|
1994
|
1 /* |
|
1
|
2 |
|
2847
|
3 Copyright (C) 1996, 1997 John W. Eaton |
|
1
|
4 |
|
|
5 This file is part of Octave. |
|
|
6 |
|
|
7 Octave is free software; you can redistribute it and/or modify it |
|
|
8 under the terms of the GNU General Public License as published by the |
|
|
9 Free Software Foundation; either version 2, or (at your option) any |
|
|
10 later version. |
|
|
11 |
|
|
12 Octave is distributed in the hope that it will be useful, but WITHOUT |
|
|
13 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
|
|
14 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
|
|
15 for more details. |
|
|
16 |
|
|
17 You should have received a copy of the GNU General Public License |
|
|
18 along with GNU CC; see the file COPYING. If not, write to the Free |
|
1315
|
19 Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. |
|
1
|
20 |
|
|
21 */ |
|
|
22 |
|
4208
|
23 %s COMMAND_START |
|
|
24 %s MATRIX_START |
|
1
|
25 |
|
|
26 %{ |
|
240
|
27 #ifdef HAVE_CONFIG_H |
|
1220
|
28 #include <config.h> |
|
240
|
29 #endif |
|
|
30 |
|
1341
|
31 #include <cctype> |
|
|
32 #include <cstring> |
|
|
33 |
|
1823
|
34 #include <string> |
|
4214
|
35 #include <stack> |
|
1823
|
36 |
|
4093
|
37 #ifdef HAVE_UNISTD_H |
|
|
38 #ifdef HAVE_SYS_TYPES_H |
|
|
39 #include <sys/types.h> |
|
|
40 #endif |
|
|
41 #include <unistd.h> |
|
|
42 #endif |
|
|
43 |
|
2926
|
44 #include "cmd-edit.h" |
|
4153
|
45 #include "quit.h" |
|
4051
|
46 #include "lo-sstream.h" |
|
2926
|
47 |
|
1497
|
48 // These would be alphabetical, but y.tab.h must be included before |
|
|
49 // oct-gperf.h and y.tab.h must be included after token.h and the tree |
|
|
50 // class declarations. We can't include y.tab.h in oct-gperf.h |
|
|
51 // because it may not be protected to allow it to be included multiple |
|
|
52 // times. |
|
|
53 |
|
3665
|
54 #include "comment-list.h" |
|
2181
|
55 #include "defun.h" |
|
1355
|
56 #include "error.h" |
|
1351
|
57 #include "input.h" |
|
1355
|
58 #include "lex.h" |
|
2891
|
59 #include "ov.h" |
|
1355
|
60 #include "parse.h" |
|
2987
|
61 #include "pt-all.h" |
|
2891
|
62 #include "symtab.h" |
|
|
63 #include "token.h" |
|
|
64 #include "toplev.h" |
|
1355
|
65 #include "utils.h" |
|
|
66 #include "variables.h" |
|
2492
|
67 #include <y.tab.h> |
|
|
68 #include <oct-gperf.h> |
|
1
|
69 |
|
2716
|
70 #if ! (defined (FLEX_SCANNER) \ |
|
|
71 && defined (YY_FLEX_MAJOR_VERSION) && YY_FLEX_MAJOR_VERSION >= 2 \ |
|
|
72 && defined (YY_FLEX_MINOR_VERSION) && YY_FLEX_MINOR_VERSION >= 5) |
|
|
73 #error lex.l requires flex version 2.5.4 or later |
|
|
74 #endif |
|
|
75 |
|
3883
|
76 // TRUE means that we have encountered EOF on the input stream. |
|
|
77 bool parser_end_of_input = false; |
|
|
78 |
|
1826
|
79 // Flags that need to be shared between the lexer and parser. |
|
|
80 lexical_feedback lexer_flags; |
|
|
81 |
|
1351
|
82 // Stack to hold tokens so that we can delete them when the parser is |
|
|
83 // reset and avoid growing forever just because we are stashing some |
|
|
84 // information. This has to appear before lex.h is included, because |
|
|
85 // one of the macros defined there uses token_stack. |
|
2614
|
86 // |
|
|
87 // XXX FIXME XXX -- this should really be static, but that causes |
|
|
88 // problems on some systems. |
|
4214
|
89 std::stack <token*> token_stack; |
|
1351
|
90 |
|
1826
|
91 // Did eat_whitespace() eat a space or tab, or a newline, or both? |
|
1
|
92 |
|
1826
|
93 typedef int yum_yum; |
|
1
|
94 |
|
1826
