1 /* PSPP - a program for statistical analysis.
2 Copyright (C) 1997-9, 2000, 2006, 2009, 2010 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/>. */
19 #include <libpspp/message.h>
28 #include <libpspp/assertion.h>
29 #include <language/command.h>
30 #include <libpspp/message.h>
31 #include <data/settings.h>
32 #include <libpspp/getl.h>
33 #include <libpspp/str.h>
34 #include <output/journal.h>
35 #include <output/text-item.h>
40 #define _(msgid) gettext (msgid)
41 #define N_(msgid) msgid
50 struct string line_buffer;
52 struct source_stream *ss;
54 int token; /* Current token. */
55 double tokval; /* T_POS_NUM, T_NEG_NUM: the token's value. */
57 char tokid [VAR_NAME_LEN + 1]; /* T_ID: the identifier. */
59 struct string tokstr; /* T_ID, T_STRING: token string value.
60 For T_ID, this is not truncated as is
63 char *prog; /* Pointer to next token in line_buffer. */
64 bool dot; /* True only if this line ends with a terminal dot. */
66 int put_token ; /* If nonzero, next token returned by lex_get().
67 Used only in exceptional circumstances. */
69 struct string put_tokstr;
74 static int parse_id (struct lexer *);
76 /* How a string represents its contents. */
79 CHARACTER_STRING, /* Characters. */
80 BINARY_STRING, /* Binary digits. */
81 OCTAL_STRING, /* Octal digits. */
82 HEX_STRING /* Hexadecimal digits. */
85 static int parse_string (struct lexer *, enum string_type);
88 static void dump_token (struct lexer *);
93 /* Initializes the lexer. */
95 lex_create (struct source_stream *ss)
97 struct lexer *lexer = xzalloc (sizeof (*lexer));
99 ds_init_empty (&lexer->tokstr);
100 ds_init_empty (&lexer->put_tokstr);
101 ds_init_empty (&lexer->line_buffer);
107 struct source_stream *
108 lex_get_source_stream (const struct lexer *lex)
114 lex_current_syntax_mode (const struct lexer *lex)
116 return source_stream_current_syntax_mode (lex->ss);
120 lex_current_error_mode (const struct lexer *lex)
122 return source_stream_current_error_mode (lex->ss);
127 lex_destroy (struct lexer *lexer)
131 ds_destroy (&lexer->put_tokstr);
132 ds_destroy (&lexer->tokstr);
133 ds_destroy (&lexer->line_buffer);
140 /* Common functions. */
142 /* Copies put_token, lexer->put_tokstr, put_tokval into token, tokstr,
143 tokval, respectively, and sets tokid appropriately. */
145 restore_token (struct lexer *lexer)
147 assert (lexer->put_token != 0);
148 lexer->token = lexer->put_token;
149 ds_assign_string (&lexer->tokstr, &lexer->put_tokstr);
150 str_copy_trunc (lexer->tokid, sizeof lexer->tokid, ds_cstr (&lexer->tokstr));
151 lexer->tokval = lexer->put_tokval;
152 lexer->put_token = 0;
155 /* Copies token, tokstr, lexer->tokval into lexer->put_token, put_tokstr,
156 put_lexer->tokval respectively. */
158 save_token (struct lexer *lexer)
160 lexer->put_token = lexer->token;
161 ds_assign_string (&lexer->put_tokstr, &lexer->tokstr);
162 lexer->put_tokval = lexer->tokval;
165 /* Parses a single token, setting appropriate global variables to
166 indicate the token's attributes. */
168 lex_get (struct lexer *lexer)
173 if (NULL == lexer->prog && ! lex_get_line (lexer) )
175 lexer->token = T_STOP;
179 /* If a token was pushed ahead, return it. */
180 if (lexer->put_token)
182 restore_token (lexer);
191 /* Skip whitespace. */
192 while (c_isspace ((unsigned char) *lexer->prog))
207 else if (!lex_get_line (lexer))
210 lexer->token = T_STOP;
217 if (lexer->put_token)
219 restore_token (lexer);
228 /* Actually parse the token. */
229 ds_clear (&lexer->tokstr);
231 switch (*lexer->prog)
234 case '0': case '1': case '2': case '3': case '4':
235 case '5': case '6': case '7': case '8': case '9':
239 /* `-' can introduce a negative number, or it can be a
240 token by itself. If it is not followed by a digit or a
241 decimal point, it is definitely not a number.
