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 <uniconv.h>
  27 #include "libguile/__scm.h"
  28
  29 \f
  30
  31 /* String representation.
  32
  33    A string is a piece of a stringbuf.  A stringbuf can be used by
  34    more than one string.  When a string is written to and the
  35    stringbuf of that string is used by more than one string, a new
  36    stringbuf is created.  That is, strings are copy-on-write.  This
  37    behavior can be used to make the substring operation quite
  38    efficient.
  39
  40    The implementation is tuned so that mutating a string is costly,
  41    but just reading it is cheap and lock-free.
  42
  43    There are also mutation-sharing strings.  They refer to a part of
  44    an ordinary string.  Writing to a mutation-sharing string just
  45    writes to the ordinary string.
  46
  47
  48    Internal, low level interface to the character arrays
  49
  50    - Use scm_is_narrow_string to determine is the string is narrow or
  51      wide.
  52
  53    - Use scm_i_string_chars or scm_i_string_wide_chars to get a
  54      pointer to the byte or scm_t_wchar array of a string for reading.
  55      Use scm_i_string_length to get the number of characters in that
  56      array.  The array is not null-terminated.
  57
  58    - The array is valid as long as the corresponding SCM object is
  59      protected but only until the next SCM_TICK.  During such a 'safe
  60      point', strings might change their representation.
  61
  62    - Use scm_i_string_start_writing to get a version of the string
  63      ready for reading and writing.  This is a potentially costly
  64      operation since it implements the copy-on-write behavior.  When
  65      done with the writing, call scm_i_string_stop_writing.  You must
  66      do this before the next SCM_TICK.  (This means, before calling
  67      almost any other scm_ function and you can't allow throws, of
  68      course.)
  69
  70    - New strings can be created with scm_i_make_string or
  71      scm_i_make_wide_string.  This gives access to a writable pointer
  72      that remains valid as long as nobody else makes a copy-on-write
  73      substring of the string.  Do not call scm_i_string_stop_writing
  74      for this pointer.
  75
  76    - Alternately, scm_i_string_ref and scm_i_string_set_x can be used
  77      to read and write strings without worrying about whether the
  78      string is narrow or wide.  scm_i_string_set_x still needs to be
  79      bracketed by scm_i_string_start_writing and
  80      scm_i_string_stop_writing.
  81
  82    Legacy interface
  83
  84    - SCM_STRINGP is just scm_is_string.
  85
  86    - SCM_STRING_CHARS uses scm_i_string_writable_chars and immediately
  87      calls scm_i_stop_writing, hoping for the best.  SCM_STRING_LENGTH
  88      is the same as scm_i_string_length.  SCM_STRING_CHARS will throw
  89      an error for for strings that are not null-terminated.  There is
  90      no wide version of this interface.
  91 */
  92
  93 SCM_API SCM scm_string_p (SCM x);
  94 SCM_API SCM scm_string (SCM chrs);
  95 SCM_API SCM scm_make_string (SCM k, SCM chr);
  96 SCM_API SCM scm_string_length (SCM str);
  97 SCM_API SCM scm_string_width (SCM str);
  98 SCM_API SCM scm_string_ref (SCM str, SCM k);
  99 SCM_API SCM scm_string_set_x (SCM str, SCM k, SCM chr);
 100 SCM_API SCM scm_substring (SCM str, SCM start, SCM end);
 101 SCM_API SCM scm_substring_read_only (SCM str, SCM start, SCM end);
 102 SCM_API SCM scm_substring_shared (SCM str, SCM start, SCM end);
 103 SCM_API SCM scm_substring_copy (SCM str, SCM start, SCM end);
 104 SCM_API SCM scm_string_append (SCM args);
 105
 106 SCM_API SCM scm_c_make_string (size_t len, SCM chr);
 107 SCM_API size_t scm_c_string_length (SCM str);
 108 SCM_API size_t scm_c_symbol_length (SCM sym);
 109 SCM_API SCM scm_c_string_ref (SCM str, size_t pos);
 110 SCM_API void scm_c_string_set_x (SCM str, size_t pos, SCM chr);
 111 SCM_API SCM scm_c_substring (SCM str, size_t start, size_t end);
 112 SCM_API SCM scm_c_substring_read_only (SCM str, size_t start, size_t end);
 113 SCM_API SCM scm_c_substring_shared (SCM str, size_t start, size_t end);
 114 SCM_API SCM scm_c_substring_copy (SCM str, size_t start, size_t end);
 115
 116 SCM_API int scm_is_string (SCM x);
