifndef PA_STRING_H #define PA_STRING_H -static const char * const IDENT_STRING_H="$Date: 2004/04/06 07:53:57 $"; +static const char * const IDENT_STRING_H="$Date: 2008/07/21 07:37:02 $"; // includes - #include "pa_types.h" #include "pa_array.h" @@ -20,6 +19,8 @@ extern "C" { // cord's author forgot to #include "cord.h" }; +// defines + // cord extension /* Returns true if x does contain */ /* char not_c at positions i..i+n. Value i,i+n must be < CORD_len(x). */ @@ -76,25 +77,26 @@ public: enum Language { L_UNSPECIFIED=0, ///< no real string has parts of this lange: it's just convinient to check when string's empty // these two must go before others, there are checks for >L_AS_IS - L_CLEAN='0', ///< clean WARNING: read above warning before changing - L_AS_IS='A', ///< leave all characters intact WARNING: read above warning before changing + L_CLEAN='0', ///< clean WARNING: read above warning before changing + L_AS_IS='A', ///< leave all characters intact WARNING: read above warning before changing L_PASS_APPENDED='P', /**< leave lang built into string being appended. just a flag, that value not stored */ - L_TAINTED='T', ///< tainted, untaint lang as assigned later + L_TAINTED='T', ///< tainted, untaint lang as assigned later // untaint langs. assigned by ^untaint[lang]{...} - L_FILE_SPEC='F', ///< file specification - L_HTTP_HEADER='h', ///< text in HTTP response header - L_MAIL_HEADER='m', ///< text in mail header - L_URI='U', ///< text in uri - L_TABLE='L', ///< ^table:set body - L_SQL='Q', ///< ^table:sql body - L_JS='J', ///< JavaScript code + L_FILE_SPEC='F', ///< file specification + L_HTTP_HEADER='h', ///< text in HTTP response header + L_MAIL_HEADER='m', ///< text in mail header + L_URI='U', ///< text in uri + L_SQL='Q', ///< ^table:sql body + L_JS='J', ///< JavaScript code L_XML='X', ///< ^dom:set xml - L_HTML='H', ///< HTML code (for editing) + L_HTML='H', ///< HTML code + L_REGEX='R', ///< RegEx expression + L_HTTP_COOKIE='C', ///< cookies encoded as %uXXXX for compartibility with js functions encode/decode // READ WARNING ABOVE BEFORE ADDING ANYTHING L_OPTIMIZE_BIT = 0x80 ///< flag, requiring cstr whitespace optimization }; @@ -108,8 +110,15 @@ public: union Languages { struct { +#ifdef PA_LITTLE_ENDIAN Language lang:8; int is_not_just_lang:sizeof(CORD)*8-8; +#elif defined(PA_BIG_ENDIAN) + int is_not_just_lang:sizeof(CORD)*8-8; + Language lang:8; +#else +# error word endianness not determined for some obscure reason +#endif } opt; CORD langs; @@ -146,6 +155,7 @@ public: public: const char* v() const; + void dump() const; Languages(): langs(0) {} Languages(Language alang) { @@ -211,7 +221,7 @@ public: if(opt.is_not_just_lang) return CORD_range_contains_chr_greater_then(langs, aoffset, alength, (unsigned)alang)==0; else - return opt.lang<=alang; + return (unsigned)opt.lang<=(unsigned)alang; } /// @returns count of blocks @@ -234,7 +244,7 @@ public: callback(opt.lang, get_length(current), info); } - bool invariant(size_t current_length) { + bool invariant(size_t current_length) const { if(!langs) return current_length==0; if(opt.is_not_just_lang) @@ -250,6 +260,7 @@ public: public: const char* v() const; + void dump() const; Body(): body(CORD_EMPTY) {} Body(CORD abody): body(abody) { @@ -367,9 +378,9 @@ private: Languages langs; ///< string characters lang const char* v() const; - -#define ASSERT_STRING_INVARIANT(string) \ - assert((string).langs.invariant((string).body.length())) + void dump() const; + #define ASSERT_STRING_INVARIANT(string) \ + assert((string).langs.invariant((string).body.length())) public: @@ -389,6 +400,7 @@ public: bool is_empty() const { return body.is_empty(); } size_t length() const { return body.length(); } + size_t length(Charset& charset) const; /// convert to CORD. if 'lang' known, forcing 'lang' to it Body cstr_to_string_body(Language lang=L_AS_IS, @@ -451,6 +463,7 @@ public: /// extracts [start, finish) piece of string String& mid(size_t substr_begin, size_t substr_end) const; + String& mid(Charset& charset, size_t from, size_t to, size_t helper_length=0) const; /** ignore lang if it's L_UNSPECIFIED @@ -466,6 +479,9 @@ public: size_t this_offset=0) const { return body.pos(c, this_offset); } + size_t pos(Charset& charset, + const String& substr, + size_t this_offset=0, Language lang=L_UNSPECIFIED) const; void split(ArrayString& result, size_t& pos_after, @@ -491,7 +507,7 @@ public: const String& regexp, const String* options, Row_action row_action, void *info, - bool& just_matched) const; + int& matches_count) const; enum Change_case_kind { CC_UPPER, CC_LOWER @@ -502,6 +518,8 @@ public: const String& trim(Trim_kind kind=TRIM_BOTH, const char* chars=0) const; double as_double() const { return pa_atod(cstr(), this); } int as_int() const { return pa_atoi(cstr(), this); } + bool as_bool() const { return as_int()!=0; } + const String& escape(Charset& source_charset) const; private: //disabled