glib/gscanner.h

   1 /* GLIB - Library of useful routines for C programming
   2  * Copyright (C) 1995-1997  Peter Mattis, Spencer Kimball and Josh MacDonald
   3  *
   4  * This library is free software; you can redistribute it and/or
   5  * modify it under the terms of the GNU Lesser General Public
   6  * License as published by the Free Software Foundation; either
   7  * version 2 of the License, or (at your option) any later version.
   8  *
   9  * This library is distributed in the hope that it will be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  12  * Lesser General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU Lesser General Public
  15  * License along with this library; if not, write to the
  16  * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  17  * Boston, MA 02111-1307, USA.
  18  */
  19
  20 /*
  21  * Modified by the GLib Team and others 1997-2000.  See the AUTHORS
  22  * file for a list of people on the GLib Team.  See the ChangeLog
  23  * files for a list of changes.  These files are distributed with
  24  * GLib at ftp://ftp.gtk.org/pub/gtk/.
  25  */
  26
  27 #ifndef __G_SCANNER_H__
  28 #define __G_SCANNER_H__
  29
  30 #include <glib/gdataset.h>
  31 #include <glib/ghash.h>
  32
  33 G_BEGIN_DECLS
  34
  35 typedef struct _GScanner        GScanner;
  36 typedef struct _GScannerConfig  GScannerConfig;
  37 typedef union  _GTokenValue     GTokenValue;
  38
  39 typedef void            (*GScannerMsgFunc)      (GScanner      *scanner,
  40                                                  gchar         *message,
  41                                                  gboolean       error);
  42
  43 /* GScanner: Flexible lexical scanner for general purpose.
  44  */
  45
  46 /* Character sets */
  47 #define G_CSET_A_2_Z    "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
  48 #define G_CSET_a_2_z    "abcdefghijklmnopqrstuvwxyz"
  49 #define G_CSET_DIGITS   "0123456789"
  50 #define G_CSET_LATINC   "\300\301\302\303\304\305\306"\
  51                         "\307\310\311\312\313\314\315\316\317\320"\
  52                         "\321\322\323\324\325\326"\
  53                         "\330\331\332\333\334\335\336"
  54 #define G_CSET_LATINS   "\337\340\341\342\343\344\345\346"\
  55                         "\347\350\351\352\353\354\355\356\357\360"\
  56                         "\361\362\363\364\365\366"\
  57                         "\370\371\372\373\374\375\376\377"
  58
  59 /* Error types */
  60 typedef enum
  61 {
  62   G_ERR_UNKNOWN,
  63   G_ERR_UNEXP_EOF,
  64   G_ERR_UNEXP_EOF_IN_STRING,
  65   G_ERR_UNEXP_EOF_IN_COMMENT,
  66   G_ERR_NON_DIGIT_IN_CONST,
  67   G_ERR_DIGIT_RADIX,
  68   G_ERR_FLOAT_RADIX,
  69   G_ERR_FLOAT_MALFORMED
  70 } GErrorType;
  71
  72 /* Token types */
  73 typedef enum
  74 {
  75   G_TOKEN_EOF                   =   0,
  76
  77   G_TOKEN_LEFT_PAREN            = '(',
  78   G_TOKEN_RIGHT_PAREN           = ')',
  79   G_TOKEN_LEFT_CURLY            = '{',
  80   G_TOKEN_RIGHT_CURLY           = '}',
  81   G_TOKEN_LEFT_BRACE            = '[',
  82   G_TOKEN_RIGHT_BRACE           = ']',
  83   G_TOKEN_EQUAL_SIGN            = '=',
  84   G_TOKEN_COMMA                 = ',',
  85
  86   G_TOKEN_NONE                  = 256,
  87
  88   G_TOKEN_ERROR,
  89
  90   G_TOKEN_CHAR,
  91   G_TOKEN_BINARY,
  92   G_TOKEN_OCTAL,
  93   G_TOKEN_INT,
  94   G_TOKEN_HEX,
  95   G_TOKEN_FLOAT,
  96   G_TOKEN_STRING,
  97
  98   G_TOKEN_SYMBOL,
  99   G_TOKEN_IDENTIFIER,
 100   G_TOKEN_IDENTIFIER_NULL,
 101
 102   G_TOKEN_COMMENT_SINGLE,
 103   G_TOKEN_COMMENT_MULTI,
 104   G_TOKEN_LAST
 105 } GTokenType;
 106
 107 union   _GTokenValue
 108 {
 109   gpointer      v_symbol;
 110   gchar         *v_identifier;
 111   gulong        v_binary;
 112   gulong        v_octal;
 113   gulong        v_int;
 114   guint64       v_int64;
 115   gdouble       v_float;
 116   gulong        v_hex;
 117   gchar         *v_string;
 118   gchar         *v_comment;
