--- parser3/src/include/pa_string.h 2003/01/23 17:05:34 1.144.2.3 +++ parser3/src/include/pa_string.h 2003/03/18 16:00:07 1.144.2.28.2.2 @@ -1,32 +1,46 @@ /** @file Parser: string class decl. - Copyright (c) 2001, 2003 ArtLebedev Group (http://www.artlebedev.com) + Copyright (c) 2001-2003 ArtLebedev Group (http://www.artlebedev.com) Author: Alexandr Petrosian (http://paf.design.ru) */ #ifndef PA_STRING_H #define PA_STRING_H -static const char* IDENT_STRING_H="$Date: 2003/01/23 17:05:34 $"; +static const char* IDENT_STRING_H="$Date: 2003/03/18 16:00:07 $"; + +// includes -#include "pa_pool.h" #include "pa_types.h" #include "pa_array.h" +#include "cord.h" + +// forwards + +class Charset; +class Table; DECLARE_OBJECT_PTR(Table); +class SQL_Connection; +class Dictionary; +class Request_charsets; +typedef Array ArrayString; DECLARE_OBJECT_PTR(ArrayString); + +// helpers + #ifndef NO_STRING_ORIGIN # define STRING_APPEND_PARAMS \ - const char *src, size_t size, \ - uchar lang, \ - const char *file, uint line + const char* src, size_t size, \ + String_UL lang, \ + const char* file, uint line /// appends piece to String @see String::real_append # define APPEND(src, size, lang, file, line) \ real_append(src, size, lang, file, line) #else # define STRING_APPEND_PARAMS \ - const char *src, \ + const char* src, \ size_t size, \ - uchar lang + String_UL lang /// appends piece to String @see String::real_append # define APPEND(src, size, lang, file, line) \ real_append(src, size, lang) @@ -43,35 +57,19 @@ static const char* IDENT_STRING_H="$Date /// handy: appends const char* piece to String @see String::real_append #define APPEND_CONST(src) APPEND_AS_IS(src, 0, 0, 0) -class Table; -class SQL_Connection; -class Dictionary; - -struct String_fragment { - const char *ptr; ///< pointer to the start - size_t size; ///< length - int/*Untaint_lang*/ lang; ///< untaint flag, later untaint language +typedef uint String_UL; -#ifndef NO_STRING_ORIGIN - /// all String pieces hold information of where they come from - struct Origin { - const char *file; ///< macros file name | load file name | sql query text - uint line; ///< file line no | record no - } origin; ///< origin -#endif -}; /** String which knows the language of all it's fragments. All pieces remember - - the file and its line they are from [can be turned off by NO_STRING_ORIGIN] - whether they are tainted or not, and the language which should be used to detaint them */ -class String: public Array { - size_t fsize; +class String: public PA_Object { + public: /** piece is tainted or not. the language to use when detaint @@ -99,31 +97,40 @@ public: UL_JS, ///< JavaScript code UL_XML, ///< ^dom:set xml UL_HTML, ///< HTML code (for editing) - UL_OPTIMIZE_BIT = 0x80 ///< flag, requiring cstr whitespace optimization + UL_OPTIMIZE_BIT = 0x8000 ///< flag, requiring cstr whitespace optimization + }; + + struct Fragment_info { + String_UL/*Untaint_lang*/ lang; ///< untaint flag, later untaint language + size_t size; ///< length + Fragment_info(String_UL alang, size_t asize): lang(alang), size(asize) {} }; +private: + + CORD body; ///< all characters of string + Array fragment_info; ///< fragment language+size info + public: - //static String& OnPool(Pool& apool, const char *local_src=0, size_t src_size=0, bool tainted=false); - String(const char *src=0, size_t src_size=0, bool tainted=false); + explicit String(const char* src=0, size_t src_size=0, bool tainted=false); String(const String& src); - bool is_empty() const { return !size(); } - size_t size() const { return fsize; } + bool is_empty() const { return body==CORD_EMPTY; } + size_t size() const { return CORD_len(body); } /// convert to C string. if 'lang' known, forcing 'lang' to it - char *cstr(Untaint_lang lang=UL_AS_IS, + const char* cstr(Untaint_lang lang=UL_AS_IS, SQL_Connection *connection=0, - Charset *cstr_charset=0, const char *cstr_charset_name=0) const { + const Request_charsets *charsets=0) const { - char *result=(char *)malloc(cstr_bufsize(lang, connection, cstr_charset)); - char *eol=store_to(result, lang, connection, cstr_charset, cstr_charset_name); + char* result=new char[cstr_bufsize(lang, connection, charsets)]; + char *eol=store_to(result, lang, connection, charsets); *eol=0; return result; } - char *cstr_debug_origins() const; /// puts pieces to buf - void serialize(size_t prolog_size, void *& buf, size_t& buf_size) const; + void serialize(size_t prolog_size, char *& buf, size_t& buf_size) const; /// appends pieces from buf to self - bool deserialize(size_t prolog_size, void *buf, size_t buf_size, const char *file); + bool deserialize(size_t prolog_size, void *buf, size_t buf_size, const char* file); /** append fragment @see APPEND_AS_IS, APPEND_CLEAN, APPEND_TAINTED, APPEND_CONST */ @@ -169,88 +176,81 @@ public: return cmp(partial, src_ptr, 0)!=0; } - String& append(const String& src, uchar lang, bool forced=false); + String& append_to(String& dest, String_UL lang, bool forced) const; + String& append(const String& src, String_UL lang, bool forced=false) { + return src.append_to(*this, lang, forced); + } String& operator << (const String& src) { return append(src, UL_PASS_APPENDED); } - String& operator << (const char *src) { return APPEND_CONST(src); } + //String& operator << (const String* src) { return append(*src, UL_PASS_APPENDED); } + String& operator << (const char* src) { return APPEND_CONST(src); } - /// simple hash code of string. used by Hash + /// simple hash code of string uint hash_code() const; /// extracts first char of a string, if any char first_char() const { - return is_empty()?0:*felements[0].ptr; + return is_empty()?0:CORD_fetch(body, 0); } /// extracts [start, finish) piece of string - String& mid(size_t start, size_t finish) const; + const String& mid(size_t start, size_t finish) const; /// @return position of substr in string, -1 means "not found" [String version] int pos(const String& substr, int this_offset=0, Untaint_lang lang=UL_UNSPECIFIED) const; /// @return position of substr in string, -1 means "not found" [const char* version] - int pos(const char *substr, size_t substr_size=0, + int pos(const char* substr, size_t substr_size=0, int this_offset=0, Untaint_lang lang=UL_UNSPECIFIED) const; - void split(Array& result, + void split(ArrayString& result, size_t *pos_after_ref, - const char *delim, size_t delim_size, - Untaint_lang lang=UL_UNSPECIFIED, int limit=-1) const; - void split(Array& result, + const char* delim, size_t delim_size, + Untaint_lang lang=UL_UNSPECIFIED, int limit=-1); + void split(ArrayString& result, size_t *pos_after_ref, const String& delim, - Untaint_lang lang=UL_UNSPECIFIED, int limit=-1) const; + Untaint_lang lang=UL_UNSPECIFIED, int limit=-1); - typedef void (*Row_action)(Table& table, Array *row, + typedef void (*Row_action)(Table* table, ArrayString& row, int prestart, int prefinish, int poststart, int postfinish, void *info); /** - @return true if fills table. + @return table of found items, if any. table format is defined and fixed[can be used by others]: @verbatim prematch/match/postmatch/1/2/3/... @endverbatim */ - bool match(Charset& source_charset, - const String *aorigin, + Table* match(Charset& source_charset, + const String& aorigin, const String& regexp, - const String *options, - Table **table, + const String& options, Row_action row_action, void *info, - bool *was_global=0) const; + bool& just_matched) const; enum Change_case_kind { CC_UPPER, CC_LOWER }; - String& change_case(Charset& source_charset, - Change_case_kind kind) const; - String& replace(Dictionary& dict) const; - double as_double() const; - int as_int() const; + const String& change_case(Charset& source_charset, + Change_case_kind kind); + const String& replace(const Dictionary& dict) const; + double as_double(); + int as_int(); - String& join_chains(char** cstr) const; + const String& join_chains() const; -#ifndef NO_STRING_ORIGIN - /// origin of string. calculated by first row - const String_fragment::Origin& origin() const; -#endif - -private: - - Untaint_lang lang_of(size_t offset) const; - - size_t cstr_bufsize(Untaint_lang lang, - SQL_Connection *connection, - Charset *buf_charset) const; + size_t cstr_bufsize(Untaint_lang lang=UL_UNSPECIFIED, + SQL_Connection *connection=0, + const Request_charsets *charsets=0) const; /// convert to C string, store to 'dest' which must be big enough for proper untaint char *store_to(char *dest, Untaint_lang lang=UL_UNSPECIFIED, SQL_Connection *connection=0, - Charset *store_to_charset=0, - const char *store_to_charset_name=0) const; + const Request_charsets *charsets=0) const; /* void join_chain( const Chunk*& achunk, const Chunk::Row*& arow, uint& acountdown, - uchar& joined_lang, const char *& joined_ptr, size_t& joined_size) const; + String_UL& joined_lang, const char* & joined_ptr, size_t& joined_size) const; */ private: //disabled @@ -258,7 +258,24 @@ private: //disabled }; -typedef object_ptr StringPtr; -typedef object_ptr ConstStringPtr; + +/// simple hash code of string. used by Hash +inline uint hash_code(const String& self) { + return self.hash_code(); +} + +// helpers + +#define STRING_PREPARED_FOREACH_FRAGMENT(code) \ + for(; fragment