242 Otherwise, it might be either, but most of the time we
243 want it as a number. When the syntax calls for a `-'
244 token, lex_negative_to_dash() must be used to break
245 negative numbers into two tokens. */
246 if (*lexer->prog == '-')
248 ds_put_char (&lexer->tokstr, *lexer->prog++);
249 while (c_isspace ((unsigned char) *lexer->prog))
252 if (!c_isdigit ((unsigned char) *lexer->prog) && *lexer->prog != '.')
257 lexer->token = T_NEG_NUM;
260 lexer->token = T_POS_NUM;
262 /* Parse the number, copying it into tokstr. */
263 while (c_isdigit ((unsigned char) *lexer->prog))
264 ds_put_char (&lexer->tokstr, *lexer->prog++);
265 if (*lexer->prog == '.')
267 ds_put_char (&lexer->tokstr, *lexer->prog++);
268 while (c_isdigit ((unsigned char) *lexer->prog))
269 ds_put_char (&lexer->tokstr, *lexer->prog++);
271 if (*lexer->prog == 'e' || *lexer->prog == 'E')
273 ds_put_char (&lexer->tokstr, *lexer->prog++);
274 if (*lexer->prog == '+' || *lexer->prog == '-')
275 ds_put_char (&lexer->tokstr, *lexer->prog++);
276 while (c_isdigit ((unsigned char) *lexer->prog))
277 ds_put_char (&lexer->tokstr, *lexer->prog++);
280 /* Parse as floating point. */
281 lexer->tokval = c_strtod (ds_cstr (&lexer->tokstr), &tail);
284 msg (SE, _("%s does not form a valid number."),
285 ds_cstr (&lexer->tokstr));
288 ds_clear (&lexer->tokstr);
289 ds_put_char (&lexer->tokstr, '0');
296 lexer->token = parse_string (lexer, CHARACTER_STRING);
299 case '(': case ')': case ',': case '=': case '+': case '/':
301 lexer->token = *lexer->prog++;
305 if (*++lexer->prog == '*')
308 lexer->token = T_EXP;
315 if (*++lexer->prog == '=')
320 else if (*lexer->prog == '>')
330 if (*++lexer->prog == '=')
340 if (*++lexer->prog == '=')
346 lexer->token = T_NOT;
351 lexer->token = T_AND;
360 if (lexer->prog[1] == '\'' || lexer->prog[1] == '"')
361 lexer->token = parse_string (lexer, BINARY_STRING);
363 lexer->token = parse_id (lexer);
367 if (lexer->prog[1] == '\'' || lexer->prog[1] == '"')
368 lexer->token = parse_string (lexer, OCTAL_STRING);
370 lexer->token = parse_id (lexer);
374 if (lexer->prog[1] == '\'' || lexer->prog[1] == '"')
375 lexer->token = parse_string (lexer, HEX_STRING);
377 lexer->token = parse_id (lexer);
381 if (lex_is_id1 (*lexer->prog))
383 lexer->token = parse_id (lexer);
388 unsigned char c = *lexer->prog++;
389 char *c_name = xasprintf (c_isgraph (c) ? "%c" : "\\%o", c);
390 msg (SE, _("Bad character in input: `%s'."), c_name);
403 /* Parses an identifier at the current position into tokid and
405 Returns the correct token type. */
407 parse_id (struct lexer *lexer)
409 struct substring rest_of_line
410 = ss_substr (ds_ss (&lexer->line_buffer),
411 ds_pointer_to_position (&lexer->line_buffer, lexer->prog),
413 struct substring id = ss_head (rest_of_line,
414 lex_id_get_length (rest_of_line));
415 lexer->prog += ss_length (id);
417 ds_assign_substring (&lexer->tokstr, id);
418 str_copy_trunc (lexer->tokid, sizeof lexer->tokid, ds_cstr (&lexer->tokstr));
419 return lex_id_to_token (id);
422 /* Reports an error to the effect that subcommand SBC may only be
425 lex_sbc_only_once (const char *sbc)
427 msg (SE, _("Subcommand %s may only be specified once."), sbc);
430 /* Reports an error to the effect that subcommand SBC is
433 lex_sbc_missing (struct lexer *lexer, const char *sbc)
435 lex_error (lexer, _("missing required subcommand %s"), sbc);
438 /* Prints a syntax error message containing the current token and
439 given message MESSAGE (if non-null). */
441 lex_error (struct lexer *lexer, const char *message, ...)
