thirdparty/libantlr3c-3.2/include/antlr3string.h

   1 /** \file
   2  * Simple string interface allows indiscriminate allocation of strings
   3  * such that they can be allocated all over the place and released in
   4  * one chunk via a string factory - saves lots of hassle in remembering what
   5  * strings were allocated where.
   6  */
   7 #ifndef _ANTLR3_STRING_H
   8 #define _ANTLR3_STRING_H
   9
  10 // [The "BSD licence"]
  11 // Copyright (c) 2005-2009 Jim Idle, Temporal Wave LLC
  12 // http://www.temporal-wave.com
  13 // http://www.linkedin.com/in/jimidle
  14 //
  15 // All rights reserved.
  16 //
  17 // Redistribution and use in source and binary forms, with or without
  18 // modification, are permitted provided that the following conditions
  19 // are met:
  20 // 1. Redistributions of source code must retain the above copyright
  21 //    notice, this list of conditions and the following disclaimer.
  22 // 2. Redistributions in binary form must reproduce the above copyright
  23 //    notice, this list of conditions and the following disclaimer in the
  24 //    documentation and/or other materials provided with the distribution.
  25 // 3. The name of the author may not be used to endorse or promote products
  26 //    derived from this software without specific prior written permission.
  27 //
  28 // THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
  29 // IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
  30 // OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
  31 // IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
  32 // INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
  33 // NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
  34 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
  35 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
  36 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
  37 // THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  38
  39 #include    <antlr3defs.h>
  40 #include    <antlr3collections.h>
  41
  42 #ifdef __cplusplus
  43 extern "C" {
  44 #endif
  45
  46 /** Base string class tracks the allocations and provides simple string
  47  *  tracking functions. Mostly you can work directly on the string for things
  48  *  that don't reallocate it, like strchr() etc. Perhaps someone will want to provide implementations for UTF8
  49  *  and so on.
  50  */
  51 typedef struct ANTLR3_STRING_struct
  52 {
  53
  54     /** The factory that created this string
  55      */
  56     pANTLR3_STRING_FACTORY      factory;
  57
  58     /** Pointer to the current string value (starts at NULL unless
  59      *  the string allocator is told to create it with a pre known size.
  60      */
  61     pANTLR3_UINT8               chars;
  62
  63     /** Current length of the string up to and not including, the trailing '\0'
  64      *  Note that the actual allocation (->size)
  65      *  is always at least one byte more than this to accommodate trailing '\0'
  66      */
  67     ANTLR3_UINT32               len;
  68
  69     /** Current size of the string in bytes including the trailing '\0'
  70      */
  71     ANTLR3_UINT32               size;
  72
  73     /** Index of string (allocation number) in case someone wants
  74      *  to explicitly release it.
  75      */
  76     ANTLR3_UINT32               index;
  77
  78     /** Occasionally it is useful to know what the encoding of the string
  79      *  actually is, hence it is stored here as one the ANTLR3_ENCODING_ values
  80      */
  81     ANTLR3_UINT8                encoding;
  82
  83     /** Pointer to function that sets the string value to a specific string in the default encoding
  84      *  for this string. For instance, if this is ASCII 8 bit, then this function is the same as set8
  85      *  but if the encoding is 16 bit, then the pointer is assumed to point to 16 bit characters not
  86      *  8 bit.
  87      */
  88     pANTLR3_UINT8   (*set)      (struct ANTLR3_STRING_struct * string, const char * chars);
  89
  90     /** Pointer to function that sets the string value to a specific 8 bit string in the default encoding
  91      *  for this string. For instance, if this is a 16 bit string, then this function is the same as set8
  92      *  but if the encoding is 16 bit, then the pointer is assumed to point to 8 bit characters that must
  93      *  be converted to 16 bit characters on the fly.
  94      */
  95     pANTLR3_UINT8   (*set8)     (struct ANTLR3_STRING_struct * string, const char * chars);
  96
  97     /** Pointer to function adds a raw char * type pointer in the default encoding
  98      *  for this string. For instance, if this is ASCII 8 bit, then this function is the same as append8
  99      *  but if the encoding is 16 bit, then the pointer is assumed to point to 16 bit characters not
 100      *  8 bit.
 101      */
 102     pANTLR3_UINT8   (*append)   (struct ANTLR3_STRING_struct * string, const char * newbit);
 103
 104     /** Pointer to function adds a raw char * type pointer in the default encoding
 105      *  for this string. For instance, if this is a 16 bit string, then this function assumes the pointer
 106      *  points to 8 bit characters that must be converted on the fly.
