libguile/strings.h

   1 /* classes: h_files */
   2
   3 #ifndef SCM_STRINGS_H
   4 #define SCM_STRINGS_H
   5
   6 /* Copyright (C) 1995,1996,1997,1998,2000,2001, 2004, 2005, 2006, 2008 Free Software Foundation, Inc.
   7  *
   8  * This library is free software; you can redistribute it and/or
   9  * modify it under the terms of the GNU Lesser General Public License
  10  * as published by the Free Software Foundation; either version 3 of
  11  * the License, or (at your option) any later version.
  12  *
  13  * This library is distributed in the hope that it will be useful, but
  14  * WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  16  * Lesser General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU Lesser General Public
  19  * License along with this library; if not, write to the Free Software
  20  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
  21  * 02110-1301 USA
  22  */
  23
  24 \f
  25
  26 #include "libguile/__scm.h"
  27
  28 \f
  29
  30 /* String representation.
  31
  32    A string is a piece of a stringbuf.  A stringbuf can be used by
  33    more than one string.  When a string is written to and the
  34    stringbuf of that string is used by more than one string, a new
  35    stringbuf is created.  That is, strings are copy-on-write.  This
  36    behavior can be used to make the substring operation quite
  37    efficient.
  38
  39    The implementation is tuned so that mutating a string is costly,
  40    but just reading it is cheap and lock-free.
  41
  42    There are also mutation-sharing strings.  They refer to a part of
  43    an ordinary string.  Writing to a mutation-sharing string just
  44    writes to the ordinary string.
  45
  46
  47    Internal, low level interface to the character arrays
  48
  49    - Use scm_i_string_chars to get a pointer to the byte array of a
  50      string for reading.  Use scm_i_string_length to get the number of
  51      bytes in that array.  The array is not null-terminated.
  52
  53    - The array is valid as long as the corresponding SCM object is
  54      protected but only until the next SCM_TICK.  During such a 'safe
  55      point', strings might change their representation.
  56
  57    - Use scm_i_string_writable_chars to get the same pointer as with
  58      scm_i_string_chars, but for reading and writing.  This is a
  59      potentially costly operation since it implements the
  60      copy-on-write behavior.  When done with the writing, call
  61      scm_i_string_stop_writing.  You must do this before the next
  62      SCM_TICK.  (This means, before calling almost any other scm_
  63      function and you can't allow throws, of course.)
  64
  65    - New strings can be created with scm_i_make_string.  This gives
  66      access to a writable pointer that remains valid as long as nobody
  67      else makes a copy-on-write substring of the string.  Do not call
  68      scm_i_string_stop_writing for this pointer.
  69
  70    Legacy interface
  71
  72    - SCM_STRINGP is just scm_is_string.
  73
  74    - SCM_STRING_CHARS uses scm_i_string_writable_chars and immediately
  75      calls scm_i_stop_writing, hoping for the best.  SCM_STRING_LENGTH
  76      is the same as scm_i_string_length.  SCM_STRING_CHARS will throw
  77      an error for for strings that are not null-terminated.
  78 */
  79
  80 SCM_API SCM scm_string_p (SCM x);
  81 SCM_API SCM scm_string (SCM chrs);
  82 SCM_API SCM scm_make_string (SCM k, SCM chr);
  83 SCM_API SCM scm_string_length (SCM str);
  84 SCM_API SCM scm_string_ref (SCM str, SCM k);
  85 SCM_API SCM scm_string_set_x (SCM str, SCM k, SCM chr);
  86 SCM_API SCM scm_substring (SCM str, SCM start, SCM end);
  87 SCM_API SCM scm_substring_read_only (SCM str, SCM start, SCM end);
  88 SCM_API SCM scm_substring_shared (SCM str, SCM start, SCM end);
  89 SCM_API SCM scm_substring_copy (SCM str, SCM start, SCM end);
  90 SCM_API SCM scm_string_append (SCM args);
  91
  92 SCM_API SCM scm_c_make_string (size_t len, SCM chr);
