Diff for /parser3/src/include/pa_string.h between versions 1.141 and 1.144.2.22

version 1.141, 2002/08/01 11:41:16 version 1.144.2.22, 2003/02/04 15:59:05
Line 1 Line 1
 /** @file  /** @file
         Parser: string class decl.          Parser: string class decl.
   
         Copyright (c) 2001, 2002 ArtLebedev Group (http://www.artlebedev.com)          Copyright (c) 2001-2003 ArtLebedev Group (http://www.artlebedev.com)
         Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)          Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)
 */  */
   
Line 10 Line 10
   
 static const char* IDENT_STRING_H="$Date$";  static const char* IDENT_STRING_H="$Date$";
   
   // includes
   
 #include "pa_pool.h"  #include "pa_pool.h"
 #include "pa_types.h"  #include "pa_types.h"
   #include "pa_array.h"
   
   // forwards
   
   class Table; DECLARE_OBJECT_PTR(Table);
   class SQL_Connection;
   class Dictionary;
   class Request_charsets;
   typedef Array<StringPtr> ArrayString; DECLARE_OBJECT_PTR(ArrayString);
   
   // helpers
   
 #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,  \
                 uchar lang, \                  String_UL lang, \
                 const char *file, uint line                  const char* file, uint line
 /// appends piece to String  @see String::real_append  /// appends piece to String  @see String::real_append
 #       define APPEND(src, size, lang, file, line) \  #       define APPEND(src, size, lang, file, line) \
                 real_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, \
                 uchar lang                  String_UL lang
 /// appends piece to String  @see String::real_append  /// appends piece to String  @see String::real_append
 #       define APPEND(src, size, lang, file, line) \  #       define APPEND(src, size, lang, file, line) \
                 real_append(src, size, lang)                  real_append(src, size, lang)
Line 42  static const char* IDENT_STRING_H="$Date Line 55  static const char* IDENT_STRING_H="$Date
 /// handy: appends const char* piece to String  @see String::real_append  /// handy: appends const char* piece to String  @see String::real_append
 #define APPEND_CONST(src) APPEND_AS_IS(src, 0, 0, 0)  #define APPEND_CONST(src) APPEND_AS_IS(src, 0, 0, 0)
   
 class Table;  typedef uint String_UL;
 class Array;  
 class SQL_Connection;  
 class Dictionary;  
   
 /**   struct String_fragment { 
         Pooled string.          String_UL/*Untaint_lang*/ lang; ///< untaint flag, later untaint language
           const char* ptr;  ///< pointer to the start
           size_t size;  ///< length
   
   #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
   };
   
         Internal structure:  /** 
         @verbatim          String which knows the language of all it's fragments.
                 String                          Chunk0  
                 ======                          ========  
                 head--------------->[ptr, size, ...]  
                 append_here-------->[ptr, size, ...]  
                                                         .  
                                                         .  
                                                         [ptr, size, ...]  
                 link_row----------->[link to the next chunk]  
         @endverbatim  
   
         All pieces remember           All pieces remember 
         - the file and its line they are from [can be turned off by NO_STRING_ORIGIN]          - the file and its line they are from [can be turned off by NO_STRING_ORIGIN]
         - whether they are tainted or not,           - whether they are tainted or not, 
           and the language which should be used to detaint them            and the language which should be used to detaint them
 */  */
 #include "pa_pragma_pack_begin.h"  
 class String : public Pooled {  
 public:  
   
         enum {  class String: public Array<String_fragment> {
                 CR_PREALLOCATED_COUNT=2, ///< default preallocated item count          size_t fsize;
                 CR_GROW_COUNT=1 ///< each time the String chunk_is_full() string expanded()          int count() const; // hiding fro accidental use instead of size()
         };  public:
   
         /** piece is tainted or not. the language to use when detaint          /** piece is tainted or not. the language to use when detaint
                 remember to change String_Untaint_lang_name @ untaint.C along                  remember to change String_Untaint_lang_name @ untaint.C along
Line 101  public: Line 110  public:
                 UL_JS,        ///< JavaScript code                  UL_JS,        ///< JavaScript code
                 UL_XML,         ///< ^dom:set xml                  UL_XML,         ///< ^dom:set xml
                 UL_HTML,      ///< HTML code (for editing)                  UL_HTML,      ///< HTML code (for editing)
                 UL_OPTIMIZE_BIT = 0x80  ///< flag, requiring cstr whitespace optimization                  UL_OPTIMIZE_BIT = 0x8000  ///< flag, requiring cstr whitespace optimization
         };          };
   
 public:  public:
   
