tizen: kdbus: make i explicitly positive in make_single_header_vector
[platform/upstream/glib.git] / glib / gscanner.c
1 /* GLIB - Library of useful routines for C programming
2  * Copyright (C) 1995-1997  Peter Mattis, Spencer Kimball and Josh MacDonald
3  *
4  * GScanner: Flexible lexical scanner for general purpose.
5  * Copyright (C) 1997, 1998 Tim Janik
6  *
7  * SPDX-License-Identifier: LGPL-2.1-or-later
8  *
9  * This library is free software; you can redistribute it and/or
10  * modify it under the terms of the GNU Lesser General Public
11  * License as published by the Free Software Foundation; either
12  * version 2.1 of the License, or (at your option) any later version.
13  *
14  * This library is distributed in the hope that it will be useful,
15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
17  * Lesser General Public License for more details.
18  *
19  * You should have received a copy of the GNU Lesser General Public
20  * License along with this library; if not, see <http://www.gnu.org/licenses/>.
21  */
22
23 /*
24  * Modified by the GLib Team and others 1997-2000.  See the AUTHORS
25  * file for a list of people on the GLib Team.  See the ChangeLog
26  * files for a list of changes.  These files are distributed with
27  * GLib at ftp://ftp.gtk.org/pub/gtk/.
28  */
29
30 /*
31  * MT safe
32  */
33
34 #include "config.h"
35
36 #include <errno.h>
37 #include <stdlib.h>
38 #include <stdarg.h>
39 #include <string.h>
40 #include <stdio.h>
41
42 #include "gscanner.h"
43
44 #include "gprintfint.h"
45 #include "gstrfuncs.h"
46 #include "gstring.h"
47 #include "gtestutils.h"
48
49 #ifdef G_OS_UNIX
50 #include <unistd.h>
51 #endif
52 #ifdef G_OS_WIN32
53 #include <io.h>
54 #endif
55
56
57 /**
58  * GScannerMsgFunc:
59  * @scanner: a #GScanner
60  * @message: the message
61  * @error: %TRUE if the message signals an error,
62  *     %FALSE if it signals a warning.
63  *
64  * Specifies the type of the message handler function.
65  */
66
67 /**
68  * G_CSET_a_2_z:
69  *
70  * The set of lowercase ASCII alphabet characters.
71  * Used for specifying valid identifier characters
72  * in #GScannerConfig.
73  */
74
75 /**
76  * G_CSET_A_2_Z:
77  *
78  * The set of uppercase ASCII alphabet characters.
79  * Used for specifying valid identifier characters
80  * in #GScannerConfig.
81  */
82
83 /**
84  * G_CSET_DIGITS:
85  *
86  * The set of ASCII digits.
87  * Used for specifying valid identifier characters
88  * in #GScannerConfig.
89  */
90
91 /**
92  * G_CSET_LATINC:
93  *
94  * The set of uppercase ISO 8859-1 alphabet characters
95  * which are not ASCII characters.
96  * Used for specifying valid identifier characters
97  * in #GScannerConfig.
98  */
99
100 /**
101  * G_CSET_LATINS:
102  *
103  * The set of lowercase ISO 8859-1 alphabet characters
104  * which are not ASCII characters.
105  * Used for specifying valid identifier characters
106  * in #GScannerConfig.
107  */
108
109 /**
110  * GTokenType:
111  * @G_TOKEN_EOF: the end of the file
112  * @G_TOKEN_LEFT_PAREN: a '(' character
113  * @G_TOKEN_LEFT_CURLY: a '{' character
114  * @G_TOKEN_LEFT_BRACE: a '[' character
115  * @G_TOKEN_RIGHT_CURLY: a '}' character
116  * @G_TOKEN_RIGHT_PAREN: a ')' character
117  * @G_TOKEN_RIGHT_BRACE: a ']' character
118  * @G_TOKEN_EQUAL_SIGN: a '=' character
119  * @G_TOKEN_COMMA: a ',' character
120  * @G_TOKEN_NONE: not a token
121  * @G_TOKEN_ERROR: an error occurred
122  * @G_TOKEN_CHAR: a character
123  * @G_TOKEN_BINARY: a binary integer
124  * @G_TOKEN_OCTAL: an octal integer
125  * @G_TOKEN_INT: an integer
126  * @G_TOKEN_HEX: a hex integer
127  * @G_TOKEN_FLOAT: a floating point number
128  * @G_TOKEN_STRING: a string
129  * @G_TOKEN_SYMBOL: a symbol
130  * @G_TOKEN_IDENTIFIER: an identifier
131  * @G_TOKEN_IDENTIFIER_NULL: a null identifier
132  * @G_TOKEN_COMMENT_SINGLE: one line comment
133  * @G_TOKEN_COMMENT_MULTI: multi line comment
134  *
135  * The possible types of token returned from each
136  * g_scanner_get_next_token() call.
137  */
138
139 /**
140  * GTokenValue:
141  * @v_symbol: token symbol value
142  * @v_identifier: token identifier value
143  * @v_binary: token binary integer value
144  * @v_octal: octal integer value
145  * @v_int: integer value
146  * @v_int64: 64-bit integer value
147  * @v_float: floating point value
148  * @v_hex: hex integer value
149  * @v_string: string value
150  * @v_comment: comment value
151  * @v_char: character value
152  * @v_error: error value
153  *
154  * A union holding the value of the token.
155  */
156
157 /**
158  * GErrorType:
159  * @G_ERR_UNKNOWN: unknown error
160  * @G_ERR_UNEXP_EOF: unexpected end of file
161  * @G_ERR_UNEXP_EOF_IN_STRING: unterminated string constant
162  * @G_ERR_UNEXP_EOF_IN_COMMENT: unterminated comment
163  * @G_ERR_NON_DIGIT_IN_CONST: non-digit character in a number
164  * @G_ERR_DIGIT_RADIX: digit beyond radix in a number
165  * @G_ERR_FLOAT_RADIX: non-decimal floating point number
166  * @G_ERR_FLOAT_MALFORMED: malformed floating point number
167  *
168  * The possible errors, used in the @v_error field
169  * of #GTokenValue, when the token is a %G_TOKEN_ERROR.
170  */
171
172 /**
173  * GScanner:
174  * @user_data: unused
175  * @max_parse_errors: unused
176  * @parse_errors: g_scanner_error() increments this field
177  * @input_name: name of input stream, featured by the default message handler
178  * @qdata: quarked data
179  * @config: link into the scanner configuration
180  * @token: token parsed by the last g_scanner_get_next_token()
181  * @value: value of the last token from g_scanner_get_next_token()
182  * @line: line number of the last token from g_scanner_get_next_token()
183  * @position: char number of the last token from g_scanner_get_next_token()
184  * @next_token: token parsed by the last g_scanner_peek_next_token()
185  * @next_value: value of the last token from g_scanner_peek_next_token()
186  * @next_line: line number of the last token from g_scanner_peek_next_token()
187  * @next_position: char number of the last token from g_scanner_peek_next_token()
188  * @msg_handler: handler function for _warn and _error
189  *
190  * `GScanner` provides a general-purpose lexical scanner.
191  *
192  * You should set @input_name after creating the scanner, since
193  * it is used by the default message handler when displaying
194  * warnings and errors. If you are scanning a file, the filename
195  * would be a good choice.
196  *
197  * The @user_data and @max_parse_errors fields are not used.
198  * If you need to associate extra data with the scanner you
199  * can place them here.
200  *
201  * If you want to use your own message handler you can set the
202  * @msg_handler field. The type of the message handler function
203  * is declared by #GScannerMsgFunc.
204  */
205
206 /**
207  * GScannerConfig:
208  * @cset_skip_characters: specifies which characters should be skipped
209  *     by the scanner (the default is the whitespace characters: space,
210  *     tab, carriage-return and line-feed).
211  * @cset_identifier_first: specifies the characters which can start
212  *     identifiers (the default is %G_CSET_a_2_z, "_", and %G_CSET_A_2_Z).
213  * @cset_identifier_nth: specifies the characters which can be used
214  *     in identifiers, after the first character (the default is
215  *     %G_CSET_a_2_z, "_0123456789", %G_CSET_A_2_Z, %G_CSET_LATINS,
216  *     %G_CSET_LATINC).
217  * @cpair_comment_single: specifies the characters at the start and
218  *     end of single-line comments. The default is "#\n" which means
219  *     that single-line comments start with a '#' and continue until
220  *     a '\n' (end of line).
221  * @case_sensitive: specifies if symbols are case sensitive (the
222  *     default is %FALSE).
223  * @skip_comment_multi: specifies if multi-line comments are skipped
224  *     and not returned as tokens (the default is %TRUE).
225  * @skip_comment_single: specifies if single-line comments are skipped
226  *     and not returned as tokens (the default is %TRUE).
227  * @scan_comment_multi: specifies if multi-line comments are recognized
228  *     (the default is %TRUE).
229  * @scan_identifier: specifies if identifiers are recognized (the
230  *     default is %TRUE).
231  * @scan_identifier_1char: specifies if single-character
232  *     identifiers are recognized (the default is %FALSE).
233  * @scan_identifier_NULL: specifies if %NULL is reported as
234  *     %G_TOKEN_IDENTIFIER_NULL (the default is %FALSE).
235  * @scan_symbols: specifies if symbols are recognized (the default
236  *     is %TRUE).
237  * @scan_binary: specifies if binary numbers are recognized (the
238  *     default is %FALSE).
239  * @scan_octal: specifies if octal numbers are recognized (the
240  *     default is %TRUE).
241  * @scan_float: specifies if floating point numbers are recognized
242  *     (the default is %TRUE).
243  * @scan_hex: specifies if hexadecimal numbers are recognized (the
244  *     default is %TRUE).