447 if (lexer->token == T_STOP)
448 ds_put_cstr (&s, _("Syntax error at end of file"));
449 else if (lexer->token == '.')
450 ds_put_cstr (&s, _("Syntax error at end of command"));
453 char *token_rep = lex_token_representation (lexer);
454 ds_put_format (&s, _("Syntax error at `%s'"), token_rep);
462 ds_put_cstr (&s, ": ");
464 va_start (args, message);
465 ds_put_vformat (&s, message, args);
469 msg (SE, "%s.", ds_cstr (&s));
473 /* Checks that we're at end of command.
474 If so, returns a successful command completion code.
475 If not, flags a syntax error and returns an error command
478 lex_end_of_command (struct lexer *lexer)
480 if (lexer->token != '.')
482 lex_error (lexer, _("expecting end of command"));
489 /* Token testing functions. */
491 /* Returns true if the current token is a number. */
493 lex_is_number (struct lexer *lexer)
495 return lexer->token == T_POS_NUM || lexer->token == T_NEG_NUM;
499 /* Returns true if the current token is a string. */
501 lex_is_string (struct lexer *lexer)
503 return lexer->token == T_STRING;
507 /* Returns the value of the current token, which must be a
508 floating point number. */
510 lex_number (struct lexer *lexer)
512 assert (lex_is_number (lexer));
513 return lexer->tokval;
516 /* Returns true iff the current token is an integer. */
518 lex_is_integer (struct lexer *lexer)
520 return (lex_is_number (lexer)
521 && lexer->tokval > LONG_MIN
522 && lexer->tokval <= LONG_MAX
523 && floor (lexer->tokval) == lexer->tokval);
526 /* Returns the value of the current token, which must be an
529 lex_integer (struct lexer *lexer)
531 assert (lex_is_integer (lexer));
532 return lexer->tokval;
535 /* Token matching functions. */
537 /* If TOK is the current token, skips it and returns true
538 Otherwise, returns false. */
540 lex_match (struct lexer *lexer, int t)
542 if (lexer->token == t)
551 /* If the current token is the identifier S, skips it and returns
552 true. The identifier may be abbreviated to its first three
554 Otherwise, returns false. */
556 lex_match_id (struct lexer *lexer, const char *s)
558 return lex_match_id_n (lexer, s, 3);
561 /* If the current token is the identifier S, skips it and returns
562 true. The identifier may be abbreviated to its first N
564 Otherwise, returns false. */
566 lex_match_id_n (struct lexer *lexer, const char *s, size_t n)
568 if (lexer->token == T_ID
569 && lex_id_match_n (ss_cstr (s), ss_cstr (lexer->tokid), n))
578 /* If the current token is integer N, skips it and returns true.
579 Otherwise, returns false. */
581 lex_match_int (struct lexer *lexer, int x)
583 if (lex_is_integer (lexer) && lex_integer (lexer) == x)
592 /* Forced matches. */
594 /* If this token is identifier S, fetches the next token and returns
596 Otherwise, reports an error and returns zero. */
598 lex_force_match_id (struct lexer *lexer, const char *s)
600 if (lex_match_id (lexer, s))
604 lex_error (lexer, _("expecting `%s'"), s);
609 /* If the current token is T, skips the token. Otherwise, reports an
610 error and returns from the current function with return value false. */
612 lex_force_match (struct lexer *lexer, int t)
614 if (lexer->token == t)
621 lex_error (lexer, _("expecting `%s'"), lex_token_name (t));
626 /* If this token is a string, does nothing and returns true.
627 Otherwise, reports an error and returns false. */
629 lex_force_string (struct lexer *lexer)
631 if (lexer->token == T_STRING)
635 lex_error (lexer, _("expecting string"));
640 /* If this token is an integer, does nothing and returns true.
641 Otherwise, reports an error and returns false. */
643 lex_force_int (struct lexer *lexer)
645 if (lex_is_integer (lexer))
649 lex_error (lexer, _("expecting integer"));
654 /* If this token is a number, does nothing and returns true.
655 Otherwise, reports an error and returns false. */
657 lex_force_num (struct lexer *lexer)
659 if (lex_is_number (lexer))
662 lex_error (lexer, _("expecting number"));
666 /* If this token is an identifier, does nothing and returns true.
