1 /* PSPP - a program for statistical analysis.
2 Copyright (C) 1997-9, 2000, 2005, 2009, 2010, 2011, 2012, 2013 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/>. */
18 This file is concerned with the definition of the PSPP syntax, NOT the
19 action of scanning/parsing code .
24 #include "data/identifier.h"
30 #include "libpspp/assertion.h"
31 #include "libpspp/cast.h"
33 #include "gl/c-ctype.h"
36 #define _(msgid) gettext (msgid)
40 /* Returns TYPE as a string, e.g. "ID" for T_ID. */
42 token_type_to_name (enum token_type type)
46 #define TOKEN_TYPE(TYPE) case T_##TYPE: return #TYPE;
51 return "unknown token type";
55 /* Returns an ASCII string that yields TOKEN if it appeared in a syntax file,
56 as a statically allocated constant string. This function returns NULL for
57 tokens that don't have any fixed string representation, such as identifier
60 token_type_to_string (enum token_type token)
153 /* Recognizing identifiers. */
156 is_ascii_id1 (unsigned char c)
158 return c_isalpha (c) || c == '@' || c == '#' || c == '$';
162 is_ascii_idn (unsigned char c)
164 return is_ascii_id1 (c) || isdigit (c) || c == '.' || c == '_';
167 /* Returns true if C may be the first byte in an identifier in the current
170 (PSPP is transitioning to using Unicode internally for syntax, so please
171 use lex_uc_is_id1() instead, if possible.) */
175 return is_ascii_id1 (c) || (unsigned char) c >= 128;
178 /* Returns true if C may be a byte in an identifier other than the first.
180 (PSPP is transitioning to using Unicode internally for syntax, so please
181 use lex_uc_is_idn() instead, if possible.) */
185 return is_ascii_idn (c) || (unsigned char) c >= 128;
188 /* Returns true if Unicode code point UC may be the first character in an
189 identifier in the current locale. */
191 lex_uc_is_id1 (ucs4_t uc)
195 : (uc_is_general_category_withtable (uc,
199 && uc != 0xfffc && uc != 0xfffd));
202 /* Returns true if Unicode code point UC may be a character in an identifier
203 other than the first. */
205 lex_uc_is_idn (ucs4_t uc)
208 ? is_ascii_id1 (uc) || isdigit (uc) || uc == '.' || uc == '_'
209 : (uc_is_general_category_withtable (uc,
214 && uc != 0xfffc && uc != 0xfffd));
217 /* Returns true if Unicode code point UC is a space that separates tokens. */
219 lex_uc_is_space (ucs4_t uc)
221 /* These are all of the Unicode characters in category Zs, Zl, or Zp. */
222 return (uc == ' ' || (uc <= 0x000d && uc >= 0x0009)
224 && (uc == 0xa0 || uc == 0x85 || uc == 0x1680 || uc == 0x180e
225 || (uc >= 0x2000 && uc <= 0x200a)
226 || uc == 0x2028 || uc == 0x2029 || uc == 0x202f
227 || uc == 0x205f || uc == 0x3000)));
231 /* Returns the length of the longest prefix of STRING that forms
232 a valid identifier. Returns zero if STRING does not begin
233 with a valid identifier. */
235 lex_id_get_length (struct substring string)
237 const uint8_t *s = CHAR_CAST (const uint8_t *, string.string);
238 size_t len = string.length;
242 for (ofs = 0; ofs < string.length; ofs += mblen)
246 mblen = u8_mbtouc (&uc, s + ofs, len - ofs);
247 if (!(ofs == 0 ? lex_uc_is_id1 (uc) : lex_uc_is_idn (uc)))
254 /* Comparing identifiers. */
256 /* Returns true if TOKEN is a case-insensitive match for KEYWORD.
258 Keywords match if one of the following is true: KEYWORD and
259 TOKEN are identical, or TOKEN is at least 3 characters long
260 and those characters are identical to KEYWORD. (Letters that
261 differ only in case are considered identical.)
263 KEYWORD must be ASCII, but TOKEN may be ASCII or UTF-8. */
265 lex_id_match (struct substring keyword, struct substring token)
267 return lex_id_match_n (keyword, token, 3);
270 /* Returns true if TOKEN is a case-insensitive match for at least
271 the first N characters of KEYWORD.
273 KEYWORD must be ASCII, but TOKEN may be ASCII or UTF-8. */
275 lex_id_match_n (struct substring keyword, struct substring token, size_t n)
277 size_t token_len = ss_length (token);
278 size_t keyword_len = ss_length (keyword);
280 if (token_len >= n && token_len < keyword_len)
281 return ss_equals_case (ss_head (keyword, token_len), token);
283 return ss_equals_case (keyword, token);
286 /* Table of keywords. */
290 const struct substring identifier;
293 static const struct keyword keywords[] =
295 { T_AND, SS_LITERAL_INITIALIZER ("AND") },
296 { T_OR, SS_LITERAL_INITIALIZER ("OR") },
297 { T_NOT, SS_LITERAL_INITIALIZER ("NOT") },
298 { T_EQ, SS_LITERAL_INITIALIZER ("EQ") },
299 { T_GE, SS_LITERAL_INITIALIZER ("GE") },
300 { T_GT, SS_LITERAL_INITIALIZER ("GT") },
301 { T_LE, SS_LITERAL_INITIALIZER ("LE") },
302 { T_LT, SS_LITERAL_INITIALIZER ("LT") },
303 { T_NE, SS_LITERAL_INITIALIZER ("NE") },
304 { T_ALL, SS_LITERAL_INITIALIZER ("ALL") },
305 { T_BY, SS_LITERAL_INITIALIZER ("BY") },
306 { T_TO, SS_LITERAL_INITIALIZER ("TO") },
307 { T_WITH, SS_LITERAL_INITIALIZER ("WITH") },
309 static const size_t keyword_cnt = sizeof keywords / sizeof *keywords;
311 /* Returns true if TOKEN is representable as a keyword. */
313 lex_is_keyword (enum token_type token)
315 const struct keyword *kw;
316 for (kw = keywords; kw < &keywords[keyword_cnt]; kw++)
317 if (kw->token == token)
322 /* Returns the proper token type, either T_ID or a reserved
323 keyword enum, for ID. */
325 lex_id_to_token (struct substring id)
327 if (ss_length (id) >= 2 && ss_length (id) <= 4)
329 const struct keyword *kw;
330 for (kw = keywords; kw < &keywords[keyword_cnt]; kw++)
331 if (ss_equals_case (kw->identifier, id))