         static String& OnPool(Pool& apool, const char *local_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(Pool& apool, const char *src=0, size_t src_size=0, bool tainted=false);  
         String(const String& src);          String(const String& src);
         bool is_empty() const { return append_here==head.chunk.rows; }          bool is_empty() const { return !size(); }
         size_t size() const;          size_t size() const { return fsize; }
         /// convert to C string. if 'lang' known, forcing 'lang' to it          /// convert to C string. if 'lang' known, forcing 'lang' to it
         char *cstr(Untaint_lang lang=UL_AS_IS,           CharPtr cstr(Untaint_lang lang=UL_AS_IS, 
                 SQL_Connection *connection=0,                  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));                  CharPtr result(new char[cstr_bufsize(lang, connection, charsets)]);
                 char *eol=store_to(result, lang, connection, cstr_charset, cstr_charset_name);                  char *eol=store_to(result.get(), lang, connection, charsets);
                 *eol=0;                  *eol=0;
                 return result;                  return result;
         }          }
         char *cstr_debug_origins() const;          char *cstr(Pool& pool, Untaint_lang lang=UL_AS_IS, 
                   SQL_Connection *connection=0,
                   const Request_charsets *charsets=0) const {
                   char *result=new(pool) char[cstr_bufsize(lang, connection, charsets)];
                   char *eol=store_to(result, lang, connection, charsets);
                   *eol=0;
                   return result;
           }
           char *cstr_debug_origins(Pool& pool) const;
         /// puts pieces to buf          /// puts pieces to buf
         void serialize(size_t prolog_size,  void *& buf, size_t& buf_size) const;          void serialize(Pool& pool, size_t prolog_size,  char *& buf, size_t& buf_size) const;
         /// appends pieces from buf to self          /// 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          /** append fragment
                 @see APPEND_AS_IS, APPEND_CLEAN, APPEND_TAINTED, APPEND_CONST                  @see APPEND_AS_IS, APPEND_CLEAN, APPEND_TAINTED, APPEND_CONST
         */          */
Line 171  public: Line 187  public:
                 return cmp(partial, src_ptr, 0)!=0;                   return cmp(partial, src_ptr, 0)!=0; 
         }          }
   
         String& append(const String& src, uchar lang, bool forced=false);          String& append_to(String& dest, uchar lang, bool forced) const;
           String& append(const String& src, uchar 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 String& src) { return append(src, UL_PASS_APPENDED); }
         String& operator << (const char *src) { return APPEND_CONST(src); }          String& operator << (StringPtr 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;          uint hash_code() const;
   
         /// extracts first char of a string          /// extracts first char of a string, if any
         char first_char() const;          char first_char() const {
                   return is_empty()?0:*felements[0].ptr;
           }
   
         /// extracts [start, finish) piece of string          /// extracts [start, finish) piece of string
         String& mid(size_t start, size_t finish) const;          StringPtr mid(size_t start, size_t finish) const;
   
         /// @return position of substr in string, -1 means "not found" [String version]          /// @return position of substr in string, -1 means "not found" [String version]
         int pos(const String& substr,           int pos(const String& substr, 
                 int this_offset=0, Untaint_lang lang=UL_UNSPECIFIED) const;                  int this_offset=0, Untaint_lang lang=UL_UNSPECIFIED) const;
         /// @return position of substr in string, -1 means "not found" [const char* version]          /// @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;                  int this_offset=0, Untaint_lang lang=UL_UNSPECIFIED) const;
   
         void split(Array& result,           void split(ArrayString& result, 
                 size_t *pos_after_ref,                   size_t *pos_after_ref, 
                 const char *delim, size_t delim_size,                   const char* delim, size_t delim_size, 
                 Untaint_lang lang=UL_UNSPECIFIED, int limit=-1) const;                  Untaint_lang lang=UL_UNSPECIFIED, int limit=-1);
         void split(Array& result,           void split(ArrayString& result, 
                 size_t *pos_after_ref,                   size_t *pos_after_ref, 
                 const String& delim,                   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 prestart, int prefinish, 
                 int poststart, int postfinish,                  int poststart, int postfinish,
                 void *info);                  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]:                   table format is defined and fixed[can be used by others]: 
                 @verbatim                  @verbatim
                         prematch/match/postmatch/1/2/3/...                          prematch/match/postmatch/1/2/3/...
                 @endverbatim                  @endverbatim
         */          */
         bool match(          TablePtr match(Charset& source_charset,
                 const String *aorigin,                            StringPtr aorigin,              
                 const String& regexp,                   const String& regexp, 
                 const String *options,                  StringPtr options,
                 Table **table,  
                 Row_action row_action, void *info,                  Row_action row_action, void *info,
                 bool *was_global=0) const;                  bool *was_global=0) const;
         enum Change_case_kind {          enum Change_case_kind {
                 CC_UPPER,                  CC_UPPER,
                 CC_LOWER                  CC_LOWER
         };          };
         String& change_case(Pool& pool,          StringPtr change_case(Pool& pool, Charset& source_charset,
                 Change_case_kind kind) const;                  Change_case_kind kind);
         String& replace(Pool& pool, Dictionary& dict) const;          StringPtr replace(Pool& pool, Dictionary& dict) const;
         double as_double() const;          double as_double();
         int as_int() const;          int as_int();
   