245  * @scan_hex_dollar: specifies if '$' is recognized as a prefix for
246  *     hexadecimal numbers (the default is %FALSE).
247  * @scan_string_sq: specifies if strings can be enclosed in single
248  *     quotes (the default is %TRUE).
249  * @scan_string_dq: specifies if strings can be enclosed in double
250  *     quotes (the default is %TRUE).
251  * @numbers_2_int: specifies if binary, octal and hexadecimal numbers
252  *     are reported as %G_TOKEN_INT (the default is %TRUE).
253  * @int_2_float: specifies if all numbers are reported as %G_TOKEN_FLOAT
254  *     (the default is %FALSE).
255  * @identifier_2_string: specifies if identifiers are reported as strings
256  *     (the default is %FALSE).
257  * @char_2_token: specifies if characters are reported by setting
258  *     `token = ch` or as %G_TOKEN_CHAR (the default is %TRUE).
259  * @symbol_2_token: specifies if symbols are reported by setting
260  *     `token = v_symbol` or as %G_TOKEN_SYMBOL (the default is %FALSE).
261  * @scope_0_fallback: specifies if a symbol is searched for in the
262  *     default scope in addition to the current scope (the default is %FALSE).
263  * @store_int64: use value.v_int64 rather than v_int
264  *
265  * Specifies the #GScanner parser configuration. Most settings can
266  * be changed during the parsing phase and will affect the lexical
267  * parsing of the next unpeeked token.
268  */
269
270 /* --- defines --- */
271 #define to_lower(c)                             ( \
272         (guchar) (                                                      \
273           ( (((guchar)(c))>='A' && ((guchar)(c))<='Z') * ('a'-'A') ) |  \
274           ( (((guchar)(c))>=192 && ((guchar)(c))<=214) * (224-192) ) |  \
275           ( (((guchar)(c))>=216 && ((guchar)(c))<=222) * (248-216) ) |  \
276           ((guchar)(c))                                                 \
277         )                                                               \
278 )
279 #define READ_BUFFER_SIZE        (4000)
280
281
282 /* --- typedefs --- */
283 typedef struct  _GScannerKey    GScannerKey;
284
285 struct  _GScannerKey
286 {
287   guint          scope_id;
288   gchar         *symbol;
289   gpointer       value;
290 };
291
292
293 /* --- variables --- */
294 static const GScannerConfig g_scanner_config_template =
295 {
296   (
297    " \t\r\n"
298    )                    /* cset_skip_characters */,
299   (
300    G_CSET_a_2_z
301    "_"
302    G_CSET_A_2_Z
303    )                    /* cset_identifier_first */,
304   (
305    G_CSET_a_2_z
306    "_"
307    G_CSET_A_2_Z
308    G_CSET_DIGITS
309    G_CSET_LATINS
310    G_CSET_LATINC
311    )                    /* cset_identifier_nth */,
312   ( "#\n" )             /* cpair_comment_single */,
313   
314   FALSE                 /* case_sensitive */,
315   
316   TRUE                  /* skip_comment_multi */,
317   TRUE                  /* skip_comment_single */,
318   TRUE                  /* scan_comment_multi */,
319   TRUE                  /* scan_identifier */,
320   FALSE                 /* scan_identifier_1char */,
321   FALSE                 /* scan_identifier_NULL */,
322   TRUE                  /* scan_symbols */,
323   FALSE                 /* scan_binary */,
324   TRUE                  /* scan_octal */,
325   TRUE                  /* scan_float */,
326   TRUE                  /* scan_hex */,
327   FALSE                 /* scan_hex_dollar */,
328   TRUE                  /* scan_string_sq */,
329   TRUE                  /* scan_string_dq */,
330   TRUE                  /* numbers_2_int */,
331   FALSE                 /* int_2_float */,
332   FALSE                 /* identifier_2_string */,
333   TRUE                  /* char_2_token */,
334   FALSE                 /* symbol_2_token */,
335   FALSE                 /* scope_0_fallback */,
336   FALSE                 /* store_int64 */,
337   0                     /* padding_dummy */
338 };
339
340
341 /* --- prototypes --- */
342 static inline
343 GScannerKey*    g_scanner_lookup_internal (GScanner     *scanner,
344                                            guint         scope_id,
345                                            const gchar  *symbol);
346 static gboolean g_scanner_key_equal       (gconstpointer v1,
347                                            gconstpointer v2);
348 static guint    g_scanner_key_hash        (gconstpointer v);
349 static void     g_scanner_get_token_ll    (GScanner     *scanner,
350                                            GTokenType   *token_p,
351                                            GTokenValue  *value_p,
352                                            guint        *line_p,
353                                            guint        *position_p);
354 static void     g_scanner_get_token_i     (GScanner     *scanner,
355                                            GTokenType   *token_p,
356                                            GTokenValue  *value_p,
357                                            guint        *line_p,
358                                            guint        *position_p);
359
360 static guchar   g_scanner_peek_next_char  (GScanner     *scanner);
361 static guchar   g_scanner_get_char        (GScanner     *scanner,
362                                            guint        *line_p,
363                                            guint        *position_p);
364 static void     g_scanner_msg_handler     (GScanner     *scanner,
365                                            gchar        *message,
366                                            gboolean      is_error);
367
368
369 /* --- functions --- */
370 static inline gint
371 g_scanner_char_2_num (guchar    c,
372                       guchar    base)
373 {
374   if (c >= '0' && c <= '9')
375     c -= '0';
376   else if (c >= 'A' && c <= 'Z')
377     c -= 'A' - 10;
378   else if (c >= 'a' && c <= 'z')
379     c -= 'a' - 10;
380   else
381     return -1;
382   
383   if (c < base)
384     return c;
385   
386   return -1;
387 }
388
389 /**
390  * g_scanner_new:
391  * @config_templ: the initial scanner settings
392  *
393  * Creates a new #GScanner.
394  *
395  * The @config_templ structure specifies the initial settings
396  * of the scanner, which are copied into the #GScanner
397  * @config field. If you pass %NULL then the default settings
398  * are used.
399  *
400  * Returns: the new #GScanner
401  */
402 GScanner *
403 g_scanner_new (const GScannerConfig *config_templ)
404 {
405   GScanner *scanner;
406   
407   if (!config_templ)
408     config_templ = &g_scanner_config_template;
409   
410   scanner = g_new0 (GScanner, 1);
411   
412   scanner->user_data = NULL;
413   scanner->max_parse_errors = 1;
414   scanner->parse_errors = 0;
415   scanner->input_name = NULL;
416   g_datalist_init (&scanner->qdata);
417   
418   scanner->config = g_new0 (GScannerConfig, 1);
419   
420   scanner->config->case_sensitive        = config_templ->case_sensitive;
421   scanner->config->cset_skip_characters  = config_templ->cset_skip_characters;
422   if (!scanner->config->cset_skip_characters)
423     scanner->config->cset_skip_characters = "";
424   scanner->config->cset_identifier_first = config_templ->cset_identifier_first;
425   scanner->config->cset_identifier_nth   = config_templ->cset_identifier_nth;
426   scanner->config->cpair_comment_single  = config_templ->cpair_comment_single;
427   scanner->config->skip_comment_multi    = config_templ->skip_comment_multi;
428   scanner->config->skip_comment_single   = config_templ->skip_comment_single;
429   scanner->config->scan_comment_multi    = config_templ->scan_comment_multi;
430   scanner->config->scan_identifier       = config_templ->scan_identifier;
431   scanner->config->scan_identifier_1char = config_templ->scan_identifier_1char;
432   scanner->config->scan_identifier_NULL  = config_templ->scan_identifier_NULL;
433   scanner->config->scan_symbols          = config_templ->scan_symbols;
434   scanner->config->scan_binary           = config_templ->scan_binary;
435   scanner->config->scan_octal            = config_templ->scan_octal;
436   scanner->config->scan_float            = config_templ->scan_float;
437   scanner->config->scan_hex              = config_templ->scan_hex;
438   scanner->config->scan_hex_dollar       = config_templ->scan_hex_dollar;
439   scanner->config->scan_string_sq        = config_templ->scan_string_sq;
440   scanner->config->scan_string_dq        = config_templ->scan_string_dq;
441   scanner->config->numbers_2_int         = config_templ->numbers_2_int;
442   scanner->config->int_2_float           = config_templ->int_2_float;
443   scanner->config->identifier_2_string   = config_templ->identifier_2_string;
444   scanner->config->char_2_token          = config_templ->char_2_token;
445   scanner->config->symbol_2_token        = config_templ->symbol_2_token;
446   scanner->config->scope_0_fallback      = config_templ->scope_0_fallback;
447   scanner->config->store_int64           = config_templ->store_int64;
448   
449   scanner->token = G_TOKEN_NONE;
450   scanner->value.v_int64 = 0;
451   scanner->line = 1;
452   scanner->position = 0;
453   
454   scanner->next_token = G_TOKEN_NONE;
455   scanner->next_value.v_int64 = 0;
456   scanner->next_line = 1;
457   scanner->next_position = 0;
458   
459   scanner->symbol_table = g_hash_table_new (g_scanner_key_hash, g_scanner_key_equal);
460   scanner->input_fd = -1;
461   scanner->text = NULL;
462   scanner->text_end = NULL;
463   scanner->buffer = NULL;
464   scanner->scope_id = 0;
465   
466   scanner->msg_handler = g_scanner_msg_handler;
467   
468   return scanner;
469 }
470
471 static inline void
472 g_scanner_free_value (GTokenType     *token_p,
473                       GTokenValue     *value_p)
474 {
475   switch (*token_p)
476     {
477     case G_TOKEN_STRING:
478     case G_TOKEN_IDENTIFIER:
479     case G_TOKEN_IDENTIFIER_NULL:
480     case G_TOKEN_COMMENT_SINGLE:
481     case G_TOKEN_COMMENT_MULTI:
482       g_free (value_p->v_string);
483       break;
484       
485     default:
486       break;
487     }
488   
489   *token_p = G_TOKEN_NONE;
490 }
491
492 static void
493 g_scanner_destroy_symbol_table_entry (gpointer _key,
494                                       gpointer _value,
495                                       gpointer _data)
496 {
497   GScannerKey *key = _key;
498   
499   g_free (key->symbol);
500   g_free (key);
501 }
502
503 /**
504  * g_scanner_destroy:
505  * @scanner: a #GScanner
506  *
507  * Frees all memory used by the #GScanner.
