1 /* GLIB - Library of useful routines for C programming
2 * Copyright (C) 1995-1997 Peter Mattis, Spencer Kimball and Josh MacDonald
3 *
4 * SPDX-License-Identifier: LGPL-2.1-or-later
5 *
6 * This library is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
10 *
11 * This library is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with this library; if not, see <http://www.gnu.org/licenses/>.
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 #if !defined (__GLIB_H_INSIDE__) && !defined (GLIB_COMPILATION)
31 #error "Only <glib.h> can be included directly."
32 #endif
33
34 #include <glib/gdataset.h>
35 #include <glib/ghash.h>
36
37 G_BEGIN_DECLS
38
39 typedef struct _GScanner GScanner;
40 typedef struct _GScannerConfig GScannerConfig;
41 typedef union _GTokenValue GTokenValue;
42
43 typedef void (*GScannerMsgFunc) (GScanner *scanner,
44 gchar *message,
45 gboolean error);
46
47 /* GScanner: Flexible lexical scanner for general purpose.
48 */
49
50 /* Character sets */
51 #define G_CSET_A_2_Z "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
52 #define G_CSET_a_2_z "abcdefghijklmnopqrstuvwxyz"
53 #define G_CSET_DIGITS "0123456789"
54 #define G_CSET_LATINC "\300\301\302\303\304\305\306"\
55 "\307\310\311\312\313\314\315\316\317\320"\
56 "\321\322\323\324\325\326"\
57 "\330\331\332\333\334\335\336"
58 #define G_CSET_LATINS "\337\340\341\342\343\344\345\346"\
59 "\347\350\351\352\353\354\355\356\357\360"\
60 "\361\362\363\364\365\366"\
61 "\370\371\372\373\374\375\376\377"
62
63 /* Error types */
64 typedef enum
65 {
66 G_ERR_UNKNOWN,
67 G_ERR_UNEXP_EOF,
68 G_ERR_UNEXP_EOF_IN_STRING,
69 G_ERR_UNEXP_EOF_IN_COMMENT,
70 G_ERR_NON_DIGIT_IN_CONST,
71 G_ERR_DIGIT_RADIX,
72 G_ERR_FLOAT_RADIX,
73 G_ERR_FLOAT_MALFORMED
74 } GErrorType;
75
76 /* Token types */
77 typedef enum
78 {
79 G_TOKEN_EOF = 0,
80
81 G_TOKEN_LEFT_PAREN = '(',
82 G_TOKEN_RIGHT_PAREN = ')',
83 G_TOKEN_LEFT_CURLY = '{',
84 G_TOKEN_RIGHT_CURLY = '}',
85 G_TOKEN_LEFT_BRACE = '[',
86 G_TOKEN_RIGHT_BRACE = ']',
87 G_TOKEN_EQUAL_SIGN = '=',
88 G_TOKEN_COMMA = ',',
89
90 G_TOKEN_NONE = 256,
91
92 G_TOKEN_ERROR,
93
94 G_TOKEN_CHAR,
95 G_TOKEN_BINARY,
96 G_TOKEN_OCTAL,
97 G_TOKEN_INT,
98 G_TOKEN_HEX,
99 G_TOKEN_FLOAT,
100 G_TOKEN_STRING,
101
102 G_TOKEN_SYMBOL,
103 G_TOKEN_IDENTIFIER,
104 G_TOKEN_IDENTIFIER_NULL,
105
106 G_TOKEN_COMMENT_SINGLE,
107 G_TOKEN_COMMENT_MULTI,
108
109 /*< private >*/
110 G_TOKEN_LAST
111 } GTokenType;
112
113 union _GTokenValue
114 {
115 gpointer v_symbol;
116 gchar *v_identifier;
117 gulong v_binary;
118 gulong v_octal;
119 gulong v_int;
120 guint64 v_int64;
121 gdouble v_float;
122 gulong v_hex;
123 gchar *v_string;
124 gchar *v_comment;
125 guchar v_char;
126 guint v_error;
127 };
128
129 struct _GScannerConfig
130 {
131 /* Character sets
132 */
133 gchar *cset_skip_characters; /* default: " \t\n" */
134 gchar *cset_identifier_first;
135 gchar *cset_identifier_nth;
136 gchar *cpair_comment_single; /* default: "#\n" */
137
138 /* Should symbol lookup work case sensitive?
