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;
50 return "unknown token type";
54 /* Returns an ASCII string that yields TOKEN if it appeared in a syntax file,
55 as a statically allocated constant string. This function returns NULL for
56 tokens that don't have any fixed string representation, such as identifier
59 token_type_to_string (enum token_type token)
151 /* Recognizing identifiers. */
154 is_ascii_id1 (unsigned char c)
156 return c_isalpha (c) || c == '@' || c == '#' || c == '$';
160 is_ascii_idn (unsigned char c)
162 return is_ascii_id1 (c) || isdigit (c) || c == '.' || c == '_';
165 /* Returns true if C may be the first byte in an identifier in the current
168 (PSPP is transitioning to using Unicode internally for syntax, so please
169 use lex_uc_is_id1() instead, if possible.) */
173 return is_ascii_id1 (c) || (unsigned char) c >= 128;
176 /* Returns true if C may be a byte in an identifier other than the first.
178 (PSPP is transitioning to using Unicode internally for syntax, so please
179 use lex_uc_is_idn() instead, if possible.) */
183 return is_ascii_idn (c) || (unsigned char) c >= 128;
186 /* Returns true if Unicode code point UC may be the first character in an
187 identifier in the current locale. */
189 lex_uc_is_id1 (ucs4_t uc)
193 : (uc_is_general_category_withtable (uc,
197 && uc != 0xfffc && uc != 0xfffd));
200 /* Returns true if Unicode code point UC may be a character in an identifier
201 other than the first. */
203 lex_uc_is_idn (ucs4_t uc)
206 ? is_ascii_id1 (uc) || isdigit (uc) || uc == '.' || uc == '_'
207 : (uc_is_general_category_withtable (uc,
212 && uc != 0xfffc && uc != 0xfffd));
215 /* Returns true if Unicode code point UC is a space that separates tokens. */
217 lex_uc_is_space (ucs4_t uc)
219 /* These are all of the Unicode characters in category Zs, Zl, or Zp. */
220 return (uc == ' ' || (uc <= 0x000d && uc >= 0x0009)
222 && (uc == 0xa0 || uc == 0x85 || uc == 0x1680 || uc == 0x180e
223 || (uc >= 0x2000 && uc <= 0x200a)
224 || uc == 0x2028 || uc == 0x2029 || uc == 0x202f
225 || uc == 0x205f || uc == 0x3000)));
229 /* Returns the length of the longest prefix of STRING that forms
230 a valid identifier. Returns zero if STRING does not begin
231 with a valid identifier. */
233 lex_id_get_length (struct substring string)
235 const uint8_t *s = CHAR_CAST (const uint8_t *, string.string);
236 size_t len = string.length;
240 for (ofs = 0; ofs < string.length; ofs += mblen)
244 mblen = u8_mbtouc (&uc, s + ofs, len - ofs);
245 if (!(ofs == 0 ? lex_uc_is_id1 (uc) : lex_uc_is_idn (uc)))
252 /* Comparing identifiers. */
254 /* Returns true if TOKEN is a case-insensitive match for KEYWORD.
256 Keywords match if one of the following is true: KEYWORD and
257 TOKEN are identical, or TOKEN is at least 3 characters long
258 and those characters are identical to KEYWORD. (Letters that
259 differ only in case are considered identical.)
261 KEYWORD must be ASCII, but TOKEN may be ASCII or UTF-8. */
263 lex_id_match (struct substring keyword, struct substring token)
265 return lex_id_match_n (keyword, token, 3);
268 /* Returns true if TOKEN is a case-insensitive match for at least
269 the first N characters of KEYWORD.
271 KEYWORD must be ASCII, but TOKEN may be ASCII or UTF-8. */
273 lex_id_match_n (struct substring keyword, struct substring token, size_t n)
275 size_t token_len = ss_length (token);
276 size_t keyword_len = ss_length (keyword);
278 if (token_len >= n && token_len < keyword_len)
279 return ss_equals_case (ss_head (keyword, token_len), token);
281 return ss_equals_case (keyword, token);
284 /* Table of keywords. */
288 const struct substring identifier;
291 static const struct keyword keywords[] =
293 { T_AND, SS_LITERAL_INITIALIZER ("AND") },
294 { T_OR, SS_LITERAL_INITIALIZER ("OR") },
295 { T_NOT, SS_LITERAL_INITIALIZER ("NOT") },
296 { T_EQ, SS_LITERAL_INITIALIZER ("EQ") },
297 { T_GE, SS_LITERAL_INITIALIZER ("GE") },
298 { T_GT, SS_LITERAL_INITIALIZER ("GT") },
299 { T_LE, SS_LITERAL_INITIALIZER ("LE") },
300 { T_LT, SS_LITERAL_INITIALIZER ("LT") },
301 { T_NE, SS_LITERAL_INITIALIZER ("NE") },
302 { T_ALL, SS_LITERAL_INITIALIZER ("ALL") },
303 { T_BY, SS_LITERAL_INITIALIZER ("BY") },
304 { T_TO, SS_LITERAL_INITIALIZER ("TO") },
305 { T_WITH, SS_LITERAL_INITIALIZER ("WITH") },
307 static const size_t keyword_cnt = sizeof keywords / sizeof *keywords;
309 /* Returns true if TOKEN is representable as a keyword. */
311 lex_is_keyword (enum token_type token)
313 const struct keyword *kw;
314 for (kw = keywords; kw < &keywords[keyword_cnt]; kw++)
315 if (kw->token == token)
320 /* Returns the proper token type, either T_ID or a reserved
321 keyword enum, for ID. */
323 lex_id_to_token (struct substring id)
325 if (ss_length (id) >= 2 && ss_length (id) <= 4)
327 const struct keyword *kw;
328 for (kw = keywords; kw < &keywords[keyword_cnt]; kw++)
329 if (ss_equals_case (kw->identifier, id))