508  */
509 void
510 g_scanner_destroy (GScanner *scanner)
511 {
512   g_return_if_fail (scanner != NULL);
513   
514   g_datalist_clear (&scanner->qdata);
515   g_hash_table_foreach (scanner->symbol_table, 
516                         g_scanner_destroy_symbol_table_entry, NULL);
517   g_hash_table_destroy (scanner->symbol_table);
518   g_scanner_free_value (&scanner->token, &scanner->value);
519   g_scanner_free_value (&scanner->next_token, &scanner->next_value);
520   g_free (scanner->config);
521   g_free (scanner->buffer);
522   g_free (scanner);
523 }
524
525 static void
526 g_scanner_msg_handler (GScanner         *scanner,
527                        gchar            *message,
528                        gboolean         is_error)
529 {
530   g_return_if_fail (scanner != NULL);
531   
532   _g_fprintf (stderr, "%s:%d: ",
533               scanner->input_name ? scanner->input_name : "<memory>",
534               scanner->line);
535   if (is_error)
536     _g_fprintf (stderr, "error: ");
537   _g_fprintf (stderr, "%s\n", message);
538 }
539
540 /**
541  * g_scanner_error:
542  * @scanner: a #GScanner
543  * @format: the message format. See the printf() documentation
544  * @...: the parameters to insert into the format string
545  *
546  * Outputs an error message, via the #GScanner message handler.
547  */
548 void
549 g_scanner_error (GScanner       *scanner,
550                  const gchar    *format,
551                  ...)
552 {
553   g_return_if_fail (scanner != NULL);
554   g_return_if_fail (format != NULL);
555   
556   scanner->parse_errors++;
557   
558   if (scanner->msg_handler)
559     {
560       va_list args;
561       gchar *string;
562       
563       va_start (args, format);
564       string = g_strdup_vprintf (format, args);
565       va_end (args);
566       
567       scanner->msg_handler (scanner, string, TRUE);
568       
569       g_free (string);
570     }
571 }
572
573 /**
574  * g_scanner_warn:
575  * @scanner: a #GScanner
576  * @format: the message format. See the printf() documentation
577  * @...: the parameters to insert into the format string
578  *
579  * Outputs a warning message, via the #GScanner message handler.
580  */
581 void
582 g_scanner_warn (GScanner       *scanner,
583                 const gchar    *format,
584                 ...)
585 {
586   g_return_if_fail (scanner != NULL);
587   g_return_if_fail (format != NULL);
588   
589   if (scanner->msg_handler)
590     {
591       va_list args;
592       gchar *string;
593       
594       va_start (args, format);
595       string = g_strdup_vprintf (format, args);
596       va_end (args);
597       
598       scanner->msg_handler (scanner, string, FALSE);
599       
600       g_free (string);
601     }
602 }
603
604 static gboolean
605 g_scanner_key_equal (gconstpointer v1,
606                      gconstpointer v2)
607 {
608   const GScannerKey *key1 = v1;
609   const GScannerKey *key2 = v2;
610   
611   return (key1->scope_id == key2->scope_id) && (strcmp (key1->symbol, key2->symbol) == 0);
612 }
613
614 static guint
615 g_scanner_key_hash (gconstpointer v)
616 {
617   const GScannerKey *key = v;
618   gchar *c;
619   guint h;
620   
621   h = key->scope_id;
622   for (c = key->symbol; *c; c++)
623     h = (h << 5) - h + *c;
624   
625   return h;
626 }
627
628 static inline GScannerKey*
629 g_scanner_lookup_internal (GScanner     *scanner,
630                            guint         scope_id,
631                            const gchar  *symbol)
632 {
633   GScannerKey   *key_p;
634   GScannerKey key;
635   
636   key.scope_id = scope_id;
637   
638   if (!scanner->config->case_sensitive)
639     {
640       gchar *d;
641       const gchar *c;
642       
643       key.symbol = g_new (gchar, strlen (symbol) + 1);
644       for (d = key.symbol, c = symbol; *c; c++, d++)
645         *d = to_lower (*c);
646       *d = 0;
647       key_p = g_hash_table_lookup (scanner->symbol_table, &key);
648       g_free (key.symbol);
649     }
650   else
651     {
652       key.symbol = (gchar*) symbol;
653       key_p = g_hash_table_lookup (scanner->symbol_table, &key);
654     }
655   
656   return key_p;
657 }
658
659 /**
660  * g_scanner_add_symbol:
661  * @scanner: a #GScanner
662  * @symbol: the symbol to add
663  * @value: the value of the symbol
664  *
665  * Adds a symbol to the default scope.
666  *
667  * Deprecated: 2.2: Use g_scanner_scope_add_symbol() instead.
668  */
669
670 /**
671  * g_scanner_scope_add_symbol:
672  * @scanner: a #GScanner
673  * @scope_id: the scope id
674  * @symbol: the symbol to add
675  * @value: the value of the symbol
676  *
677  * Adds a symbol to the given scope.
678  */
679 void
680 g_scanner_scope_add_symbol (GScanner    *scanner,
681                             guint        scope_id,
682                             const gchar *symbol,
683                             gpointer     value)
684 {
685   GScannerKey   *key;
686   
687   g_return_if_fail (scanner != NULL);
688   g_return_if_fail (symbol != NULL);
689   
690   key = g_scanner_lookup_internal (scanner, scope_id, symbol);
691   
692   if (!key)
693     {
694       key = g_new (GScannerKey, 1);
695       key->scope_id = scope_id;
696       key->symbol = g_strdup (symbol);
697       key->value = value;
698       if (!scanner->config->case_sensitive)
699         {
700           gchar *c;
701           
702           c = key->symbol;
703           while (*c != 0)
704             {
705               *c = to_lower (*c);
706               c++;
707             }
708         }
709       g_hash_table_add (scanner->symbol_table, key);
710     }
711   else
712     key->value = value;
713 }
714
715 /**
716  * g_scanner_remove_symbol:
717  * @scanner: a #GScanner
718  * @symbol: the symbol to remove
719  *
720  * Removes a symbol from the default scope.
721  *
722  * Deprecated: 2.2: Use g_scanner_scope_remove_symbol() instead.
723  */
724
725 /**
726  * g_scanner_scope_remove_symbol:
727  * @scanner: a #GScanner
728  * @scope_id: the scope id
729  * @symbol: the symbol to remove
730  *
731  * Removes a symbol from a scope.
732  */
733 void
734 g_scanner_scope_remove_symbol (GScanner    *scanner,
735                                guint        scope_id,
736                                const gchar *symbol)
737 {
738   GScannerKey   *key;
739   
740   g_return_if_fail (scanner != NULL);
741   g_return_if_fail (symbol != NULL);
742   
743   key = g_scanner_lookup_internal (scanner, scope_id, symbol);
744   
745   if (key)
746     {
747       g_hash_table_remove (scanner->symbol_table, key);
748       g_free (key->symbol);
749       g_free (key);
750     }
751 }
752
753 /**
754  * g_scanner_freeze_symbol_table:
755  * @scanner: a #GScanner
756  *
757  * There is no reason to use this macro, since it does nothing.
758  *
759  * Deprecated: 2.2: This macro does nothing.
760  */
761
762 /**
763  * g_scanner_thaw_symbol_table:
764  * @scanner: a #GScanner
765  *
766  * There is no reason to use this macro, since it does nothing.
767  *
768  * Deprecated: 2.2: This macro does nothing.
769  */
770
771 /**
772  * g_scanner_lookup_symbol:
773  * @scanner: a #GScanner
774  * @symbol: the symbol to look up
775  *
776  * Looks up a symbol in the current scope and return its value.
777  * If the symbol is not bound in the current scope, %NULL is
778  * returned.
779  *
780  * Returns: the value of @symbol in the current scope, or %NULL
781  *     if @symbol is not bound in the current scope
782  */
783 gpointer
784 g_scanner_lookup_symbol (GScanner       *scanner,
785                          const gchar    *symbol)
786 {
787   GScannerKey   *key;
788   guint scope_id;
789   
790   g_return_val_if_fail (scanner != NULL, NULL);
791   
792   if (!symbol)
793     return NULL;
794   
795   scope_id = scanner->scope_id;
796   key = g_scanner_lookup_internal (scanner, scope_id, symbol);
797   if (!key && scope_id && scanner->config->scope_0_fallback)
798     key = g_scanner_lookup_internal (scanner, 0, symbol);
799   
800   if (key)
801     return key->value;
802   else
803     return NULL;
804 }
805
806 /**
807  * g_scanner_scope_lookup_symbol:
808  * @scanner: a #GScanner
809  * @scope_id: the scope id
810  * @symbol: the symbol to look up
811  *
812  * Looks up a symbol in a scope and return its value. If the
813  * symbol is not bound in the scope, %NULL is returned.
814  *
815  * Returns: the value of @symbol in the given scope, or %NULL
816  *     if @symbol is not bound in the given scope.