139 */
140 guint case_sensitive : 1;
141
142 /* Boolean values to be adjusted "on the fly"
143 * to configure scanning behaviour.
144 */
145 guint skip_comment_multi : 1; /* C like comment */
146 guint skip_comment_single : 1; /* single line comment */
147 guint scan_comment_multi : 1; /* scan multi line comments? */
148 guint scan_identifier : 1;
149 guint scan_identifier_1char : 1;
150 guint scan_identifier_NULL : 1;
151 guint scan_symbols : 1;
152 guint scan_binary : 1;
153 guint scan_octal : 1;
154 guint scan_float : 1;
155 guint scan_hex : 1; /* '0x0ff0' */
156 guint scan_hex_dollar : 1; /* '$0ff0' */
157 guint scan_string_sq : 1; /* string: 'anything' */
158 guint scan_string_dq : 1; /* string: "\\-escapes!\n" */
159 guint numbers_2_int : 1; /* bin, octal, hex => int */
160 guint int_2_float : 1; /* int => G_TOKEN_FLOAT? */
161 guint identifier_2_string : 1;
162 guint char_2_token : 1; /* return G_TOKEN_CHAR? */
163 guint symbol_2_token : 1;
164 guint scope_0_fallback : 1; /* try scope 0 on lookups? */
165 guint store_int64 : 1; /* use value.v_int64 rather than v_int */
166
167 /*< private >*/
168 guint padding_dummy;
169 };
170
171 struct _GScanner
172 {
173 /* unused fields */
174 gpointer user_data;
175 guint max_parse_errors;
176
177 /* g_scanner_error() increments this field */
178 guint parse_errors;
179
180 /* name of input stream, featured by the default message handler */
181 const gchar *input_name;
182
183 /* quarked data */
184 GData *qdata;
185
186 /* link into the scanner configuration */
187 GScannerConfig *config;
188
189 /* fields filled in after g_scanner_get_next_token() */
190 GTokenType token;
191 GTokenValue value;
192 guint line;
193 guint position;
194
195 /* fields filled in after g_scanner_peek_next_token() */
196 GTokenType next_token;
197 GTokenValue next_value;
198 guint next_line;
199 guint next_position;
200
201 /*< private >*/
202 /* to be considered private */
203 GHashTable *symbol_table;
204 gint input_fd;
205 const gchar *text;
206 const gchar *text_end;
207 gchar *buffer;
208 guint scope_id;
209
210 /*< public >*/
211 /* handler function for _warn and _error */
212 GScannerMsgFunc msg_handler;
213 };
214
215 GLIB_AVAILABLE_IN_ALL
216 GScanner* g_scanner_new (const GScannerConfig *config_templ);
217 GLIB_AVAILABLE_IN_ALL
218 void g_scanner_destroy (GScanner *scanner);
219 GLIB_AVAILABLE_IN_ALL
220 void g_scanner_input_file (GScanner *scanner,
221 gint input_fd);
222 GLIB_AVAILABLE_IN_ALL
223 void g_scanner_sync_file_offset (GScanner *scanner);
224 GLIB_AVAILABLE_IN_ALL
225 void g_scanner_input_text (GScanner *scanner,
226 const gchar *text,
227 guint text_len);
228 GLIB_AVAILABLE_IN_ALL
229 GTokenType g_scanner_get_next_token (GScanner *scanner);
230 GLIB_AVAILABLE_IN_ALL