 107      */
 108     pANTLR3_UINT8   (*append8)  (struct ANTLR3_STRING_struct * string, const char * newbit);
 109
 110     /** Pointer to function that inserts the supplied string at the specified
 111      *  offset in the current string in the default encoding for this string. For instance, if this is an 8
 112      *  bit string, then this is the same as insert8, but if this is a 16 bit string, then the poitner
 113      *  must point to 16 bit characters.
 114      *
 115      */
 116     pANTLR3_UINT8   (*insert)   (struct ANTLR3_STRING_struct * string, ANTLR3_UINT32 point, const char * newbit);
 117
 118     /** Pointer to function that inserts the supplied string at the specified
 119      *  offset in the current string in the default encoding for this string. For instance, if this is a 16 bit string
 120      *  then the pointer is assumed to point at 8 bit characteres that must be converted on the fly.
 121      */
 122     pANTLR3_UINT8   (*insert8)  (struct ANTLR3_STRING_struct * string, ANTLR3_UINT32 point, const char * newbit);
 123
 124     /** Pointer to function that sets the string value to a copy of the supplied string (strings must be in the
 125      *  same encoding.
 126      */
 127     pANTLR3_UINT8   (*setS)     (struct ANTLR3_STRING_struct * string, struct ANTLR3_STRING_struct * chars);
 128
 129     /** Pointer to function appends a copy of the characters contained in another string. Strings must be in the
 130      *  same encoding.
 131      */
 132     pANTLR3_UINT8   (*appendS)  (struct ANTLR3_STRING_struct * string, struct ANTLR3_STRING_struct * newbit);
 133
 134     /** Pointer to function that inserts a copy of the characters in the supplied string at the specified
 135      *  offset in the current string. strings must be in the same encoding.
 136      */
 137     pANTLR3_UINT8   (*insertS)  (struct ANTLR3_STRING_struct * string, ANTLR3_UINT32 point, struct ANTLR3_STRING_struct * newbit);
 138
 139     /** Pointer to function that inserts the supplied integer in string form at the specified
 140      *  offset in the current string.
 141      */
 142     pANTLR3_UINT8   (*inserti)  (struct ANTLR3_STRING_struct * string, ANTLR3_UINT32 point, ANTLR3_INT32 i);
 143
 144     /** Pointer to function that adds a single character to the end of the string, in the encoding of the
 145      *  string - 8 bit, 16 bit, utf-8 etc. Input is a single UTF32 (32 bits wide integer) character.
 146      */
 147     pANTLR3_UINT8   (*addc)     (struct ANTLR3_STRING_struct * string, ANTLR3_UINT32 c);
 148
 149     /** Pointer to function that adds the stringified representation of an integer
 150      *  to the string.
 151      */
 152     pANTLR3_UINT8   (*addi)     (struct ANTLR3_STRING_struct * string, ANTLR3_INT32 i);
 153
 154     /** Pointer to function that compares the text of a string to the supplied
 155      *  8 bit character string and returns a result a la strcmp()
 156      */
 157     ANTLR3_UINT32   (*compare8) (struct ANTLR3_STRING_struct * string, const char * compStr);
 158
 159     /** Pointer to a function that compares the text of a string with the supplied character string
 160      *  (which is assumed to be in the same encoding as the string itself) and returns a result
 161      *  a la strcmp()
 162      */
 163     ANTLR3_UINT32   (*compare)  (struct ANTLR3_STRING_struct * string, const char * compStr);
 164
 165     /** Pointer to a function that compares the text of a string with the supplied string
 166      *  (which is assumed to be in the same encoding as the string itself) and returns a result
 167      *  a la strcmp()
 168      */
 169     ANTLR3_UINT32   (*compareS) (struct ANTLR3_STRING_struct * string, struct ANTLR3_STRING_struct * compStr);
 170
 171     /** Pointer to a function that returns the character indexed at the supplied
 172      *  offset as a 32 bit character.
 173      */
 174     ANTLR3_UCHAR    (*charAt)   (struct ANTLR3_STRING_struct * string, ANTLR3_UINT32 offset);
 175
 176     /** Pointer to a function that returns a substring of the supplied string a la .subString(s,e)
 177      *  in the Java language.