817  *
818  */
819 gpointer
820 g_scanner_scope_lookup_symbol (GScanner       *scanner,
821                                guint           scope_id,
822                                const gchar    *symbol)
823 {
824   GScannerKey   *key;
825   
826   g_return_val_if_fail (scanner != NULL, NULL);
827   
828   if (!symbol)
829     return NULL;
830   
831   key = g_scanner_lookup_internal (scanner, scope_id, symbol);
832   
833   if (key)
834     return key->value;
835   else
836     return NULL;
837 }
838
839 /**
840  * g_scanner_set_scope:
841  * @scanner: a #GScanner
842  * @scope_id: the new scope id
843  *
844  * Sets the current scope.
845  *
846  * Returns: the old scope id
847  */
848 guint
849 g_scanner_set_scope (GScanner       *scanner,
850                      guint           scope_id)
851 {
852   guint old_scope_id;
853   
854   g_return_val_if_fail (scanner != NULL, 0);
855   
856   old_scope_id = scanner->scope_id;
857   scanner->scope_id = scope_id;
858   
859   return old_scope_id;
860 }
861
862 static void
863 g_scanner_foreach_internal (gpointer  _key,
864                             gpointer  _value,
865                             gpointer  _user_data)
866 {
867   GScannerKey *key;
868   gpointer *d;
869   GHFunc func;
870   gpointer user_data;
871   guint *scope_id;
872   
873   d = _user_data;
874   func = (GHFunc) d[0];
875   user_data = d[1];
876   scope_id = d[2];
877   key = _value;
878   
879   if (key->scope_id == *scope_id)
880     func (key->symbol, key->value, user_data);
881 }
882
883 /**
884  * g_scanner_foreach_symbol:
885  * @scanner: a #GScanner
886  * @func: the function to call with each symbol
887  * @data: data to pass to the function
888  *
889  * Calls a function for each symbol in the default scope.
890  *
891  * Deprecated: 2.2: Use g_scanner_scope_foreach_symbol() instead.
892  */
893
894 /**
895  * g_scanner_scope_foreach_symbol:
896  * @scanner: a #GScanner
897  * @scope_id: the scope id
898  * @func: (scope call): the function to call for each symbol/value pair
899  * @user_data: user data to pass to the function
900  *
901  * Calls the given function for each of the symbol/value pairs
902  * in the given scope of the #GScanner. The function is passed
903  * the symbol and value of each pair, and the given @user_data
904  * parameter.
905  */
906 void
907 g_scanner_scope_foreach_symbol (GScanner       *scanner,
908                                 guint           scope_id,
909                                 GHFunc          func,
910                                 gpointer        user_data)
911 {
912   gpointer d[3];
913   
914   g_return_if_fail (scanner != NULL);
915   
916   d[0] = (gpointer) func;
917   d[1] = user_data;
918   d[2] = &scope_id;
919   
920   g_hash_table_foreach (scanner->symbol_table, g_scanner_foreach_internal, d);
921 }
922
923 /**
924  * g_scanner_peek_next_token:
925  * @scanner: a #GScanner
926  *
927  * Parses the next token, without removing it from the input stream.
928  * The token data is placed in the @next_token, @next_value, @next_line,
929  * and @next_position fields of the #GScanner structure.
930  *
931  * Note that, while the token is not removed from the input stream
932  * (i.e. the next call to g_scanner_get_next_token() will return the
933  * same token), it will not be reevaluated. This can lead to surprising
934  * results when changing scope or the scanner configuration after peeking
935  * the next token. Getting the next token after switching the scope or
936  * configuration will return whatever was peeked before, regardless of
937  * any symbols that may have been added or removed in the new scope.
938  *
939  * Returns: the type of the token
940  */
941 GTokenType
942 g_scanner_peek_next_token (GScanner     *scanner)
943 {
944   g_return_val_if_fail (scanner != NULL, G_TOKEN_EOF);
945   
946   if (scanner->next_token == G_TOKEN_NONE)
947     {
948       scanner->next_line = scanner->line;
949       scanner->next_position = scanner->position;
950       g_scanner_get_token_i (scanner,
951                              &scanner->next_token,
952                              &scanner->next_value,
953                              &scanner->next_line,
954                              &scanner->next_position);
955     }
956   
957   return scanner->next_token;
958 }
959
960 /**
961  * g_scanner_get_next_token:
962  * @scanner: a #GScanner
963  *
964  * Parses the next token just like g_scanner_peek_next_token()
965  * and also removes it from the input stream. The token data is
966  * placed in the @token, @value, @line, and @position fields of
967  * the #GScanner structure.
968  *
969  * Returns: the type of the token
970  */
971 GTokenType
972 g_scanner_get_next_token (GScanner      *scanner)
973 {
974   g_return_val_if_fail (scanner != NULL, G_TOKEN_EOF);
975   
976   if (scanner->next_token != G_TOKEN_NONE)
977     {
978       g_scanner_free_value (&scanner->token, &scanner->value);
979       
980       scanner->token = scanner->next_token;
981       scanner->value = scanner->next_value;
982       scanner->line = scanner->next_line;
983       scanner->position = scanner->next_position;
984       scanner->next_token = G_TOKEN_NONE;
985     }
986   else
987     g_scanner_get_token_i (scanner,
988                            &scanner->token,
989                            &scanner->value,
990                            &scanner->line,
991                            &scanner->position);
992   
993   return scanner->token;
994 }
995
996 /**
997  * g_scanner_cur_token:
998  * @scanner: a #GScanner
999  *
1000  * Gets the current token type. This is simply the @token
1001  * field in the #GScanner structure.
1002  *
1003  * Returns: the current token type
1004  */
1005 GTokenType
1006 g_scanner_cur_token (GScanner *scanner)
1007 {
1008   g_return_val_if_fail (scanner != NULL, G_TOKEN_EOF);
1009   
1010   return scanner->token;
1011 }
1012
1013 /**
1014  * g_scanner_cur_value:
1015  * @scanner: a #GScanner
1016  *
1017  * Gets the current token value. This is simply the @value
1018  * field in the #GScanner structure.
1019  *
1020  * Returns: the current token value
1021  */
1022 GTokenValue
1023 g_scanner_cur_value (GScanner *scanner)
1024 {
1025   GTokenValue v;
1026   
1027   v.v_int64 = 0;
1028   
1029   g_return_val_if_fail (scanner != NULL, v);
1030
1031   /* MSC isn't capable of handling return scanner->value; ? */
1032
1033   v = scanner->value;
1034
1035   return v;
1036 }
1037
1038 /**
1039  * g_scanner_cur_line:
1040  * @scanner: a #GScanner
1041  *
1042  * Returns the current line in the input stream (counting
1043  * from 1). This is the line of the last token parsed via
1044  * g_scanner_get_next_token().
1045  *
1046  * Returns: the current line
1047  */
1048 guint
1049 g_scanner_cur_line (GScanner *scanner)
1050 {
1051   g_return_val_if_fail (scanner != NULL, 0);
1052   
1053   return scanner->line;
1054 }
1055
1056 /**
1057  * g_scanner_cur_position:
1058  * @scanner: a #GScanner
1059  *
1060  * Returns the current position in the current line (counting
1061  * from 0). This is the position of the last token parsed via
1062  * g_scanner_get_next_token().
1063  *
1064  * Returns: the current position on the line
1065  */
1066 guint
1067 g_scanner_cur_position (GScanner *scanner)
1068 {
1069   g_return_val_if_fail (scanner != NULL, 0);
1070   
1071   return scanner->position;
1072 }
1073
1074 /**
1075  * g_scanner_eof:
1076  * @scanner: a #GScanner
1077  *
1078  * Returns %TRUE if the scanner has reached the end of
1079  * the file or text buffer.
1080  *
1081  * Returns: %TRUE if the scanner has reached the end of
1082  *     the file or text buffer
1083  */
1084 gboolean
1085 g_scanner_eof (GScanner *scanner)
1086 {
1087   g_return_val_if_fail (scanner != NULL, TRUE);
1088   
1089   return scanner->token == G_TOKEN_EOF || scanner->token == G_TOKEN_ERROR;
1090 }
1091
1092 /**
1093  * g_scanner_input_file:
1094  * @scanner: a #GScanner
1095  * @input_fd: a file descriptor
1096  *
1097  * Prepares to scan a file.
1098  */
1099 void
1100 g_scanner_input_file (GScanner *scanner,
1101                       gint      input_fd)
1102 {
1103   g_return_if_fail (scanner != NULL);
1104   g_return_if_fail (input_fd >= 0);
1105
1106   if (scanner->input_fd >= 0)
1107     g_scanner_sync_file_offset (scanner);
1108
1109   scanner->token = G_TOKEN_NONE;
1110   scanner->value.v_int64 = 0;
1111   scanner->line = 1;
1112   scanner->position = 0;
1113   scanner->next_token = G_TOKEN_NONE;
1114
1115   scanner->input_fd = input_fd;
1116   scanner->text = NULL;
1117   scanner->text_end = NULL;
1118
1119   if (!scanner->buffer)
1120     scanner->buffer = g_new (gchar, READ_BUFFER_SIZE + 1);
1121 }
1122
1123 /**
1124  * g_scanner_input_text:
1125  * @scanner: a #GScanner
1126  * @text: the text buffer to scan
1127  * @text_len: the length of the text buffer
1128  *
1129  * Prepares to scan a text buffer.
