1 /* PSPP - a program for statistical analysis.
2 Copyright (C) 1997-9, 2000, 2010, 2011, 2013, 2014 Free Software Foundation, Inc.
4 This program is free software: you can redistribute it and/or modify
5 it under the terms of the GNU General Public License as published by
6 the Free Software Foundation, either version 3 of the License, or
7 (at your option) any later version.
9 This program 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
12 GNU General Public License for more details.
14 You should have received a copy of the GNU General Public License
15 along with this program. If not, see <http://www.gnu.org/licenses/>. */
24 #include "data/identifier.h"
25 #include "data/variable.h"
26 #include "language/lexer/segment.h"
27 #include "libpspp/cast.h"
28 #include "libpspp/compiler.h"
29 #include "libpspp/message.h"
30 #include "libpspp/prompt.h"
31 #include "libpspp/str.h"
37 /* Handling of errors. */
40 LEX_ERROR_TERMINAL, /* Discard input line and continue reading. */
41 LEX_ERROR_CONTINUE, /* Continue to next command, except for
42 cascading failures. */
43 LEX_ERROR_IGNORE, /* Continue, even for cascading failures. */
44 LEX_ERROR_STOP, /* Stop processing. */
47 /* Reads a single syntax file as a stream of bytes encoded in UTF-8.
52 const struct lex_reader_class *class;
53 enum segmenter_mode syntax;
54 enum lex_error_mode error;
56 char *file_name; /* NULL if not associated with a file. */
57 int line_number; /* 1-based initial line number, 0 if none. */
61 /* An implementation of a lex_reader. */
62 struct lex_reader_class
64 /* Reads up to N bytes of data from READER into N. Returns the positive
65 number of bytes read if successful, or zero at end of input or on
68 STYLE provides a hint to interactive readers as to what kind of syntax
69 is being read right now. */
70 size_t (*read) (struct lex_reader *reader, char *buf, size_t n,
71 enum prompt_style style);
73 /* Closes and destroys READER, releasing any allocated storage.
75 The caller will free the 'file_name' member of READER, so the
76 implementation should not do so. */
77 void (*destroy) (struct lex_reader *reader);
80 /* Helper functions for lex_reader. */
81 void lex_reader_init (struct lex_reader *, const struct lex_reader_class *);
82 void lex_reader_set_file_name (struct lex_reader *, const char *file_name);
84 /* Creating various kinds of lex_readers. */
85 struct lex_reader *lex_reader_for_file (const char *file_name,
87 enum segmenter_mode syntax,
88 enum lex_error_mode error);
89 struct lex_reader *lex_reader_for_string (const char *, const char *encoding);
90 struct lex_reader *lex_reader_for_format (const char *, const char *, ...)
92 struct lex_reader *lex_reader_for_substring_nocopy (struct substring, const char *encoding);
95 struct lexer *lex_create (void);
96 void lex_destroy (struct lexer *);
99 void lex_define_macro (struct lexer *, struct macro *);
102 void lex_include (struct lexer *, struct lex_reader *);
103 void lex_append (struct lexer *, struct lex_reader *);
106 void lex_get (struct lexer *);
107 void lex_get_n (struct lexer *, size_t n);
109 /* Token testing functions. */
110 bool lex_is_number (const struct lexer *);
111 double lex_number (const struct lexer *);
112 bool lex_is_integer (const struct lexer *);
113 long lex_integer (const struct lexer *);
114 bool lex_is_string (const struct lexer *);
116 /* Token testing functions with lookahead. */
117 bool lex_next_is_number (const struct lexer *, int n);
118 double lex_next_number (const struct lexer *, int n);
119 bool lex_next_is_integer (const struct lexer *, int n);
120 long lex_next_integer (const struct lexer *, int n);
121 bool lex_next_is_string (const struct lexer *, int n);
123 /* Token matching functions. */
124 bool lex_match (struct lexer *, enum token_type);
125 bool lex_match_id (struct lexer *, const char *);
126 bool lex_match_id_n (struct lexer *, const char *, size_t n);
127 bool lex_match_int (struct lexer *, int);
128 bool lex_at_phrase (struct lexer *, const char *s);
129 bool lex_match_phrase (struct lexer *, const char *s);
131 /* Forcible matching functions. */
132 bool lex_force_match (struct lexer *, enum token_type) WARN_UNUSED_RESULT;
133 bool lex_force_match_id (struct lexer *, const char *) WARN_UNUSED_RESULT;
134 bool lex_force_int (struct lexer *) WARN_UNUSED_RESULT;