  93 SCM_API size_t scm_c_string_length (SCM str);
  94 SCM_API size_t scm_c_symbol_length (SCM sym);
  95 SCM_API SCM scm_c_string_ref (SCM str, size_t pos);
  96 SCM_API void scm_c_string_set_x (SCM str, size_t pos, SCM chr);
  97 SCM_API SCM scm_c_substring (SCM str, size_t start, size_t end);
  98 SCM_API SCM scm_c_substring_read_only (SCM str, size_t start, size_t end);
  99 SCM_API SCM scm_c_substring_shared (SCM str, size_t start, size_t end);
 100 SCM_API SCM scm_c_substring_copy (SCM str, size_t start, size_t end);
 101
 102 SCM_API int scm_is_string (SCM x);
 103 SCM_API SCM scm_from_locale_string (const char *str);
 104 SCM_API SCM scm_from_locale_stringn (const char *str, size_t len);
 105 SCM_API SCM scm_take_locale_string (char *str);
 106 SCM_API SCM scm_take_locale_stringn (char *str, size_t len);
 107 SCM_API char *scm_to_locale_string (SCM str);
 108 SCM_API char *scm_to_locale_stringn (SCM str, size_t *lenp);
 109 SCM_API size_t scm_to_locale_stringbuf (SCM str, char *buf, size_t max_len);
 110
 111 SCM_API SCM scm_makfromstrs (int argc, char **argv);
 112
 113 /* internal accessor functions.  Arguments must be valid. */
 114
 115 SCM_INTERNAL SCM scm_i_make_string (size_t len, char **datap);
 116 SCM_INTERNAL SCM scm_i_substring (SCM str, size_t start, size_t end);
 117 SCM_INTERNAL SCM scm_i_substring_read_only (SCM str, size_t start, size_t end);
 118 SCM_INTERNAL SCM scm_i_substring_shared (SCM str, size_t start, size_t end);
 119 SCM_INTERNAL SCM scm_i_substring_copy (SCM str, size_t start, size_t end);
 120 SCM_INTERNAL size_t scm_i_string_length (SCM str);
 121 SCM_API /* FIXME: not internal */ const char *scm_i_string_chars (SCM str);
 122 SCM_API /* FIXME: not internal */ char *scm_i_string_writable_chars (SCM str);
 123 SCM_INTERNAL void scm_i_string_stop_writing (void);
 124
 125 /* internal functions related to symbols. */
 126
 127 SCM_INTERNAL SCM scm_i_make_symbol (SCM name, scm_t_bits flags,
 128                                     unsigned long hash, SCM props);
 129 SCM_INTERNAL SCM
 130 scm_i_c_make_symbol (const char *name, size_t len,
 131                      scm_t_bits flags, unsigned long hash, SCM props);
 132 SCM_INTERNAL SCM
 133 scm_i_c_take_symbol (char *name, size_t len,
 134                      scm_t_bits flags, unsigned long hash, SCM props);
 135 SCM_INTERNAL const char *scm_i_symbol_chars (SCM sym);
 136 SCM_INTERNAL size_t scm_i_symbol_length (SCM sym);
 137 SCM_INTERNAL SCM scm_i_symbol_substring (SCM sym, size_t start, size_t end);
 138
 139 /* internal GC functions. */
 140
 141 SCM_INTERNAL SCM scm_i_string_mark (SCM str);
 142 SCM_INTERNAL SCM scm_i_stringbuf_mark (SCM buf);
 143 SCM_INTERNAL SCM scm_i_symbol_mark (SCM buf);
 144 SCM_INTERNAL void scm_i_string_free (SCM str);
 145 SCM_INTERNAL void scm_i_stringbuf_free (SCM buf);
 146 SCM_INTERNAL void scm_i_symbol_free (SCM sym);
 147
 148 /* internal utility functions. */
 149
 150 SCM_INTERNAL char **scm_i_allocate_string_pointers (SCM list);
 151 SCM_INTERNAL void scm_i_free_string_pointers (char **pointers);
 152 SCM_INTERNAL void scm_i_get_substring_spec (size_t len,
 153                                             SCM start, size_t *cstart,
 154                                             SCM end, size_t *cend);
 155 SCM_INTERNAL SCM scm_i_take_stringbufn (char *str, size_t len);
 156
 157 /* deprecated stuff */
 158
 159 #if SCM_ENABLE_DEPRECATED
 160
 161 SCM_API int scm_i_deprecated_stringp (SCM obj);
 162 SCM_API char *scm_i_deprecated_string_chars (SCM str);
 163 SCM_API size_t scm_i_deprecated_string_length (SCM str);
 164
 165 #define SCM_STRINGP(x)       scm_i_deprecated_stringp(x)
 166 #define SCM_STRING_CHARS(x)  scm_i_deprecated_string_chars(x)
 167 #define SCM_STRING_LENGTH(x) scm_i_deprecated_string_length(x)
 168 #define SCM_STRING_UCHARS(str) ((unsigned char *)SCM_STRING_CHARS (str))
 169
 170 #endif
 171
 172 SCM_INTERNAL void scm_init_strings (void);
 173
 174 #endif  /* SCM_STRINGS_H */
 175
 176 /*
 177   Local Variables:
 178   c-file-style: "gnu"
 179   End:
 180 */