 178      */
 179     struct ANTLR3_STRING_struct *
 180                                         (*subString)    (struct ANTLR3_STRING_struct * string, ANTLR3_UINT32 startIndex, ANTLR3_UINT32 endIndex);
 181
 182     /** Pointer to a function that returns the integer representation of any numeric characters
 183      *  at the beginning of the string
 184      */
 185     ANTLR3_INT32        (*toInt32)          (struct ANTLR3_STRING_struct * string);
 186
 187     /** Pointer to a function that yields an 8 bit string regardless of the encoding of the supplied
 188      *  string. This is useful when you want to use the text of a token in some way that requires an 8 bit
 189      *  value, such as the key for a hashtable. The function is required to produce a usable string even
 190      *  if the text given as input has characters that do not fit in 8 bit space, it will replace them
 191      *  with some arbitrary character such as '?'
 192      */
 193     struct ANTLR3_STRING_struct *
 194                                         (*to8)      (struct ANTLR3_STRING_struct * string);
 195
 196         /// Pointer to a function that yields a UT8 encoded string of the current string,
 197         /// regardless of the current encoding of the string. Because there is currently no UTF8
 198         /// handling in the string class, it creates therefore, a string that is useful only for read only
 199         /// applications as it will not contain methods that deal with UTF8 at the moment.
 200         ///
 201         struct ANTLR3_STRING_struct *
 202                                         (*toUTF8)       (struct ANTLR3_STRING_struct * string);
 203
 204 }
 205     ANTLR3_STRING;
 206
 207 /** Definition of the string factory interface, which creates and tracks
 208  *  strings for you of various shapes and sizes.
 209  */
 210 typedef struct  ANTLR3_STRING_FACTORY_struct
 211 {
 212     /** List of all the strings that have been allocated by the factory
 213      */
 214     pANTLR3_VECTOR    strings;
 215
 216     /* Index of next string that we allocate
 217      */
 218     ANTLR3_UINT32   index;
 219
 220     /** Pointer to function that manufactures an empty string
 221      */
 222     pANTLR3_STRING  (*newRaw)   (struct ANTLR3_STRING_FACTORY_struct * factory);
 223
 224     /** Pointer to function that manufactures a raw string with no text in it but space for size
 225      *  characters.
 226      */
 227     pANTLR3_STRING  (*newSize)  (struct ANTLR3_STRING_FACTORY_struct * factory, ANTLR3_UINT32 size);
 228
 229     /** Pointer to function that manufactures a string from a given pointer and length. The pointer is assumed
 230      *  to point to characters in the same encoding as the string type, hence if this is a 16 bit string the
 231      *  pointer should point to 16 bit characters.
 232      */
 233     pANTLR3_STRING  (*newPtr)   (struct ANTLR3_STRING_FACTORY_struct * factory, pANTLR3_UINT8 string, ANTLR3_UINT32 size);
 234
 235     /** Pointer to function that manufactures a string from a given pointer and length. The pointer is assumed to
 236      *  point at 8 bit characters which must be converted on the fly to the encoding of the actual string.
 237      */
 238     pANTLR3_STRING  (*newPtr8)  (struct ANTLR3_STRING_FACTORY_struct * factory, pANTLR3_UINT8 string, ANTLR3_UINT32 size);
 239
 240     /** Pointer to function that manufactures a string from a given pointer and works out the length. The pointer is
 241      *  assumed to point to characters in the same encoding as the string itself, i.e. 16 bit if a 16 bit
 242      *  string and so on.
 243      */
 244     pANTLR3_STRING  (*newStr)   (struct ANTLR3_STRING_FACTORY_struct * factory, pANTLR3_UINT8 string);
 245
 246     /** Pointer to function that manufactures a string from a given pointer and length. The pointer should
 247      *  point to 8 bit characters regardless of the actual encoding of the string. The 8 bit characters
 248      *  will be converted to the actual string encoding on the fly.
 249      */
 250     pANTLR3_STRING  (*newStr8)  (struct ANTLR3_STRING_FACTORY_struct * factory, pANTLR3_UINT8 string);
 251
 252     /** Pointer to function that deletes the string altogether
 253      */
 254     void            (*destroy)  (struct ANTLR3_STRING_FACTORY_struct * factory, pANTLR3_STRING string);
 255
 256     /** Pointer to function that returns a copy of the string in printable form without any control
 257      *  characters in it.
 258      */
 259     pANTLR3_STRING  (*printable)(struct ANTLR3_STRING_FACTORY_struct * factory, pANTLR3_STRING string);
 260
 261     /** Pointer to function that closes the factory
 262      */
 263     void            (*close)    (struct ANTLR3_STRING_FACTORY_struct * factory);
 264
 265 }
 266     ANTLR3_STRING_FACTORY;
 267
 268 #ifdef __cplusplus
 269 }
 270 #endif
 271
 272 #endif
 273