         String& join_chains(Pool& pool, char** cstr) const;          StringPtr join_chains(Pool& pool) const;
   
 #ifndef NO_STRING_ORIGIN  #ifndef NO_STRING_ORIGIN
         /// origin of string. calculated by first row          /// origin of string. calculated by first row
         const Origin& origin() const;          const String_fragment::Origin& origin() const;
 #endif  #endif
   
 private:          size_t cstr_bufsize(Untaint_lang lang=UL_UNSPECIFIED,
                   SQL_Connection *connection=0,
         /** several String fragments                  const Request_charsets *charsets=0) const;
         */  
         struct Chunk {  
                 typedef uchar count_type;  
                 count_type count; ///< the number of rows in chunk  
                 // here could be some padding bytes  
                 /// string fragment or a link to next chunk union  
                 typedef union Row {  
                         typedef uchar item_size_type;  
                         /// fragment  
                         struct {   
                                 const char *ptr;  ///< pointer to the start  
                                 item_size_type size;  ///< length  
                                 uchar/*Untaint_lang*/ lang; ///< untaint flag, later untaint language  
 #ifndef NO_STRING_ORIGIN  
                                 Origin origin;  ///< origin  
 #endif  
                         } item;  
                         // we are using the fact that there's no padding before this field!  
                         Chunk *link;  ///< link to the next chunk in chain  
                 } rows_type[CR_PREALLOCATED_COUNT];  
                 rows_type rows;  
         };  
         /**  
                 'mutable' because can write after it's end, after it was appended to somebody   
                 @see String::append  
         */  
         mutable struct {  
                 Chunk chunk;  
                 Chunk *link_storage;  
         } head;  ///< the head chunk of the chunk chain  
   
         /// next append would write to this record  
         Chunk::Row *append_here;  
           
 private:  
         /// last chunk  
         mutable Chunk *last_chunk;  
   
 private:  
   
         bool chunk_is_full() {  
                 return append_here == last_chunk->rows+last_chunk->count;  
         }  
         uint used_rows() const;  
         void expand();  
           
         Untaint_lang lang_of(size_t offset) const;  
   
         size_t cstr_bufsize(Untaint_lang lang,  
                 SQL_Connection *connection,  
                 Charset *buf_charset) const;  
         /// convert to C string, store to 'dest' which must be big enough for proper untaint          /// 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,           char *store_to(char *dest, Untaint_lang lang=UL_UNSPECIFIED, 
                 SQL_Connection *connection=0,                  SQL_Connection *connection=0,
                 Charset *store_to_charset=0,                  const Request_charsets *charsets=0) const;
                 const char *store_to_charset_name=0) const;  /*
           void join_chain(
         void join_chain(Pool& pool,   
                                            const Chunk*& achunk, const Chunk::Row*& arow, uint& acountdown,                                              const Chunk*& achunk, const Chunk::Row*& arow, uint& acountdown, 
                                            uchar& joined_lang, const char *& joined_ptr, size_t& joined_size) const;                                             uchar& joined_lang, const char* & joined_ptr, size_t& joined_size) const;
   */
 private: //disabled  private: //disabled
   
         String& operator = (const String&) { return *this; }          String& operator = (const String&) { return *this; }
   
 };  };
 #include "pa_pragma_pack_end.h"  
   
 #define STRING_PREPARED_FOREACH_ROW(self, body) \  
         while(row!=(self).append_here) { \  /// simple hash code of string. used by Hash
                 if(countdown==0) { \  inline uint hash_code(const String& self) {
                         chunk=row->link; \          return self.hash_code();
                         row=chunk->rows; \  
                         countdown=chunk->count; \  
                 }; \  
                 { body } \  
                 row++; countdown--; \  
         }   
   
 #define STRING_PREFIX_FOREACH_ROW(self, body) { \  
         const Chunk *chunk=&(self).head.chunk;  \  
         const Chunk::Row *row=chunk->rows; \  
         uint countdown=chunk->count; \  
         STRING_PREPARED_FOREACH_ROW(self, body) \  
 }  }
   /// simple hash code of string. used by Hash
   inline uint hash_code(StringPtr self) {
           return self->hash_code();
   }
   
   // helpers
   
 #define STRING_FOREACH_ROW(body) STRING_PREFIX_FOREACH_ROW(*this, body)  #define STRING_PREPARED_FOREACH_FRAGMENT(code) \
 #define STRING_SRC_FOREACH_ROW(body) STRING_PREFIX_FOREACH_ROW(src, body)          for(; fragment<last; fragment++) { \
                   code; \
           }
   
   #define STRING_FOREACH_FRAGMENT(code) \
           { \
                   element_type *last = felements+fused; \
                   element_type *fragment = felements; \
                   STRING_PREPARED_FOREACH_FRAGMENT(code) \
           }
   
 #endif  #endif

Removed from v.1.141  
changed lines
  Added in v.1.144.2.22


E-mail: