|
|
| version 1.40, 2001/03/18 20:31:26 | version 1.62, 2001/04/03 05:23:40 |
|---|---|
| Line 1 | Line 1 |
| /* | /** @file |
| Parser | Parser: string class decl. |
| Copyright (c) 2001 ArtLebedev Group (http://www.artlebedev.com) | Copyright (c) 2001 ArtLebedev Group (http://www.artlebedev.com) |
| Author: Alexander Petrosyan <paf@design.ru> (http://design.ru/paf) | Author: Alexander Petrosyan <paf@design.ru> (http://design.ru/paf) |
| $Id$ | $Id$ |
| */ | */ |
| /* | |
| String Chunk0 | |
| ====== ======== | |
| head--------------->[ptr, size, ...] | |
| append_here-------->[ptr, size, ...] | |
| . | |
| . | |
| [ptr, size, ...] | |
| link_row----------->[link to the next chunk] | |
| */ | |
| #ifndef PA_STRING_H | #ifndef PA_STRING_H |
| #define PA_STRING_H | #define PA_STRING_H |
| #ifdef HAVE_CONFIG_H | #include "pa_config_includes.h" |
| # include "pa_config.h" | |
| #endif | |
| #include <stddef.h> | #include <stddef.h> |
| #include "pa_pool.h" | #include "pa_pool.h" |
| #include "pa_types.h" | #include "pa_types.h" |
| /** | |
| $MAIN:html-typo table elements must enlarge string not more that that | |
| that's a tradeoff - otherwise we'd have to scan string twice: | |
| - first for buffer length | |
| - second for replacements themselves | |
| */ | |
| #define UNTAINT_TIMES_BIGGER 10 | #define UNTAINT_TIMES_BIGGER 10 |
| #ifndef NO_STRING_ORIGIN | #ifndef NO_STRING_ORIGIN |
| # define STRING_APPEND_PARAMS \ | # define STRING_APPEND_PARAMS \ |
| const char *src, size_t size, \ | const char *src, size_t size, \ |
| bool tainted, \ | String::Untaint_lang lang, \ |
| const char *file, uint line | const char *file, uint line |
| # define APPEND(src, size, file, line) real_append(src, size, false, file, line) | /// appends piece to String @see String::real_append |
| # define APPEND_TAINTED(src, size, file, line) real_append(src, size, true, file, line) | # define APPEND(src, size, lang, file, line) \ |
| real_append(src, size, lang, file, line) | |
| #else | #else |
| # define STRING_APPEND_PARAMS \ | # define STRING_APPEND_PARAMS \ |
| const char *src, \ | const char *src, \ |
| size_t size, \ | size_t size, \ |
| bool tainted | String::Untaint_lang lang |
| # define APPEND(src, size, file, line) real_append(src, size, false) | /// appends piece to String @see String::real_append |
| # define APPEND_TAINTED(src, size, file, line) real_append(src, size, true) | # define APPEND(src, size, lang, file, line) \ |
| real_append(src, size, lang) | |
| #endif | #endif |
| #define APPEND_CONST(src) APPEND(src, 0, 0, 0) | /// appends clean piece to String @see String::real_append |
| #define APPEND_CLEAN(src, size, file, line) \ | |
| APPEND(src, size, String::UL_CLEAN, file, line) | |
| /// appends tainted piece to String @see String::real_append | |
| #define APPEND_TAINTED(src, size, file, line) \ | |
| APPEND(src, size, String::UL_TAINTED, file, line) | |
| /// handy: appends const char* piece to String @see String::real_append | |
| #define APPEND_CONST(src) APPEND_CLEAN(src, 0, 0, 0) | |
| class Array; | |
| /** | |
| Pooled string. | |
| Internal structure: | |
| @verbatim | |
| String Chunk0 | |
| ====== ======== | |
| head--------------->[ptr, size, ...] | |
| append_here-------->[ptr, size, ...] | |
| . | |
| . | |
| [ptr, size, ...] | |
| link_row----------->[link to the next chunk] | |
| @endverbatim | |
| 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 Pooled { | class String : public Pooled { |
| public: | public: |
| enum { | enum { |
| CR_PREALLOCATED_COUNT=5, | CR_PREALLOCATED_COUNT=5, ///< default preallocated item count |
| CR_GROW_PERCENT=60 | CR_GROW_PERCENT=60 ///< each time the Array chunk_is_full() array expanded() |
| }; | }; |
| /// piece is tainted or not. the language to use when detaint | |
| enum Untaint_lang { | enum Untaint_lang { |
| UNKNOWN=0, // when get by name fails | UL_UNKNOWN=0, ///< zero value handy for hash lookup @see untaint_lang_name2enum |
| NO, // clean | UL_CLEAN, ///< clean |
| YES, // tainted, untaint language as assigned later | UL_TAINTED, ///< tainted, untaint language as assigned later |
| // untaint languages. assigned by ^untaint[lang]{...} | // untaint languages. assigned by ^untaint[lang]{...} |
| PASS_APPENDED, | UL_PASS_APPENDED, |
| // leave language built into string being appended | /**< |
| // just a flag, that value not stored | leave language built into string being appended. |
| AS_IS, | just a flag, that value not stored |
| HEADER, | */ |
| URI, | UL_AS_IS, ///< leave all characters intact |
| TABLE, | UL_FILE_NAME, ///< filename |
| SQL, | UL_HEADER, ///< text in response header |
| JS, | UL_URI, ///< text in uri |
| HTML, | UL_TABLE, ///< ^table:set body |
| HTML_TYPO | UL_SQL, ///< ^table:sql body |
| UL_JS, ///< JavaScript code | |