1130  */
1131 void
1132 g_scanner_input_text (GScanner    *scanner,
1133                       const gchar *text,
1134                       guint        text_len)
1135 {
1136   g_return_if_fail (scanner != NULL);
1137   if (text_len)
1138     g_return_if_fail (text != NULL);
1139   else
1140     text = NULL;
1141
1142   if (scanner->input_fd >= 0)
1143     g_scanner_sync_file_offset (scanner);
1144
1145   scanner->token = G_TOKEN_NONE;
1146   scanner->value.v_int64 = 0;
1147   scanner->line = 1;
1148   scanner->position = 0;
1149   scanner->next_token = G_TOKEN_NONE;
1150
1151   scanner->input_fd = -1;
1152   scanner->text = text;
1153   scanner->text_end = text + text_len;
1154
1155   if (scanner->buffer)
1156     {
1157       g_free (scanner->buffer);
1158       scanner->buffer = NULL;
1159     }
1160 }
1161
1162 static guchar
1163 g_scanner_peek_next_char (GScanner *scanner)
1164 {
1165   if (scanner->text < scanner->text_end)
1166     {
1167       return *scanner->text;
1168     }
1169   else if (scanner->input_fd >= 0)
1170     {
1171       gint count;
1172       gchar *buffer;
1173
1174       buffer = scanner->buffer;
1175       do
1176         {
1177           count = read (scanner->input_fd, buffer, READ_BUFFER_SIZE);
1178         }
1179       while (count == -1 && (errno == EINTR || errno == EAGAIN));
1180
1181       if (count < 1)
1182         {
1183           scanner->input_fd = -1;
1184
1185           return 0;
1186         }
1187       else
1188         {
1189           scanner->text = buffer;
1190           scanner->text_end = buffer + count;
1191
1192           return *buffer;
1193         }
1194     }
1195   else
1196     return 0;
1197 }
1198
1199 /**
1200  * g_scanner_sync_file_offset:
1201  * @scanner: a #GScanner
1202  *
1203  * Rewinds the filedescriptor to the current buffer position
1204  * and blows the file read ahead buffer. This is useful for
1205  * third party uses of the scanners filedescriptor, which hooks
1206  * onto the current scanning position.
1207  */
1208 void
1209 g_scanner_sync_file_offset (GScanner *scanner)
1210 {
1211   g_return_if_fail (scanner != NULL);
1212
1213   /* for file input, rewind the filedescriptor to the current
1214    * buffer position and blow the file read ahead buffer. useful
1215    * for third party uses of our file descriptor, which hooks 
1216    * onto the current scanning position.
1217    */
1218
1219   if (scanner->input_fd >= 0 && scanner->text_end > scanner->text)
1220     {
1221       gint buffered;
1222
1223       buffered = scanner->text_end - scanner->text;
1224       if (lseek (scanner->input_fd, - buffered, SEEK_CUR) >= 0)
1225         {
1226           /* we succeeded, blow our buffer's contents now */
1227           scanner->text = NULL;
1228           scanner->text_end = NULL;
1229         }
1230       else
1231         errno = 0;
1232     }
1233 }
1234
1235 static guchar
1236 g_scanner_get_char (GScanner    *scanner,
1237                     guint       *line_p,
1238                     guint       *position_p)
1239 {
1240   guchar fchar;
1241
1242   if (scanner->text < scanner->text_end)
1243     fchar = *(scanner->text++);
1244   else if (scanner->input_fd >= 0)
1245     {
1246       gint count;
1247       gchar *buffer;
1248
1249       buffer = scanner->buffer;
1250       do
1251         {
1252           count = read (scanner->input_fd, buffer, READ_BUFFER_SIZE);
1253         }
1254       while (count == -1 && (errno == EINTR || errno == EAGAIN));
1255
1256       if (count < 1)
1257         {
1258           scanner->input_fd = -1;
1259           fchar = 0;
1260         }
1261       else
1262         {
1263           scanner->text = buffer + 1;
1264           scanner->text_end = buffer + count;
1265           fchar = *buffer;
1266           if (!fchar)
1267             {
1268               g_scanner_sync_file_offset (scanner);
1269               scanner->text_end = scanner->text;
1270               scanner->input_fd = -1;
1271             }
1272         }
1273     }
1274   else
1275     fchar = 0;
1276   
1277   if (fchar == '\n')
1278     {
1279       (*position_p) = 0;
1280       (*line_p)++;
1281     }
1282   else if (fchar)
1283     {
1284       (*position_p)++;
1285     }
1286   
1287   return fchar;
1288 }
1289
1290 /**
1291  * g_scanner_unexp_token:
1292  * @scanner: a #GScanner
1293  * @expected_token: the expected token
1294  * @identifier_spec: a string describing how the scanner's user
1295  *     refers to identifiers (%NULL defaults to "identifier").
1296  *     This is used if @expected_token is %G_TOKEN_IDENTIFIER or
1297  *     %G_TOKEN_IDENTIFIER_NULL.
1298  * @symbol_spec: a string describing how the scanner's user refers
1299  *     to symbols (%NULL defaults to "symbol"). This is used if
1300  *     @expected_token is %G_TOKEN_SYMBOL or any token value greater
1301  *     than %G_TOKEN_LAST.
1302  * @symbol_name: the name of the symbol, if the scanner's current
1303  *     token is a symbol.
1304  * @message: a message string to output at the end of the
1305  *     warning/error, or %NULL.
1306  * @is_error: if %TRUE it is output as an error. If %FALSE it is
1307  *     output as a warning.
1308  *
1309  * Outputs a message through the scanner's msg_handler,
1310  * resulting from an unexpected token in the input stream.
1311  * Note that you should not call g_scanner_peek_next_token()
1312  * followed by g_scanner_unexp_token() without an intermediate
1313  * call to g_scanner_get_next_token(), as g_scanner_unexp_token()
1314  * evaluates the scanner's current token (not the peeked token)
1315  * to construct part of the message.
1316  */
1317 void
1318 g_scanner_unexp_token (GScanner         *scanner,
1319                        GTokenType        expected_token,
1320                        const gchar      *identifier_spec,
1321                        const gchar      *symbol_spec,
1322                        const gchar      *symbol_name,
1323                        const gchar      *message,
1324                        gint              is_error)
1325 {
1326   gchar *token_string;
1327   guint token_string_len;
1328   gchar *expected_string;
1329   guint expected_string_len;
1330   gchar *message_prefix;
1331   gboolean print_unexp;
1332   void (*msg_handler)   (GScanner*, const gchar*, ...);
1333   
1334   g_return_if_fail (scanner != NULL);
1335   
1336   if (is_error)
1337     msg_handler = g_scanner_error;
1338   else
1339     msg_handler = g_scanner_warn;
1340   
1341   if (!identifier_spec)
1342     identifier_spec = "identifier";
1343   if (!symbol_spec)
1344     symbol_spec = "symbol";
1345   
1346   token_string_len = 56;
1347   token_string = g_new (gchar, token_string_len + 1);
1348   expected_string_len = 64;
1349   expected_string = g_new (gchar, expected_string_len + 1);
1350   print_unexp = TRUE;
1351   
1352   switch (scanner->token)
1353     {
1354     case G_TOKEN_EOF:
1355       _g_snprintf (token_string, token_string_len, "end of file");
1356       break;
1357       
1358     default:
1359       if (scanner->token >= 1 && scanner->token <= 255)
1360         {
1361           if ((scanner->token >= ' ' && scanner->token <= '~') ||
1362               strchr (scanner->config->cset_identifier_first, scanner->token) ||
1363               strchr (scanner->config->cset_identifier_nth, scanner->token))
1364             _g_snprintf (token_string, token_string_len, "character '%c'", scanner->token);
1365           else
1366             _g_snprintf (token_string, token_string_len, "character '\\%o'", scanner->token);
1367           break;
1368         }
1369       else if (!scanner->config->symbol_2_token)
1370         {
1371           _g_snprintf (token_string, token_string_len, "(unknown) token <%d>", scanner->token);
1372           break;
1373         }
1374       G_GNUC_FALLTHROUGH;
1375     case G_TOKEN_SYMBOL:
1376       if (expected_token == G_TOKEN_SYMBOL ||
1377           (scanner->config->symbol_2_token &&
1378            expected_token > G_TOKEN_LAST))
1379         print_unexp = FALSE;
1380       if (symbol_name)
1381         _g_snprintf (token_string,
1382                      token_string_len,
1383                      "%s%s '%s'",
1384                      print_unexp ? "" : "invalid ",
1385                      symbol_spec,
1386                      symbol_name);
1387       else
1388         _g_snprintf (token_string,
1389                      token_string_len,
1390                      "%s%s",
1391                      print_unexp ? "" : "invalid ",
1392                      symbol_spec);
1393       break;
1394       
1395     case G_TOKEN_ERROR:
1396       print_unexp = FALSE;
1397       expected_token = G_TOKEN_NONE;
1398       switch (scanner->value.v_error)
1399         {
1400         case G_ERR_UNEXP_EOF:
1401           _g_snprintf (token_string, token_string_len, "scanner: unexpected end of file");
1402           break;
1403           
1404         case G_ERR_UNEXP_EOF_IN_STRING:
1405           _g_snprintf (token_string, token_string_len, "scanner: unterminated string constant");
1406           break;
1407           
1408         case G_ERR_UNEXP_EOF_IN_COMMENT:
1409           _g_snprintf (token_string, token_string_len, "scanner: unterminated comment");
1410           break;
1411           
1412         case G_ERR_NON_DIGIT_IN_CONST:
1413           _g_snprintf (token_string, token_string_len, "scanner: non digit in constant");
1414           break;
1415           
1416         case G_ERR_FLOAT_RADIX:
1417           _g_snprintf (token_string, token_string_len, "scanner: invalid radix for floating constant");
1418           break;
1419           
1420         case G_ERR_FLOAT_MALFORMED:
1421           _g_snprintf (token_string, token_string_len, "scanner: malformed floating constant");
1422           break;
1423           
1424         case G_ERR_DIGIT_RADIX:
1425           _g_snprintf (token_string, token_string_len, "scanner: digit is beyond radix");
1426           break;
1427           
1428         case G_ERR_UNKNOWN:
1429         default:
1430           _g_snprintf (token_string, token_string_len, "scanner: unknown error");
1431           break;
1432         }
1433       break;
1434       
1435     case G_TOKEN_CHAR:
1436       _g_snprintf (token_string, token_string_len, "character '%c'", scanner->value.v_char);
1437       break;
1438       
1439     case G_TOKEN_IDENTIFIER:
1440     case G_TOKEN_IDENTIFIER_NULL:
1441       if (expected_token == G_TOKEN_IDENTIFIER ||
1442           expected_token == G_TOKEN_IDENTIFIER_NULL)
1443         print_unexp = FALSE;
1444       _g_snprintf (token_string,
1445                   token_string_len,
1446                   "%s%s '%s'",
1447                   print_unexp ? "" : "invalid ",
1448                   identifier_spec,
1449                   scanner->token == G_TOKEN_IDENTIFIER ? scanner->value.v_string : "null");
1450       break;
1451       
1452     case G_TOKEN_BINARY:
1453     case G_TOKEN_OCTAL:
1454     case G_TOKEN_INT:
1455     case G_TOKEN_HEX:
1456       if (scanner->config->store_int64)
1457         _g_snprintf (token_string, token_string_len, "number '%" G_GUINT64_FORMAT "'", scanner->value.v_int64);
1458       else
1459         _g_snprintf (token_string, token_string_len, "number '%lu'", scanner->value.v_int);
1460       break;
1461       
1462     case G_TOKEN_FLOAT:
1463       _g_snprintf (token_string, token_string_len, "number '%.3f'", scanner->value.v_float);
1464       break;
1465       
1466     case G_TOKEN_STRING:
1467       if (expected_token == G_TOKEN_STRING)
1468         print_unexp = FALSE;
1469       _g_snprintf (token_string,
1470                    token_string_len,
1471                    "%s%sstring constant \"%s\"",
1472                    print_unexp ? "" : "invalid ",
1473                    scanner->value.v_string[0] == 0 ? "empty " : "",
1474                    scanner->value.v_string);
1475       token_string[token_string_len - 2] = '"';
1476       token_string[token_string_len - 1] = 0;
1477       break;
1478       
1479     case G_TOKEN_COMMENT_SINGLE:
1480     case G_TOKEN_COMMENT_MULTI:
1481       _g_snprintf (token_string, token_string_len, "comment");
1482       break;
1483       
1484     case G_TOKEN_NONE:
1485       /* somehow the user's parsing code is screwed, there isn't much
1486        * we can do about it.