 117 SCM_API SCM scm_from_locale_string (const char *str);
 118 SCM_API SCM scm_from_locale_stringn (const char *str, size_t len);
 119 SCM_API SCM scm_take_locale_string (char *str);
 120 SCM_API SCM scm_take_locale_stringn (char *str, size_t len);
 121 SCM_API char *scm_to_locale_string (SCM str);
 122 SCM_API char *scm_to_locale_stringn (SCM str, size_t *lenp);
 123 SCM_INTERNAL char *scm_to_stringn (SCM str, size_t *lenp,
 124                                    const char *encoding,
 125                                    enum iconv_ilseq_handler handler);
 126 SCM_API size_t scm_to_locale_stringbuf (SCM str, char *buf, size_t max_len);
 127
 128 SCM_API SCM scm_makfromstrs (int argc, char **argv);
 129
 130 /* internal accessor functions.  Arguments must be valid. */
 131
 132 SCM_INTERNAL SCM scm_i_make_string (size_t len, char **datap);
 133 SCM_INTERNAL SCM scm_i_make_wide_string (size_t len, scm_t_wchar **datap);
 134 SCM_INTERNAL SCM scm_i_substring (SCM str, size_t start, size_t end);
 135 SCM_INTERNAL SCM scm_i_substring_read_only (SCM str, size_t start, size_t end);
 136 SCM_INTERNAL SCM scm_i_substring_shared (SCM str, size_t start, size_t end);
 137 SCM_INTERNAL SCM scm_i_substring_copy (SCM str, size_t start, size_t end);
 138 SCM_INTERNAL size_t scm_i_string_length (SCM str);
 139 SCM_API /* FIXME: not internal */ const char *scm_i_string_chars (SCM str);
 140 SCM_API const scm_t_wchar *scm_i_string_wide_chars (SCM str);
 141 SCM_API /* FIXME: not internal */ char *scm_i_string_writable_chars (SCM str);
 142 SCM_INTERNAL SCM scm_i_string_start_writing (SCM str);
 143 SCM_INTERNAL void scm_i_string_stop_writing (void);
 144 SCM_INTERNAL int scm_i_is_narrow_string (SCM str);
 145 SCM_INTERNAL scm_t_wchar scm_i_string_ref (SCM str, size_t x);
 146 SCM_INTERNAL void scm_i_string_set_x (SCM str, size_t p, scm_t_wchar chr);
 147 /* internal functions related to symbols. */
 148
 149 SCM_INTERNAL SCM scm_i_make_symbol (SCM name, scm_t_bits flags,
 150                                     unsigned long hash, SCM props);
 151 SCM_INTERNAL SCM
 152 scm_i_c_make_symbol (const char *name, size_t len,
 153                      scm_t_bits flags, unsigned long hash, SCM props);
 154 SCM_INTERNAL SCM
 155 scm_i_c_take_symbol (char *name, size_t len,
 156                      scm_t_bits flags, unsigned long hash, SCM props);
 157 SCM_INTERNAL const char *scm_i_symbol_chars (SCM sym);
 158 SCM_INTERNAL const scm_t_wchar *scm_i_symbol_wide_chars (SCM sym);
 159 SCM_INTERNAL size_t scm_i_symbol_length (SCM sym);
 160 SCM_INTERNAL int scm_i_is_narrow_symbol (SCM str);
 161 SCM_INTERNAL SCM scm_i_symbol_substring (SCM sym, size_t start, size_t end);
 162 SCM_INTERNAL scm_t_wchar scm_i_symbol_ref (SCM sym, size_t x);
 163
 164 /* internal GC functions. */
 165
 166 SCM_INTERNAL SCM scm_i_string_mark (SCM str);
 167 SCM_INTERNAL SCM scm_i_stringbuf_mark (SCM buf);
 168 SCM_INTERNAL SCM scm_i_symbol_mark (SCM buf);
 169 SCM_INTERNAL void scm_i_string_free (SCM str);
 170 SCM_INTERNAL void scm_i_stringbuf_free (SCM buf);
 171 SCM_INTERNAL void scm_i_symbol_free (SCM sym);
 172
 173 /* internal utility functions. */
 174
 175 SCM_INTERNAL char **scm_i_allocate_string_pointers (SCM list);
 176 SCM_INTERNAL void scm_i_free_string_pointers (char **pointers);
 177 SCM_INTERNAL void scm_i_get_substring_spec (size_t len,
 178                                             SCM start, size_t *cstart,
 179                                             SCM end, size_t *cend);
 180 SCM_INTERNAL SCM scm_i_take_stringbufn (char *str, size_t len);
 181
 182 /* deprecated stuff */
 183
 184 #if SCM_ENABLE_DEPRECATED
 185
 186 SCM_API int scm_i_deprecated_stringp (SCM obj);
 187 SCM_API char *scm_i_deprecated_string_chars (SCM str);
 188 SCM_API size_t scm_i_deprecated_string_length (SCM str);
 189
 190 #define SCM_STRINGP(x)       scm_i_deprecated_stringp(x)
 191 #define SCM_STRING_CHARS(x)  scm_i_deprecated_string_chars(x)
 192 #define SCM_STRING_LENGTH(x) scm_i_deprecated_string_length(x)
 193 #define SCM_STRING_UCHARS(str) ((unsigned char *)SCM_STRING_CHARS (str))
 194
 195 #endif
 196
 197 SCM_INTERNAL void scm_init_strings (void);
 198
 199 #endif  /* SCM_STRINGS_H */
 200
 201 /*
 202   Local Variables:
 203   c-file-style: "gnu"
 204   End:
 205 */