 119   guchar        v_char;
 120   guint         v_error;
 121 };
 122
 123 struct  _GScannerConfig
 124 {
 125   /* Character sets
 126    */
 127   gchar         *cset_skip_characters;          /* default: " \t\n" */
 128   gchar         *cset_identifier_first;
 129   gchar         *cset_identifier_nth;
 130   gchar         *cpair_comment_single;          /* default: "#\n" */
 131
 132   /* Should symbol lookup work case sensitive?
 133    */
 134   guint         case_sensitive : 1;
 135
 136   /* Boolean values to be adjusted "on the fly"
 137    * to configure scanning behaviour.
 138    */
 139   guint         skip_comment_multi : 1;         /* C like comment */
 140   guint         skip_comment_single : 1;        /* single line comment */
 141   guint         scan_comment_multi : 1;         /* scan multi line comments? */
 142   guint         scan_identifier : 1;
 143   guint         scan_identifier_1char : 1;
 144   guint         scan_identifier_NULL : 1;
 145   guint         scan_symbols : 1;
 146   guint         scan_binary : 1;
 147   guint         scan_octal : 1;
 148   guint         scan_float : 1;
 149   guint         scan_hex : 1;                   /* `0x0ff0' */
 150   guint         scan_hex_dollar : 1;            /* `$0ff0' */
 151   guint         scan_string_sq : 1;             /* string: 'anything' */
 152   guint         scan_string_dq : 1;             /* string: "\\-escapes!\n" */
 153   guint         numbers_2_int : 1;              /* bin, octal, hex => int */
 154   guint         int_2_float : 1;                /* int => G_TOKEN_FLOAT? */
 155   guint         identifier_2_string : 1;
 156   guint         char_2_token : 1;               /* return G_TOKEN_CHAR? */
 157   guint         symbol_2_token : 1;
 158   guint         scope_0_fallback : 1;           /* try scope 0 on lookups? */
 159   guint         store_int64 : 1;                /* use value.v_int64 rather than v_int */
 160   guint         padding_dummy;
 161 };
 162
 163 struct  _GScanner
 164 {
 165   /* unused fields */
 166   gpointer              user_data;
 167   guint                 max_parse_errors;
 168
 169   /* g_scanner_error() increments this field */
 170   guint                 parse_errors;
 171
 172   /* name of input stream, featured by the default message handler */
 173   const gchar           *input_name;
 174
 175   /* quarked data */
 176   GData                 *qdata;
 177
 178   /* link into the scanner configuration */
 179   GScannerConfig        *config;
 180
 181   /* fields filled in after g_scanner_get_next_token() */
 182   GTokenType            token;
 183   GTokenValue           value;
 184   guint                 line;
 185   guint                 position;
 186
 187   /* fields filled in after g_scanner_peek_next_token() */
 188   GTokenType            next_token;
 189   GTokenValue           next_value;
 190   guint                 next_line;
 191   guint                 next_position;
 192
 193   /* to be considered private */
 194   GHashTable            *symbol_table;
 195   gint                  input_fd;
 196   const gchar           *text;
 197   const gchar           *text_end;
 198   gchar                 *buffer;
 199   guint                 scope_id;
 200
 201   /* handler function for _warn and _error */
 202   GScannerMsgFunc       msg_handler;
 203 };
 204
 205 GScanner*       g_scanner_new                   (const GScannerConfig *config_templ);
 206 void            g_scanner_destroy               (GScanner       *scanner);
 207 void            g_scanner_input_file            (GScanner       *scanner,
 208                                                  gint           input_fd);
 209 void            g_scanner_sync_file_offset      (GScanner       *scanner);
 210 void            g_scanner_input_text            (GScanner       *scanner,
 211                                                  const  gchar   *text,
 212                                                  guint          text_len);