1487        * Note, a common case to trigger this is
1488        * g_scanner_peek_next_token(); g_scanner_unexp_token();
1489        * without an intermediate g_scanner_get_next_token().
1490        */
1491       g_assert_not_reached ();
1492       break;
1493     }
1494   
1495   
1496   switch (expected_token)
1497     {
1498       gboolean need_valid;
1499       gchar *tstring;
1500     case G_TOKEN_EOF:
1501       _g_snprintf (expected_string, expected_string_len, "end of file");
1502       break;
1503     default:
1504       if (expected_token >= 1 && expected_token <= 255)
1505         {
1506           if ((expected_token >= ' ' && expected_token <= '~') ||
1507               strchr (scanner->config->cset_identifier_first, expected_token) ||
1508               strchr (scanner->config->cset_identifier_nth, expected_token))
1509             _g_snprintf (expected_string, expected_string_len, "character '%c'", expected_token);
1510           else
1511             _g_snprintf (expected_string, expected_string_len, "character '\\%o'", expected_token);
1512           break;
1513         }
1514       else if (!scanner->config->symbol_2_token)
1515         {
1516           _g_snprintf (expected_string, expected_string_len, "(unknown) token <%d>", expected_token);
1517           break;
1518         }
1519       G_GNUC_FALLTHROUGH;
1520     case G_TOKEN_SYMBOL:
1521       need_valid = (scanner->token == G_TOKEN_SYMBOL ||
1522                     (scanner->config->symbol_2_token &&
1523                      scanner->token > G_TOKEN_LAST));
1524       _g_snprintf (expected_string,
1525                    expected_string_len,
1526                    "%s%s",
1527                    need_valid ? "valid " : "",
1528                    symbol_spec);
1529       /* FIXME: should we attempt to look up the symbol_name for symbol_2_token? */
1530       break;
1531     case G_TOKEN_CHAR:
1532       _g_snprintf (expected_string, expected_string_len, "%scharacter",
1533                    scanner->token == G_TOKEN_CHAR ? "valid " : "");
1534       break;
1535     case G_TOKEN_BINARY:
1536       tstring = "binary";
1537       _g_snprintf (expected_string, expected_string_len, "%snumber (%s)",
1538                    scanner->token == expected_token ? "valid " : "", tstring);
1539       break;
1540     case G_TOKEN_OCTAL:
1541       tstring = "octal";
1542       _g_snprintf (expected_string, expected_string_len, "%snumber (%s)",
1543                    scanner->token == expected_token ? "valid " : "", tstring);
1544       break;
1545     case G_TOKEN_INT:
1546       tstring = "integer";
1547       _g_snprintf (expected_string, expected_string_len, "%snumber (%s)",
1548                    scanner->token == expected_token ? "valid " : "", tstring);
1549       break;
1550     case G_TOKEN_HEX:
1551       tstring = "hexadecimal";
1552       _g_snprintf (expected_string, expected_string_len, "%snumber (%s)",
1553                    scanner->token == expected_token ? "valid " : "", tstring);
1554       break;
1555     case G_TOKEN_FLOAT:
1556       tstring = "float";
1557       _g_snprintf (expected_string, expected_string_len, "%snumber (%s)",
1558                    scanner->token == expected_token ? "valid " : "", tstring);
1559       break;
1560     case G_TOKEN_STRING:
1561       _g_snprintf (expected_string,
1562                    expected_string_len,
1563                    "%sstring constant",
1564                    scanner->token == G_TOKEN_STRING ? "valid " : "");
1565       break;
1566     case G_TOKEN_IDENTIFIER:
1567     case G_TOKEN_IDENTIFIER_NULL:
1568       need_valid = (scanner->token == G_TOKEN_IDENTIFIER_NULL ||
1569                     scanner->token == G_TOKEN_IDENTIFIER);
1570       _g_snprintf (expected_string,
1571                    expected_string_len,
1572                    "%s%s",
1573                    need_valid ? "valid " : "",
1574                    identifier_spec);
1575       break;
1576     case G_TOKEN_COMMENT_SINGLE:
1577       tstring = "single-line";
1578       _g_snprintf (expected_string, expected_string_len, "%scomment (%s)",
1579                    scanner->token == expected_token ? "valid " : "", tstring);
1580       break;
1581     case G_TOKEN_COMMENT_MULTI:
1582       tstring = "multi-line";
1583       _g_snprintf (expected_string, expected_string_len, "%scomment (%s)",
1584                    scanner->token == expected_token ? "valid " : "", tstring);
1585       break;
1586     case G_TOKEN_NONE:
1587     case G_TOKEN_ERROR:
1588       /* this is handled upon printout */
1589       break;
1590     }
1591   
1592   if (message && message[0] != 0)
1593     message_prefix = " - ";
1594   else
1595     {
1596       message_prefix = "";
1597       message = "";
1598     }
1599   if (expected_token == G_TOKEN_ERROR)
1600     {
1601       msg_handler (scanner,
1602                    "failure around %s%s%s",
1603                    token_string,
1604                    message_prefix,
1605                    message);
1606     }
1607   else if (expected_token == G_TOKEN_NONE)
1608     {
1609       if (print_unexp)
1610         msg_handler (scanner,
1611                      "unexpected %s%s%s",
1612                      token_string,
1613                      message_prefix,
1614                      message);
1615       else
1616         msg_handler (scanner,
1617                      "%s%s%s",
1618                      token_string,
1619                      message_prefix,
1620                      message);
1621     }
1622   else
1623     {
1624       if (print_unexp)
1625         msg_handler (scanner,
1626                      "unexpected %s, expected %s%s%s",
1627                      token_string,
1628                      expected_string,
1629                      message_prefix,
1630                      message);
1631       else
1632         msg_handler (scanner,
1633                      "%s, expected %s%s%s",
1634                      token_string,
1635                      expected_string,
1636                      message_prefix,
1637                      message);
1638     }
1639   
1640   g_free (token_string);
1641   g_free (expected_string);
1642 }
1643
1644 static void
1645 g_scanner_get_token_i (GScanner *scanner,
1646                        GTokenType       *token_p,
1647                        GTokenValue      *value_p,
1648                        guint            *line_p,
1649                        guint            *position_p)
1650 {
1651   do
1652     {
1653       g_scanner_free_value (token_p, value_p);
1654       g_scanner_get_token_ll (scanner, token_p, value_p, line_p, position_p);
1655     }
1656   while (((*token_p > 0 && *token_p < 256) &&
1657           strchr (scanner->config->cset_skip_characters, *token_p)) ||
1658          (*token_p == G_TOKEN_CHAR &&
1659           strchr (scanner->config->cset_skip_characters, value_p->v_char)) ||
1660          (*token_p == G_TOKEN_COMMENT_MULTI &&
1661           scanner->config->skip_comment_multi) ||
1662          (*token_p == G_TOKEN_COMMENT_SINGLE &&
1663           scanner->config->skip_comment_single));
1664   
1665   switch (*token_p)
1666     {
1667     case G_TOKEN_IDENTIFIER:
1668       if (scanner->config->identifier_2_string)
1669         *token_p = G_TOKEN_STRING;
1670       break;
1671       
1672     case G_TOKEN_SYMBOL:
1673       if (scanner->config->symbol_2_token)
1674         *token_p = (GTokenType) ((size_t) value_p->v_symbol);
1675       break;
1676       
1677     case G_TOKEN_BINARY:
1678     case G_TOKEN_OCTAL:
1679     case G_TOKEN_HEX:
1680       if (scanner->config->numbers_2_int)
1681         *token_p = G_TOKEN_INT;
1682       break;
1683       
1684     default:
1685       break;
1686     }
1687   
1688   if (*token_p == G_TOKEN_INT &&
1689       scanner->config->int_2_float)
1690     {
1691       *token_p = G_TOKEN_FLOAT;
1692
1693       /* Have to assign through a temporary variable to avoid undefined behaviour
1694        * by copying between potentially-overlapping union members. */
1695       if (scanner->config->store_int64)
1696         {
1697           gint64 temp = value_p->v_int64;
1698           value_p->v_float = temp;
1699         }
1700       else
1701         {
1702           gint temp = value_p->v_int;
1703           value_p->v_float = temp;
1704         }
1705     }
1706   
1707   errno = 0;
1708 }
1709
1710 static void
1711 g_scanner_get_token_ll  (GScanner       *scanner,
1712                          GTokenType     *token_p,
1713                          GTokenValue    *value_p,
1714                          guint          *line_p,
1715                          guint          *position_p)
1716 {
1717   GScannerConfig *config;
1718   GTokenType       token;
1719   gboolean         in_comment_multi;
1720   gboolean         in_comment_single;
1721   gboolean         in_string_sq;
1722   gboolean         in_string_dq;
1723   GString         *gstring;
1724   GTokenValue      value;
1725   guchar           ch;
1726   
1727   config = scanner->config;
1728   (*value_p).v_int64 = 0;
1729   
1730   if ((scanner->text >= scanner->text_end && scanner->input_fd < 0) ||
1731       scanner->token == G_TOKEN_EOF)
1732     {
1733       *token_p = G_TOKEN_EOF;
1734       return;
1735     }
1736   
1737   in_comment_multi = FALSE;
1738   in_comment_single = FALSE;
1739   in_string_sq = FALSE;
1740   in_string_dq = FALSE;
1741   gstring = NULL;
1742   
1743   do /* while (ch != 0) */
1744     {
1745       gboolean dotted_float = FALSE;
1746       
1747       ch = g_scanner_get_char (scanner, line_p, position_p);
1748       
1749       value.v_int64 = 0;
1750       token = G_TOKEN_NONE;
1751       
1752       /* this is *evil*, but needed ;(
1753        * we first check for identifier first character, because  it
1754        * might interfere with other key chars like slashes or numbers
1755        */
1756       if (config->scan_identifier &&