| UL_HTML, ///< HTML code (for editing) | |
| UL_HTML_TYPO ///< HTML code with TYPOgraphic replacements (for showing) | |
| }; | }; |
| public: | public: |
| Line 80 public: | Line 109 public: |
| String(Pool& apool, const char *src=0, bool tainted=false); | String(Pool& apool, const char *src=0, bool tainted=false); |
| String(const String& src); | String(const String& src); |
| size_t size() const { return fsize; } | size_t size() const { return fsize; } |
| int used_rows() const { return fused_rows; } | /// convert to C string |
| char *cstr() const; | char *cstr() const { |
| char *result=(char *)malloc(size()*UNTAINT_TIMES_BIGGER+1); | |
| char *eol=store_to(result); | |
| *eol=0; | |
| return result; | |
| } | |
| /** append fragment | |
| @see APPEND_CLEAN, APPEND_TAINTED, APPEND_CONST | |
| */ | |
| String& real_append(STRING_APPEND_PARAMS); | String& real_append(STRING_APPEND_PARAMS); |
| int cmp (const String& src) const; | /// @return <0 ==0 or >0 depending on comparison result |
| bool operator < (const String& src) const { return cmp(src)<0; } | int cmp (int& partial, const String& src, |
| bool operator > (const String& src) const { return cmp(src)>0; } | size_t this_offset=0, Untaint_lang lang=UL_UNKNOWN) const; |
| bool operator <= (const String& src) const { return cmp(src)<=0; } | bool operator < (const String& src) const { int p; return cmp(p, src)<0; } |
| bool operator >= (const String& src) const { return cmp(src)>=0; } | bool operator > (const String& src) const { int p; return cmp(p, src)>0; } |
| bool operator <= (const String& src) const { int p; return cmp(p, src)<=0; } | |
| bool operator >= (const String& src) const { int p; return cmp(p, src)>=0; } | |
| bool operator == (const String& src) const { | bool operator == (const String& src) const { |
| if(size()!=src.size()) // can speed up in trivial case | if(size()!=src.size()) // can speed up in trivial case |
| return false; | return false; |
| return cmp(src)==0; | int p; return cmp(p, src)==0; |
| } | |
| bool operator != (const String& src) const { int p; return cmp(p, src)!=0; } | |
| /** | |
| @param partial | |
| returns partial match status. | |
| - -1: strings too different | |
| - 0: full match | |
| - 1: means @c this starts @c src | |
| - 2: means @src starts @this | |
| */ | |
| int cmp(int& partial, const char* src_ptr, size_t src_size=0, | |
| size_t this_offset=0, Untaint_lang lang=UL_UNKNOWN) const; | |
| bool operator == (const char* src_ptr) const { | |
| size_t src_size=src_ptr?strlen(src_ptr):0; | |
| if(size() != src_size) | |
| return false; | |
| int partial; // unused | |
| return cmp(partial, src_ptr, src_size)==0; | |
| } | } |
| bool operator != (const String& src) const { return cmp(src)!=0; } | |
| bool operator == (const char* b_ptr) const; | /** |
| appends other String. | |
| marking all tainted pieces of it with @a lang. | |
| or marking ALL pieces of it with a @a lang when @a forced to. | |
| */ | |
| String& append(const String& src, Untaint_lang lang, bool forced=false); | String& append(const String& src, Untaint_lang lang, bool forced=false); |
| /// simple hash code of string. used by Hash | |
| uint hash_code() const; | uint hash_code() const; |
| const Origin& origin() const { return head.rows[0].item.origin; } | /// extracts [start, finish) piece of string |
| String& piece(size_t start, size_t finish) const; | |
| /// @return position of substr in string, -1 means "not found" [String version] | |
| int pos(const String& substr, | |
| size_t this_offset=0, Untaint_lang lang=UL_UNKNOWN) const; | |
| /// @return position of substr in string, -1 means "not found" [const char* version] | |
| int pos(const char *substr, size_t substr_size, | |
| size_t this_offset=0, Untaint_lang lang=UL_UNKNOWN) const; | |
| void split(Array& result, | |
| size_t *pos_after_ref, | |
| const char *delim, size_t delim_size, | |
| Untaint_lang lang, int limit=-1) const; | |
| void split(Array& result, | |
| size_t *pos_after_ref, | |
| const String& delim, | |
| Untaint_lang lang, int limit=-1) const; | |
| // void change_lang(Untaint_lang lang); | #ifndef NO_STRING_ORIGIN |
| /// origin of string. calculated by first row | |
| const Origin& origin() const; | |
| #endif | |
| private: | private: |
| struct Chunk { | struct Chunk { |
| // the number of rows in chunk | // the number of rows in chunk |
| int count; | size_t count; |
| union Row { | union Row { |
| // fragment | // fragment |
| struct { | struct { |
| Line 149 private: | Line 233 private: |
| return append_here == link_row; | return append_here == link_row; |
| } | } |
| void expand(); | void expand(); |
| void set_lang(Chunk::Row *row, Untaint_lang lang, bool forced, size_t size); | char *String::store_to(char *dest) const; |
| private: //disabled | private: //disabled |