135 bool lex_force_int_range (struct lexer *, const char *name,
136 long min, long max) WARN_UNUSED_RESULT;
137 bool lex_force_num (struct lexer *) WARN_UNUSED_RESULT;
138 bool lex_force_num_range_closed (struct lexer *, const char *name,
139 double min, double max) WARN_UNUSED_RESULT;
140 bool lex_force_num_range_halfopen (struct lexer *, const char *name,
141 double min, double max) WARN_UNUSED_RESULT;
142 bool lex_force_num_range_open (struct lexer *, const char *name,
143 double min, double max) WARN_UNUSED_RESULT;
144 bool lex_force_id (struct lexer *) WARN_UNUSED_RESULT;
145 bool lex_force_string (struct lexer *) WARN_UNUSED_RESULT;
146 bool lex_force_string_or_id (struct lexer *) WARN_UNUSED_RESULT;
148 /* Token accessors. */
149 enum token_type lex_token (const struct lexer *);
150 double lex_tokval (const struct lexer *);
151 const char *lex_tokcstr (const struct lexer *);
152 struct substring lex_tokss (const struct lexer *);
155 const struct token *lex_next (const struct lexer *, int n);
156 enum token_type lex_next_token (const struct lexer *, int n);
157 const char *lex_next_tokcstr (const struct lexer *, int n);
158 double lex_next_tokval (const struct lexer *, int n);
159 struct substring lex_next_tokss (const struct lexer *, int n);
161 /* Looking at the current command, including lookahead and lookbehind. */
162 int lex_ofs (const struct lexer *);
163 int lex_max_ofs (const struct lexer *);
164 const struct token *lex_ofs_token (const struct lexer *, int ofs);
165 struct msg_location *lex_ofs_location (const struct lexer *, int ofs0, int ofs1);
166 struct msg_point lex_ofs_start_point (const struct lexer *, int ofs);
167 struct msg_point lex_ofs_end_point (const struct lexer *, int ofs);
169 /* Token representation. */
170 char *lex_next_representation (const struct lexer *, int n0, int n1);
171 char *lex_ofs_representation (const struct lexer *, int ofs0, int ofs1);
172 bool lex_next_is_from_macro (const struct lexer *, int n);
174 /* Current position. */
175 const char *lex_get_file_name (const struct lexer *);
176 struct msg_location *lex_get_location (const struct lexer *, int n0, int n1);
177 const char *lex_get_encoding (const struct lexer *);
178 const struct lex_source *lex_source (const struct lexer *);
180 /* Issuing errors and warnings. */
181 void lex_error (struct lexer *, const char *, ...) PRINTF_FORMAT (2, 3);
182 void lex_next_error (struct lexer *, int n0, int n1, const char *, ...)
183 PRINTF_FORMAT (4, 5);
184 void lex_ofs_error (struct lexer *, int ofs0, int ofs1, const char *, ...)
185 PRINTF_FORMAT (4, 5);
187 void lex_msg (struct lexer *, enum msg_class, const char *, ...)
188 PRINTF_FORMAT (3, 4);
189 void lex_next_msg (struct lexer *, enum msg_class, int n0, int n1,
191 PRINTF_FORMAT (5, 6);
192 void lex_ofs_msg (struct lexer *, enum msg_class, int ofs0, int ofs1,
194 PRINTF_FORMAT (5, 6);
195 void lex_ofs_msg_valist (struct lexer *lexer, enum msg_class,
196 int ofs0, int ofs1, const char *format, va_list)
197 PRINTF_FORMAT (5, 0);
199 int lex_end_of_command (struct lexer *);
201 void lex_error_expecting (struct lexer *, ...) SENTINEL(0);
202 #define lex_error_expecting(...) \
203 lex_error_expecting(__VA_ARGS__, NULL_SENTINEL)
204 void lex_error_expecting_valist (struct lexer *, va_list);
205 void lex_error_expecting_array (struct lexer *, const char **, size_t n);
207 void lex_sbc_only_once (struct lexer *, const char *);
208 void lex_sbc_missing (struct lexer *, const char *);
210 void lex_spec_only_once (struct lexer *, const char *subcommand,
211 const char *specification);
212 void lex_spec_missing (struct lexer *, const char *subcommand,
213 const char *specification);
215 /* Error handling. */
216 enum segmenter_mode lex_get_syntax_mode (const struct lexer *);
217 enum lex_error_mode lex_get_error_mode (const struct lexer *);
218 void lex_discard_rest_of_command (struct lexer *);
219 void lex_interactive_reset (struct lexer *);
220 void lex_discard_noninteractive (struct lexer *);
222 /* Source code access. */
223 void lex_set_message_handler (struct lexer *,
224 void (*output_msg) (const struct msg *,
226 struct lex_source *lex_source_ref (const struct lex_source *);
227 void lex_source_unref (struct lex_source *);
228 struct substring lex_source_get_line (const struct lex_source *, int line);