1757           ch && strchr (config->cset_identifier_first, ch))
1758         goto identifier_precedence;
1759       
1760       switch (ch)
1761         {
1762         case 0:
1763           token = G_TOKEN_EOF;
1764           (*position_p)++;
1765           /* ch = 0; */
1766           break;
1767           
1768         case '/':
1769           if (!config->scan_comment_multi ||
1770               g_scanner_peek_next_char (scanner) != '*')
1771             goto default_case;
1772           g_scanner_get_char (scanner, line_p, position_p);
1773           token = G_TOKEN_COMMENT_MULTI;
1774           in_comment_multi = TRUE;
1775           gstring = g_string_new (NULL);
1776           while ((ch = g_scanner_get_char (scanner, line_p, position_p)) != 0)
1777             {
1778               if (ch == '*' && g_scanner_peek_next_char (scanner) == '/')
1779                 {
1780                   g_scanner_get_char (scanner, line_p, position_p);
1781                   in_comment_multi = FALSE;
1782                   break;
1783                 }
1784               else
1785                 gstring = g_string_append_c (gstring, ch);
1786             }
1787           ch = 0;
1788           break;
1789           
1790         case '\'':
1791           if (!config->scan_string_sq)
1792             goto default_case;
1793           token = G_TOKEN_STRING;
1794           in_string_sq = TRUE;
1795           gstring = g_string_new (NULL);
1796           while ((ch = g_scanner_get_char (scanner, line_p, position_p)) != 0)
1797             {
1798               if (ch == '\'')
1799                 {
1800                   in_string_sq = FALSE;
1801                   break;
1802                 }
1803               else
1804                 gstring = g_string_append_c (gstring, ch);
1805             }
1806           ch = 0;
1807           break;
1808           
1809         case '"':
1810           if (!config->scan_string_dq)
1811             goto default_case;
1812           token = G_TOKEN_STRING;
1813           in_string_dq = TRUE;
1814           gstring = g_string_new (NULL);
1815           while ((ch = g_scanner_get_char (scanner, line_p, position_p)) != 0)
1816             {
1817               if (ch == '"')
1818                 {
1819                   in_string_dq = FALSE;
1820                   break;
1821                 }
1822               else
1823                 {
1824                   if (ch == '\\')
1825                     {
1826                       ch = g_scanner_get_char (scanner, line_p, position_p);
1827                       switch (ch)
1828                         {
1829                           guint i;
1830                           guint fchar;
1831                           
1832                         case 0:
1833                           break;
1834                           
1835                         case '\\':
1836                           gstring = g_string_append_c (gstring, '\\');
1837                           break;
1838                           
1839                         case 'n':
1840                           gstring = g_string_append_c (gstring, '\n');
1841                           break;
1842                           
1843                         case 't':
1844                           gstring = g_string_append_c (gstring, '\t');
1845                           break;
1846                           
1847                         case 'r':
1848                           gstring = g_string_append_c (gstring, '\r');
1849                           break;
1850                           
1851                         case 'b':
1852                           gstring = g_string_append_c (gstring, '\b');
1853                           break;
1854                           
1855                         case 'f':
1856                           gstring = g_string_append_c (gstring, '\f');
1857                           break;
1858                           
1859                         case '0':
1860                         case '1':
1861                         case '2':
1862                         case '3':
1863                         case '4':
1864                         case '5':
1865                         case '6':
1866                         case '7':
1867                           i = ch - '0';
1868                           fchar = g_scanner_peek_next_char (scanner);
1869                           if (fchar >= '0' && fchar <= '7')
1870                             {
1871                               ch = g_scanner_get_char (scanner, line_p, position_p);
1872                               i = i * 8 + ch - '0';
1873                               fchar = g_scanner_peek_next_char (scanner);
1874                               if (fchar >= '0' && fchar <= '7')
1875                                 {
1876                                   ch = g_scanner_get_char (scanner, line_p, position_p);
1877                                   i = i * 8 + ch - '0';
1878                                 }
1879                             }
1880                           gstring = g_string_append_c (gstring, i);
1881                           break;
1882                           
1883                         default:
1884                           gstring = g_string_append_c (gstring, ch);
1885                           break;
1886                         }
1887                     }
1888                   else
1889                     gstring = g_string_append_c (gstring, ch);
1890                 }
1891             }
1892           ch = 0;
1893           break;
1894           
1895         case '.':
1896           if (!config->scan_float)
1897             goto default_case;
1898           token = G_TOKEN_FLOAT;
1899           dotted_float = TRUE;
1900           ch = g_scanner_get_char (scanner, line_p, position_p);
1901           goto number_parsing;
1902           
1903         case '$':
1904           if (!config->scan_hex_dollar)
1905             goto default_case;
1906           token = G_TOKEN_HEX;
1907           ch = g_scanner_get_char (scanner, line_p, position_p);
1908           goto number_parsing;
1909           
1910         case '0':
1911           if (config->scan_octal)
1912             token = G_TOKEN_OCTAL;
1913           else
1914             token = G_TOKEN_INT;
1915           ch = g_scanner_peek_next_char (scanner);
1916           if (config->scan_hex && (ch == 'x' || ch == 'X'))
1917             {
1918               token = G_TOKEN_HEX;
1919               g_scanner_get_char (scanner, line_p, position_p);
1920               ch = g_scanner_get_char (scanner, line_p, position_p);
1921               if (ch == 0)
1922                 {
1923                   token = G_TOKEN_ERROR;
1924                   value.v_error = G_ERR_UNEXP_EOF;
1925                   (*position_p)++;
1926                   break;
1927                 }
1928               if (g_scanner_char_2_num (ch, 16) < 0)
1929                 {
1930                   token = G_TOKEN_ERROR;
1931                   value.v_error = G_ERR_DIGIT_RADIX;
1932                   ch = 0;
1933                   break;
1934                 }
1935             }
1936           else if (config->scan_binary && (ch == 'b' || ch == 'B'))
1937             {
1938               token = G_TOKEN_BINARY;
1939               g_scanner_get_char (scanner, line_p, position_p);
1940               ch = g_scanner_get_char (scanner, line_p, position_p);
1941               if (ch == 0)
1942                 {
1943                   token = G_TOKEN_ERROR;
1944                   value.v_error = G_ERR_UNEXP_EOF;
1945                   (*position_p)++;
1946                   break;
1947                 }
1948               if (g_scanner_char_2_num (ch, 10) < 0)
1949                 {
1950                   token = G_TOKEN_ERROR;
1951                   value.v_error = G_ERR_NON_DIGIT_IN_CONST;
1952                   ch = 0;
1953                   break;
1954                 }
1955             }
1956           else
1957             ch = '0';
1958           G_GNUC_FALLTHROUGH;
1959         case '1':
1960         case '2':
1961         case '3':
1962         case '4':
1963         case '5':
1964         case '6':
1965         case '7':
1966         case '8':
1967         case '9':
1968         number_parsing:
1969         {
1970           gboolean in_number = TRUE;
1971           gchar *endptr;
1972           
1973           if (token == G_TOKEN_NONE)
1974             token = G_TOKEN_INT;
1975           
1976           gstring = g_string_new (dotted_float ? "0." : "");
1977           gstring = g_string_append_c (gstring, ch);
1978           
1979           do /* while (in_number) */
1980             {
1981               gboolean is_E;
1982               
1983               is_E = token == G_TOKEN_FLOAT && (ch == 'e' || ch == 'E');
1984               
1985               ch = g_scanner_peek_next_char (scanner);
1986               
1987               if (g_scanner_char_2_num (ch, 36) >= 0 ||
1988                   (config->scan_float && ch == '.') ||
1989                   (is_E && (ch == '+' || ch == '-')))
1990                 {
1991                   ch = g_scanner_get_char (scanner, line_p, position_p);
1992                   
1993                   switch (ch)
1994                     {
1995                     case '.':
1996                       if (token != G_TOKEN_INT && token != G_TOKEN_OCTAL)
1997                         {
1998                           value.v_error = token == G_TOKEN_FLOAT ? G_ERR_FLOAT_MALFORMED : G_ERR_FLOAT_RADIX;
1999                           token = G_TOKEN_ERROR;
2000                           in_number = FALSE;
2001                         }
2002                       else
2003                         {
2004                           token = G_TOKEN_FLOAT;
2005                           gstring = g_string_append_c (gstring, ch);
2006                         }
2007                       break;
2008                       