667 Otherwise, reports an error and returns false. */
669 lex_force_id (struct lexer *lexer)
671 if (lexer->token == T_ID)
674 lex_error (lexer, _("expecting identifier"));
678 /* Weird token functions. */
680 /* Returns the first character of the next token, except that if the
681 next token is not an identifier, the character returned will not be
682 a character that can begin an identifier. Specifically, the
683 hexstring lead-in X' causes lookahead() to return '. Note that an
684 alphanumeric return value doesn't guarantee an ID token, it could
685 also be a reserved-word token. */
687 lex_look_ahead (struct lexer *lexer)
689 if (lexer->put_token)
690 return lexer->put_token;
694 if (NULL == lexer->prog && ! lex_get_line (lexer) )
699 while (c_isspace ((unsigned char) *lexer->prog))
706 else if (!lex_get_line (lexer))
709 if (lexer->put_token)
710 return lexer->put_token;
713 if ((toupper ((unsigned char) *lexer->prog) == 'X'
714 || toupper ((unsigned char) *lexer->prog) == 'B'
715 || toupper ((unsigned char) *lexer->prog) == 'O')
716 && (lexer->prog[1] == '\'' || lexer->prog[1] == '"'))
723 /* Makes the current token become the next token to be read; the
724 current token is set to T. */
726 lex_put_back (struct lexer *lexer, int t)
732 /* Makes the current token become the next token to be read; the
733 current token is set to the identifier ID. */
735 lex_put_back_id (struct lexer *lexer, const char *id)
737 assert (lex_id_to_token (ss_cstr (id)) == T_ID);
740 ds_assign_cstr (&lexer->tokstr, id);
741 str_copy_trunc (lexer->tokid, sizeof lexer->tokid, ds_cstr (&lexer->tokstr));
744 /* Weird line processing functions. */
746 /* Returns the entire contents of the current line. */
748 lex_entire_line (const struct lexer *lexer)
750 return ds_cstr (&lexer->line_buffer);
753 const struct string *
754 lex_entire_line_ds (const struct lexer *lexer)
756 return &lexer->line_buffer;
759 /* As lex_entire_line(), but only returns the part of the current line
760 that hasn't already been tokenized. */
762 lex_rest_of_line (const struct lexer *lexer)
767 /* Returns true if the current line ends in a terminal dot,
770 lex_end_dot (const struct lexer *lexer)
775 /* Causes the rest of the current input line to be ignored for
776 tokenization purposes. */
778 lex_discard_line (struct lexer *lexer)
780 ds_cstr (&lexer->line_buffer); /* Ensures ds_end points to something valid */
781 lexer->prog = ds_end (&lexer->line_buffer);
783 lexer->put_token = 0;
787 /* Discards the rest of the current command.
788 When we're reading commands from a file, we skip tokens until
789 a terminal dot or EOF.
790 When we're reading commands interactively from the user,
791 that's just discarding the current line, because presumably
792 the user doesn't want to finish typing a command that will be
795 lex_discard_rest_of_command (struct lexer *lexer)
797 if (!getl_is_interactive (lexer->ss))
799 while (lexer->token != T_STOP && lexer->token != '.')