 213 GTokenType      g_scanner_get_next_token        (GScanner       *scanner);
 214 GTokenType      g_scanner_peek_next_token       (GScanner       *scanner);
 215 GTokenType      g_scanner_cur_token             (GScanner       *scanner);
 216 GTokenValue     g_scanner_cur_value             (GScanner       *scanner);
 217 guint           g_scanner_cur_line              (GScanner       *scanner);
 218 guint           g_scanner_cur_position          (GScanner       *scanner);
 219 gboolean        g_scanner_eof                   (GScanner       *scanner);
 220 guint           g_scanner_set_scope             (GScanner       *scanner,
 221                                                  guint           scope_id);
 222 void            g_scanner_scope_add_symbol      (GScanner       *scanner,
 223                                                  guint           scope_id,
 224                                                  const gchar    *symbol,
 225                                                  gpointer       value);
 226 void            g_scanner_scope_remove_symbol   (GScanner       *scanner,
 227                                                  guint           scope_id,
 228                                                  const gchar    *symbol);
 229 gpointer        g_scanner_scope_lookup_symbol   (GScanner       *scanner,
 230                                                  guint           scope_id,
 231                                                  const gchar    *symbol);
 232 void            g_scanner_scope_foreach_symbol  (GScanner       *scanner,
 233                                                  guint           scope_id,
 234                                                  GHFunc          func,
 235                                                  gpointer        user_data);
 236 gpointer        g_scanner_lookup_symbol         (GScanner       *scanner,
 237                                                  const gchar    *symbol);
 238 void            g_scanner_unexp_token           (GScanner       *scanner,
 239                                                  GTokenType     expected_token,
 240                                                  const gchar    *identifier_spec,
 241                                                  const gchar    *symbol_spec,
 242                                                  const gchar    *symbol_name,
 243                                                  const gchar    *message,
 244                                                  gint            is_error);
 245 void            g_scanner_error                 (GScanner       *scanner,
 246                                                  const gchar    *format,
 247                                                  ...) G_GNUC_PRINTF (2,3);
 248 void            g_scanner_warn                  (GScanner       *scanner,
 249                                                  const gchar    *format,
 250                                                  ...) G_GNUC_PRINTF (2,3);
 251
 252 #ifndef G_DISABLE_DEPRECATED
 253
 254 /* keep downward source compatibility */
 255 #define         g_scanner_add_symbol( scanner, symbol, value )  G_STMT_START { \
 256   g_scanner_scope_add_symbol ((scanner), 0, (symbol), (value)); \
 257 } G_STMT_END
 258 #define         g_scanner_remove_symbol( scanner, symbol )      G_STMT_START { \
 259   g_scanner_scope_remove_symbol ((scanner), 0, (symbol)); \
 260 } G_STMT_END
 261 #define         g_scanner_foreach_symbol( scanner, func, data ) G_STMT_START { \
 262   g_scanner_scope_foreach_symbol ((scanner), 0, (func), (data)); \
 263 } G_STMT_END
 264
 265 /* The following two functions are deprecated and will be removed in
 266  * the next major release. They do no good. */
 267 #define g_scanner_freeze_symbol_table(scanner) ((void)0)
 268 #define g_scanner_thaw_symbol_table(scanner) ((void)0)
 269
 270 #endif /* G_DISABLE_DEPRECATED */
 271
 272 G_END_DECLS
 273
 274 #endif /* __G_SCANNER_H__ */
 275