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_STOP /* Stop processing. */
46 /* Reads a single syntax file as a stream of bytes encoded in UTF-8.
51 const struct lex_reader_class *class;
52 enum segmenter_mode syntax;
53 enum lex_error_mode error;
55 char *file_name; /* NULL if not associated with a file. */
56 int line_number; /* 1-based initial line number, 0 if none. */
60 /* An implementation of a lex_reader. */
61 struct lex_reader_class
63 /* Reads up to N bytes of data from READER into N. Returns the positive
64 number of bytes read if successful, or zero at end of input or on
67 STYLE provides a hint to interactive readers as to what kind of syntax
68 is being read right now. */
69 size_t (*read) (struct lex_reader *reader, char *buf, size_t n,
70 enum prompt_style style);
72 /* Closes and destroys READER, releasing any allocated storage.
74 The caller will free the 'file_name' member of READER, so the
75 implementation should not do so. */
76 void (*destroy) (struct lex_reader *reader);
79 /* Helper functions for lex_reader. */
80 void lex_reader_init (struct lex_reader *, const struct lex_reader_class *);
81 void lex_reader_set_file_name (struct lex_reader *, const char *file_name);
83 /* Creating various kinds of lex_readers. */
84 struct lex_reader *lex_reader_for_file (const char *file_name,
86 enum segmenter_mode syntax,
87 enum lex_error_mode error);
88 struct lex_reader *lex_reader_for_string (const char *, const char *encoding);
89 struct lex_reader *lex_reader_for_format (const char *, const char *, ...)
91 struct lex_reader *lex_reader_for_substring_nocopy (struct substring, const char *encoding);
94 struct lexer *lex_create (void);
95 void lex_destroy (struct lexer *);
98 void lex_define_macro (struct lexer *, struct macro *);
101 void lex_include (struct lexer *, struct lex_reader *);
102 void lex_append (struct lexer *, struct lex_reader *);
105 void lex_get (struct lexer *);
106 void lex_get_n (struct lexer *, size_t n);
108 /* Token testing functions. */
109 bool lex_is_number (const struct lexer *);
110 double lex_number (const struct lexer *);
111 bool lex_is_integer (const struct lexer *);
112 long lex_integer (const struct lexer *);
113 bool lex_is_string (const struct lexer *);
115 /* Token testing functions with lookahead. */
116 bool lex_next_is_number (const struct lexer *, int n);
117 double lex_next_number (const struct lexer *, int n);
118 bool lex_next_is_integer (const struct lexer *, int n);
119 long lex_next_integer (const struct lexer *, int n);
120 bool lex_next_is_string (const struct lexer *, int n);
122 /* Token matching functions. */
123 bool lex_match (struct lexer *, enum token_type);
124 bool lex_match_id (struct lexer *, const char *);
125 bool lex_match_id_n (struct lexer *, const char *, size_t n);
126 bool lex_match_int (struct lexer *, int);
127 bool lex_at_phrase (struct lexer *, const char *s);
128 bool lex_match_phrase (struct lexer *, const char *s);
130 /* Forcible matching functions. */
131 bool lex_force_match (struct lexer *, enum token_type) WARN_UNUSED_RESULT;
132 bool lex_force_match_id (struct lexer *, const char *) WARN_UNUSED_RESULT;
133 bool lex_force_int (struct lexer *) WARN_UNUSED_RESULT;
134 bool lex_force_int_range (struct lexer *, const char *name,
135 long min, long max) WARN_UNUSED_RESULT;
136 bool lex_force_num (struct lexer *) WARN_UNUSED_RESULT;
137 bool lex_force_num_range_closed (struct lexer *, const char *name,
138 double min, double max) WARN_UNUSED_RESULT;
139 bool lex_force_num_range_halfopen (struct lexer *, const char *name,
140 double min, double max) WARN_UNUSED_RESULT;
141 bool lex_force_num_range_open (struct lexer *, const char *name,
142 double min, double max) WARN_UNUSED_RESULT;
143 bool lex_force_id (struct lexer *) WARN_UNUSED_RESULT;
144 bool lex_force_string (struct lexer *) WARN_UNUSED_RESULT;
145 bool lex_force_string_or_id (struct lexer *) WARN_UNUSED_RESULT;
147 /* Token accessors. */
148 enum token_type lex_token (const struct lexer *);
149 double lex_tokval (const struct lexer *);
150 const char *lex_tokcstr (const struct lexer *);
151 struct substring lex_tokss (const struct lexer *);
154 const struct token *lex_next (const struct lexer *, int n);
155 enum token_type lex_next_token (const struct lexer *, int n);
156 const char *lex_next_tokcstr (const struct lexer *, int n);
157 double lex_next_tokval (const struct lexer *, int n);
158 struct substring lex_next_tokss (const struct lexer *, int n);
160 /* Looking at the current command, including lookahead and lookbehind. */
161 int lex_ofs (const struct lexer *);
162 const struct token *lex_ofs_token (const struct lexer *, int ofs);
163 struct msg_location *lex_ofs_location (const struct lexer *, int ofs0, int ofs1);
164 struct msg_point lex_ofs_start_point (const struct lexer *, int ofs);
165 struct msg_point lex_ofs_end_point (const struct lexer *, int ofs);
167 /* Token representation. */
168 char *lex_next_representation (const struct lexer *, int n0, int n1);
169 char *lex_ofs_representation (const struct lexer *, int ofs0, int ofs1);
170 bool lex_next_is_from_macro (const struct lexer *, int n);
172 /* Current position. */
173 const char *lex_get_file_name (const struct lexer *);
174 struct msg_location *lex_get_location (const struct lexer *, int n0, int n1);
175 const char *lex_get_encoding (const struct lexer *);
177 /* Issuing errors. */
178 void lex_error (struct lexer *, const char *, ...) PRINTF_FORMAT (2, 3);
179 void lex_next_error (struct lexer *, int n0, int n1, const char *, ...)
180 PRINTF_FORMAT (4, 5);
181 int lex_end_of_command (struct lexer *);
183 void lex_error_expecting (struct lexer *, ...) SENTINEL(0);
184 #define lex_error_expecting(...) \
185 lex_error_expecting(__VA_ARGS__, NULL_SENTINEL)
186 void lex_error_expecting_valist (struct lexer *, va_list);
187 void lex_error_expecting_array (struct lexer *, const char **, size_t n);
189 void lex_sbc_only_once (const char *);
190 void lex_sbc_missing (const char *);
192 void lex_spec_only_once (struct lexer *, const char *subcommand,
193 const char *specification);
194 void lex_spec_missing (struct lexer *, const char *subcommand,
195 const char *specification);
197 void lex_error_valist (struct lexer *, const char *, va_list)
198 PRINTF_FORMAT (2, 0);
199 void lex_next_error_valist (struct lexer *lexer, int n0, int n1,
200 const char *format, va_list)
201 PRINTF_FORMAT (4, 0);
203 /* Error handling. */
204 enum segmenter_mode lex_get_syntax_mode (const struct lexer *);
205 enum lex_error_mode lex_get_error_mode (const struct lexer *);
206 void lex_discard_rest_of_command (struct lexer *);
207 void lex_interactive_reset (struct lexer *);
208 void lex_discard_noninteractive (struct lexer *);
210 /* Source code access. */
211 void lex_set_message_handler (struct lexer *,
212 void (*output_msg) (const struct msg *,
214 void lex_source_ref (const struct lex_source *);
215 void lex_source_unref (struct lex_source *);
216 struct substring lex_source_get_line (const struct lex_source *, int line);