803 lex_discard_line (lexer);
806 /* Weird line reading functions. */
808 /* Remove C-style comments in STRING, begun by slash-star and
809 terminated by star-slash or newline. */
811 strip_comments (struct string *string)
819 for (cp = ds_cstr (string); *cp; )
821 /* If we're not in a comment, check for quote marks. */
826 else if (*cp == '\'' || *cp == '"')
830 /* If we're not inside a quotation, check for comment. */
833 if (cp[0] == '/' && cp[1] == '*')
840 else if (in_comment && cp[0] == '*' && cp[1] == '/')
849 /* Check commenting. */
856 /* Prepares LINE, which is subject to the given SYNTAX rules, for
857 tokenization by stripping comments and determining whether it
858 is the beginning or end of a command and storing into
859 *LINE_STARTS_COMMAND and *LINE_ENDS_COMMAND appropriately. */
861 lex_preprocess_line (struct string *line,
862 enum syntax_mode syntax,
863 bool *line_starts_command,
864 bool *line_ends_command)
866 strip_comments (line);
867 ds_rtrim (line, ss_cstr (CC_SPACES));
868 *line_ends_command = (ds_chomp (line, settings_get_endcmd ())
869 || (ds_is_empty (line) && settings_get_nulline ()));
870 *line_starts_command = false;
871 if (syntax == GETL_BATCH)
873 int first = ds_first (line);
874 *line_starts_command = !c_isspace (first);
875 if (first == '+' || first == '-')
876 *ds_data (line) = ' ';
880 /* Reads a line, without performing any preprocessing. */
882 lex_get_line_raw (struct lexer *lexer)
884 bool ok = getl_read_line (lexer->ss, &lexer->line_buffer);
887 const char *line = ds_cstr (&lexer->line_buffer);
888 text_item_submit (text_item_create (TEXT_ITEM_SYNTAX, line));
893 /* Reads a line for use by the tokenizer, and preprocesses it by
894 removing comments, stripping trailing whitespace and the
895 terminal dot, and removing leading indentors. */
897 lex_get_line (struct lexer *lexer)
899 bool line_starts_command;
901 if (!lex_get_line_raw (lexer))
907 lex_preprocess_line (&lexer->line_buffer,
908 lex_current_syntax_mode (lexer),
909 &line_starts_command, &lexer->dot);
911 if (line_starts_command)
912 lexer->put_token = '.';
914 lexer->prog = ds_cstr (&lexer->line_buffer);
920 /* Returns the name of a token. */
922 lex_token_name (int token)
924 if (lex_is_keyword (token))
925 return lex_id_name (token);
926 else if (token < 256)
928 static char t[256][2];
938 /* Returns an ASCII representation of the current token as a
939 malloc()'d string. */
941 lex_token_representation (struct lexer *lexer)
945 switch (lexer->token)
950 return ds_xstrdup (&lexer->tokstr);
958 for (sp = ds_cstr (&lexer->tokstr); sp < ds_end (&lexer->tokstr); sp++)
959 if (!c_isprint ((unsigned char) *sp))
965 token_rep = xmalloc (2 + ds_length (&lexer->tokstr) * 2 + 1 + 1);
973 for (sp = ds_cstr (&lexer->tokstr); *sp; )
977 *dp++ = (unsigned char) *sp++;
980 for (sp = ds_cstr (&lexer->tokstr); sp < ds_end (&lexer->tokstr); sp++)
982 *dp++ = (((unsigned char) *sp) >> 4)["0123456789ABCDEF"];
983 *dp++ = (((unsigned char) *sp) & 15)["0123456789ABCDEF"];
993 token_rep = xmalloc (1);
998 return xstrdup ("**");
1001 return xstrdup (lex_token_name (lexer->token));
1007 /* Really weird functions. */
1009 /* Most of the time, a `-' is a lead-in to a negative number. But
1010 sometimes it's actually part of the syntax. If a dash can be part
1011 of syntax then this function is called to rip it off of a
1014 lex_negative_to_dash (struct lexer *lexer)
1016 if (lexer->token == T_NEG_NUM)
1018 lexer->token = T_POS_NUM;
1019 lexer->tokval = -lexer->tokval;
1020 ds_assign_substring (&lexer->tokstr, ds_substr (&lexer->tokstr, 1, SIZE_MAX));
1026 /* Skip a COMMENT command. */
1028 lex_skip_comment (struct lexer *lexer)
1032 if (!lex_get_line (lexer))
1034 lexer->put_token = T_STOP;
1039 if (lexer->put_token == '.')
1042 ds_cstr (&lexer->line_buffer); /* Ensures ds_end will point to a valid char */
1043 lexer->prog = ds_end (&lexer->line_buffer);
1049 /* Private functions. */
1051 /* When invoked, tokstr contains a string of binary, octal, or
1052 hex digits, according to TYPE. The string is converted to
1053 characters having the specified values. */
1055 convert_numeric_string_to_char_string (struct lexer *lexer,
1056 enum string_type type)
1058 const char *base_name;
1068 base_name = _("binary");
1073 base_name = _("octal");
1078 base_name = _("hex");
1086 byte_cnt = ds_length (&lexer->tokstr) / chars_per_byte;
1087 if (ds_length (&lexer->tokstr) % chars_per_byte)
1088 msg (SE, _("String of %s digits has %zu characters, which is not a "
1090 base_name, ds_length (&lexer->tokstr), chars_per_byte);
1092 p = ds_cstr (&lexer->tokstr);
1093 for (i = 0; i < byte_cnt; i++)
1099 for (j = 0; j < chars_per_byte; j++, p++)
1103 if (*p >= '0' && *p <= '9')
1107 static const char alpha[] = "abcdef";
1108 const char *q = strchr (alpha, tolower ((unsigned char) *p));
1117 msg (SE, _("`%c' is not a valid %s digit."), *p, base_name);
1119 value = value * base + v;
1122 ds_cstr (&lexer->tokstr)[i] = (unsigned char) value;
1125 ds_truncate (&lexer->tokstr, byte_cnt);
1128 /* Parses a string from the input buffer into tokstr. The input
1129 buffer pointer lexer->prog must point to the initial single or double
1130 quote. TYPE indicates the type of string to be parsed.