2009                     case '0':
2010                     case '1':
2011                     case '2':
2012                     case '3':
2013                     case '4':
2014                     case '5':
2015                     case '6':
2016                     case '7':
2017                     case '8':
2018                     case '9':
2019                       gstring = g_string_append_c (gstring, ch);
2020                       break;
2021                       
2022                     case '-':
2023                     case '+':
2024                       if (token != G_TOKEN_FLOAT)
2025                         {
2026                           token = G_TOKEN_ERROR;
2027                           value.v_error = G_ERR_NON_DIGIT_IN_CONST;
2028                           in_number = FALSE;
2029                         }
2030                       else
2031                         gstring = g_string_append_c (gstring, ch);
2032                       break;
2033                       
2034                     case 'e':
2035                     case 'E':
2036                       if ((token != G_TOKEN_HEX && !config->scan_float) ||
2037                           (token != G_TOKEN_HEX &&
2038                            token != G_TOKEN_OCTAL &&
2039                            token != G_TOKEN_FLOAT &&
2040                            token != G_TOKEN_INT))
2041                         {
2042                           token = G_TOKEN_ERROR;
2043                           value.v_error = G_ERR_NON_DIGIT_IN_CONST;
2044                           in_number = FALSE;
2045                         }
2046                       else
2047                         {
2048                           if (token != G_TOKEN_HEX)
2049                             token = G_TOKEN_FLOAT;
2050                           gstring = g_string_append_c (gstring, ch);
2051                         }
2052                       break;
2053                       
2054                     default:
2055                       if (token != G_TOKEN_HEX)
2056                         {
2057                           token = G_TOKEN_ERROR;
2058                           value.v_error = G_ERR_NON_DIGIT_IN_CONST;
2059                           in_number = FALSE;
2060                         }
2061                       else
2062                         gstring = g_string_append_c (gstring, ch);
2063                       break;
2064                     }
2065                 }
2066               else
2067                 in_number = FALSE;
2068             }
2069           while (in_number);
2070           
2071           endptr = NULL;
2072           if (token == G_TOKEN_FLOAT)
2073             value.v_float = g_strtod (gstring->str, &endptr);
2074           else
2075             {
2076               guint64 ui64 = 0;
2077               switch (token)
2078                 {
2079                 case G_TOKEN_BINARY:
2080                   ui64 = g_ascii_strtoull (gstring->str, &endptr, 2);
2081                   break;
2082                 case G_TOKEN_OCTAL:
2083                   ui64 = g_ascii_strtoull (gstring->str, &endptr, 8);
2084                   break;
2085                 case G_TOKEN_INT:
2086                   ui64 = g_ascii_strtoull (gstring->str, &endptr, 10);
2087                   break;
2088                 case G_TOKEN_HEX:
2089                   ui64 = g_ascii_strtoull (gstring->str, &endptr, 16);
2090                   break;
2091                 default: ;
2092                 }
2093               if (scanner->config->store_int64)
2094                 value.v_int64 = ui64;
2095               else
2096                 value.v_int = ui64;
2097             }
2098           if (endptr && *endptr)
2099             {
2100               token = G_TOKEN_ERROR;
2101               if (*endptr == 'e' || *endptr == 'E')
2102                 value.v_error = G_ERR_NON_DIGIT_IN_CONST;
2103               else
2104                 value.v_error = G_ERR_DIGIT_RADIX;
2105             }
2106           g_string_free (gstring, TRUE);
2107           gstring = NULL;
2108           ch = 0;
2109         } /* number_parsing:... */
2110         break;
2111         
2112         default:
2113         default_case:
2114         {
2115           if (config->cpair_comment_single &&
2116               ch == config->cpair_comment_single[0])
2117             {
2118               token = G_TOKEN_COMMENT_SINGLE;
2119               in_comment_single = TRUE;
2120               gstring = g_string_new (NULL);
2121               ch = g_scanner_get_char (scanner, line_p, position_p);
2122               while (ch != 0)
2123                 {
2124                   if (ch == config->cpair_comment_single[1])
2125                     {
2126                       in_comment_single = FALSE;
2127                       ch = 0;
2128                       break;
2129                     }
2130                   
2131                   gstring = g_string_append_c (gstring, ch);
2132                   ch = g_scanner_get_char (scanner, line_p, position_p);
2133                 }
2134               /* ignore a missing newline at EOF for single line comments */
2135               if (in_comment_single &&
2136                   config->cpair_comment_single[1] == '\n')
2137                 in_comment_single = FALSE;
2138             }
2139           else if (config->scan_identifier && ch &&
2140                    strchr (config->cset_identifier_first, ch))
2141             {
2142             identifier_precedence:
2143               
2144               if (config->cset_identifier_nth && ch &&
2145                   strchr (config->cset_identifier_nth,
2146                           g_scanner_peek_next_char (scanner)))
2147                 {
2148                   token = G_TOKEN_IDENTIFIER;
2149                   gstring = g_string_new (NULL);
2150                   gstring = g_string_append_c (gstring, ch);
2151                   do
2152                     {
2153                       ch = g_scanner_get_char (scanner, line_p, position_p);
2154                       gstring = g_string_append_c (gstring, ch);
2155                       ch = g_scanner_peek_next_char (scanner);
2156                     }
2157                   while (ch && strchr (config->cset_identifier_nth, ch));
2158                   ch = 0;
2159                 }
2160               else if (config->scan_identifier_1char)
2161                 {
2162                   token = G_TOKEN_IDENTIFIER;
2163                   value.v_identifier = g_new0 (gchar, 2);
2164                   value.v_identifier[0] = ch;
2165                   ch = 0;
2166                 }
2167             }
2168           if (ch)
2169             {
2170               if (config->char_2_token)
2171                 token = ch;
2172               else
2173                 {
2174                   token = G_TOKEN_CHAR;
2175                   value.v_char = ch;
2176                 }
2177               ch = 0;
2178             }
2179         } /* default_case:... */
2180         break;
2181         }
2182       g_assert (ch == 0 && token != G_TOKEN_NONE); /* paranoid */
2183     }
2184   while (ch != 0);
2185   
2186   if (in_comment_multi || in_comment_single ||
2187       in_string_sq || in_string_dq)
2188     {
2189       token = G_TOKEN_ERROR;
2190       if (gstring)
2191         {
2192           g_string_free (gstring, TRUE);
2193           gstring = NULL;
2194         }
2195       (*position_p)++;
2196       if (in_comment_multi || in_comment_single)
2197         value.v_error = G_ERR_UNEXP_EOF_IN_COMMENT;
2198       else /* (in_string_sq || in_string_dq) */
2199         value.v_error = G_ERR_UNEXP_EOF_IN_STRING;
2200     }
2201   
2202   if (gstring)
2203     {
2204       value.v_string = g_string_free (gstring, FALSE);
2205       gstring = NULL;
2206     }
2207   
2208   if (token == G_TOKEN_IDENTIFIER)
2209     {
2210       if (config->scan_symbols)
2211         {
2212           GScannerKey *key;
2213           guint scope_id;
2214           
2215           scope_id = scanner->scope_id;
2216           key = g_scanner_lookup_internal (scanner, scope_id, value.v_identifier);
2217           if (!key && scope_id && scanner->config->scope_0_fallback)
2218             key = g_scanner_lookup_internal (scanner, 0, value.v_identifier);
2219           
2220           if (key)
2221             {
2222               g_free (value.v_identifier);
2223               token = G_TOKEN_SYMBOL;
2224               value.v_symbol = key->value;
2225             }
2226         }
2227       
2228       if (token == G_TOKEN_IDENTIFIER &&
2229           config->scan_identifier_NULL &&
2230           strlen (value.v_identifier) == 4)
2231         {
2232           gchar *null_upper = "NULL";
2233           gchar *null_lower = "null";
2234           
2235           if (scanner->config->case_sensitive)
2236             {
2237               if (value.v_identifier[0] == null_upper[0] &&
2238                   value.v_identifier[1] == null_upper[1] &&
2239                   value.v_identifier[2] == null_upper[2] &&
2240                   value.v_identifier[3] == null_upper[3])
2241                 token = G_TOKEN_IDENTIFIER_NULL;
2242             }
2243           else
2244             {
2245               if ((value.v_identifier[0] == null_upper[0] ||
2246                    value.v_identifier[0] == null_lower[0]) &&
2247                   (value.v_identifier[1] == null_upper[1] ||
2248                    value.v_identifier[1] == null_lower[1]) &&
2249                   (value.v_identifier[2] == null_upper[2] ||
2250                    value.v_identifier[2] == null_lower[2]) &&
2251                   (value.v_identifier[3] == null_upper[3] ||
2252                    value.v_identifier[3] == null_lower[3]))
2253                 token = G_TOKEN_IDENTIFIER_NULL;
2254             }
2255         }
2256     }
2257   
2258   *token_p = token;
2259   *value_p = value;
2260 }