231 GTokenType g_scanner_peek_next_token (GScanner *scanner);
232 GLIB_AVAILABLE_IN_ALL
233 GTokenType g_scanner_cur_token (GScanner *scanner);
234 GLIB_AVAILABLE_IN_ALL
235 GTokenValue g_scanner_cur_value (GScanner *scanner);
236 GLIB_AVAILABLE_IN_ALL
237 guint g_scanner_cur_line (GScanner *scanner);
238 GLIB_AVAILABLE_IN_ALL
239 guint g_scanner_cur_position (GScanner *scanner);
240 GLIB_AVAILABLE_IN_ALL
241 gboolean g_scanner_eof (GScanner *scanner);
242 GLIB_AVAILABLE_IN_ALL
243 guint g_scanner_set_scope (GScanner *scanner,
244 guint scope_id);
245 GLIB_AVAILABLE_IN_ALL
246 void g_scanner_scope_add_symbol (GScanner *scanner,
247 guint scope_id,
248 const gchar *symbol,
249 gpointer value);
250 GLIB_AVAILABLE_IN_ALL
251 void g_scanner_scope_remove_symbol (GScanner *scanner,
252 guint scope_id,
253 const gchar *symbol);
254 GLIB_AVAILABLE_IN_ALL
255 gpointer g_scanner_scope_lookup_symbol (GScanner *scanner,
256 guint scope_id,
257 const gchar *symbol);
258 GLIB_AVAILABLE_IN_ALL
259 void g_scanner_scope_foreach_symbol (GScanner *scanner,
260 guint scope_id,
261 GHFunc func,
262 gpointer user_data);
263 GLIB_AVAILABLE_IN_ALL
264 gpointer g_scanner_lookup_symbol (GScanner *scanner,
265 const gchar *symbol);
266 GLIB_AVAILABLE_IN_ALL
267 void g_scanner_unexp_token (GScanner *scanner,
268 GTokenType expected_token,
269 const gchar *identifier_spec,
270 const gchar *symbol_spec,
271 const gchar *symbol_name,
272 const gchar *message,
273 gint is_error);
274 GLIB_AVAILABLE_IN_ALL
275 void g_scanner_error (GScanner *scanner,
276 const gchar *format,
277 ...) G_GNUC_PRINTF (2,3);
278 GLIB_AVAILABLE_IN_ALL
279 void g_scanner_warn (GScanner *scanner,
280 const gchar *format,
281 ...) G_GNUC_PRINTF (2,3);
282
283 /* keep downward source compatibility */
284 #define g_scanner_add_symbol( scanner, symbol, value ) G_STMT_START { \
285 g_scanner_scope_add_symbol ((scanner), 0, (symbol), (value)); \
286 } G_STMT_END GLIB_DEPRECATED_MACRO_IN_2_26_FOR(g_scanner_scope_add_symbol)
287 #define g_scanner_remove_symbol( scanner, symbol ) G_STMT_START { \
288 g_scanner_scope_remove_symbol ((scanner), 0, (symbol)); \
289 } G_STMT_END GLIB_DEPRECATED_MACRO_IN_2_26_FOR(g_scanner_scope_remove_symbol)
290 #define g_scanner_foreach_symbol( scanner, func, data ) G_STMT_START { \
291 g_scanner_scope_foreach_symbol ((scanner), 0, (func), (data)); \
292 } G_STMT_END GLIB_DEPRECATED_MACRO_IN_2_26_FOR(g_scanner_scope_foreach_symbol)
293
294 /* The following two functions are deprecated and will be removed in
295 * the next major release. They do no good. */
296 #define g_scanner_freeze_symbol_table(scanner) ((void)0) GLIB_DEPRECATED_MACRO_IN_2_26
297 #define g_scanner_thaw_symbol_table(scanner) ((void)0) GLIB_DEPRECATED_MACRO_IN_2_26
298
299 G_END_DECLS
300
301 #endif /* __G_SCANNER_H__ */