1131 Returns token type. */
1133 parse_string (struct lexer *lexer, enum string_type type)
1135 if (type != CHARACTER_STRING)
1138 /* Accumulate the entire string, joining sections indicated by +
1142 /* Single or double quote. */
1143 int c = *lexer->prog++;
1145 /* Accumulate section. */
1148 /* Check end of line. */
1149 if (*lexer->prog == '\0')
1151 msg (SE, _("Unterminated string constant."));
1155 /* Double quote characters to embed them in strings. */
1156 if (*lexer->prog == c)
1158 if (lexer->prog[1] == c)
1164 ds_put_char (&lexer->tokstr, *lexer->prog++);
1168 /* Skip whitespace after final quote mark. */
1169 if (lexer->prog == NULL)
1173 while (c_isspace ((unsigned char) *lexer->prog))
1181 if (!lex_get_line (lexer))
1185 /* Skip plus sign. */
1186 if (*lexer->prog != '+')
1190 /* Skip whitespace after plus sign. */
1191 if (lexer->prog == NULL)
1195 while (c_isspace ((unsigned char) *lexer->prog))
1203 if (!lex_get_line (lexer))
1205 msg (SE, _("Unexpected end of file in string concatenation."));
1210 /* Ensure that a valid string follows. */
1211 if (*lexer->prog != '\'' && *lexer->prog != '"')
1213 msg (SE, _("String expected following `+'."));
1218 /* We come here when we've finished concatenating all the string sections
1219 into one large string. */
1221 if (type != CHARACTER_STRING)
1222 convert_numeric_string_to_char_string (lexer, type);
1228 /* Reads one token from the lexer and writes a textual representation
1229 on stdout for debugging purposes. */
1231 dump_token (struct lexer *lexer)
1237 curln = getl_source_location (lexer->ss);
1238 curfn = getl_source_name (lexer->ss);
1240 fprintf (stderr, "%s:%d\t", curfn, curln);
1243 switch (lexer->token)
1246 fprintf (stderr, "ID\t%s\n", lexer->tokid);
1251 fprintf (stderr, "NUM\t%f\n", lexer->tokval);
1255 fprintf (stderr, "STRING\t`%s'\n", ds_cstr (&lexer->tokstr));
1259 fprintf (stderr, "STOP\n");
1263 fprintf (stderr, "MISC\tEXP\"");
1267 fprintf (stderr, "MISC\tEOF\n");
1271 if (lex_is_keyword (lexer->token))
1272 fprintf (stderr, "KEYWORD\t%s\n", lex_token_name (lexer->token));
1274 fprintf (stderr, "PUNCT\t%c\n", lexer->token);
1278 #endif /* DUMP_TOKENS */
1281 /* Token Accessor Functions */
1284 lex_token (const struct lexer *lexer)
1286 return lexer->token;
1290 lex_tokval (const struct lexer *lexer)
1292 return lexer->tokval;
1296 lex_tokid (const struct lexer *lexer)
1298 return lexer->tokid;
1301 const struct string *
1302 lex_tokstr (const struct lexer *lexer)
1304 return &lexer->tokstr;
1307 /* If the lexer is positioned at the (pseudo)identifier S, which
1308 may contain a hyphen ('-'), skips it and returns true. Each
1309 half of the identifier may be abbreviated to its first three
1311 Otherwise, returns false. */
1313 lex_match_hyphenated_word (struct lexer *lexer, const char *s)
1315 const char *hyphen = strchr (s, '-');
1317 return lex_match_id (lexer, s);
1318 else if (lexer->token != T_ID
1319 || !lex_id_match (ss_buffer (s, hyphen - s), ss_cstr (lexer->tokid))
1320 || lex_look_ahead (lexer) != '-')
1325 lex_force_match (lexer, '-');
1326 lex_force_match_id (lexer, hyphen + 1);