|
95 const yum_yum ATE_NOTHING = 0; |
|
|
96 const yum_yum ATE_SPACE_OR_TAB = 1; |
|
|
97 const yum_yum ATE_NEWLINE = 2; |
|
1088
|
98 |
|
3351
|
99 // Is the closest nesting level a square bracket, squiggly brace or a paren? |
|
1826
|
100 |
|
4214
|
101 class bracket_brace_paren_nesting_level |
|
1826
|
102 { |
|
|
103 public: |
|
|
104 |
|
4214
|
105 bracket_brace_paren_nesting_level (void) : context () { } |
|
1826
|
106 |
|
3351
|
107 ~bracket_brace_paren_nesting_level (void) { } |
|
|
108 |
|
4214
|
109 void bracket (void) { context.push (BRACKET); } |
|
|
110 bool is_bracket (void) |
|
|
111 { return ! context.empty () && context.top () == BRACKET; } |
|
|
112 |
|
|
113 void brace (void) { context.push (BRACE); } |
|
|
114 bool is_brace (void) |
|
|
115 { return ! context.empty () && context.top () == BRACE; } |
|
|
116 |
|
|
117 void paren (void) { context.push (PAREN); } |
|
|
118 bool is_paren (void) |
|
|
119 { return ! context.empty () && context.top () == PAREN; } |
|
|
120 |
|
|
121 bool none (void) { return context.empty (); } |
|
|
122 |
|
|
123 void remove (void) { if (! context.empty ()) context.pop (); } |
|
|
124 |
|
|
125 void clear (void) { while (! context.empty ()) context.pop (); } |
|
1826
|
126 |
|
|
127 private: |
|
|
128 |
|
4214
|
129 std::stack<int> context; |
|
|
130 |
|
3351
|
131 enum { BRACKET = 1, BRACE = 2, PAREN = 3 }; |
|
1826
|
132 |
|
3351
|
133 bracket_brace_paren_nesting_level (const bracket_brace_paren_nesting_level&); |
|
1826
|
134 |
|
3351
|
135 bracket_brace_paren_nesting_level& |
|
|
136 operator = (const bracket_brace_paren_nesting_level&); |
|
1826
|
137 }; |
|
|
138 |
|
3351
|
139 static bracket_brace_paren_nesting_level nesting_level; |
|
1
|
140 |
|
2167
|
141 // Should whitespace in a literal matrix list be automatically |
|
|
142 // converted to commas and semicolons? |
|
|
143 // |
|
|
144 // user specifies value of var |
|
|
145 // -------------- ------------ |
|
|
146 // "ignore" 2 |
|
|
147 // "traditional" 1 |
|
|
148 // anything else 0 |
|
|
149 // |
|
|
150 // Octave will never insert a comma in a literal matrix list if the |
|
|
151 // user specifies "ignore". For example, the statement [1 2] will |
|
|
152 // result in an error instead of being treated the same as [1, 2], and |
|
|
153 // the statement |
|
|
154 // |
|
|
155 // [ 1, 2, |
|
|
156 // 3, 4 ] |
|
|
157 // |
|
|
158 // will result in the vector [1 2 3 4] instead of a matrix. |
|
|
159 // |
|
|
160 // Traditional behavior makes Octave convert spaces to a comma between |
|
|
161 // identifiers and `('. For example, the statement |
|
|
162 // |
|
|
163 // [eye (2)] |
|
|
164 // |
|
|
165 // will be parsed as |
|
|
166 // |
|
|
167 // [eye, (2)] |
|
|
168 // |
|
|
169 // and will result in an error since the `eye' function will be |
|
|
170 // called with no arguments. To get around this, you would have to |
|
|
171 // omit the space between `eye' and the `('. |
|
|
172 // |
|
|
173 // The default value is 0, which results in behavior that is the same |
|
|
174 // as traditional, except that Octave does not convert spaces to a |
|
|
175 // comma between identifiers and `('. For example, the statement |
|
|
176 // |
|
|
177 // [eye (2)] |
|
|
178 // |
|
|
179 // will result in a call to `eye' with the argument `2'. |
|
|
180 |
|
|
181 static int Vwhitespace_in_literal_matrix; |
|
|
182 |
|
4037
|
183 static bool Vwarn_matlab_incompatible = false; |
|
|
184 |
|
3388
|
185 static bool Vwarn_separator_insert = false; |
|
|
186 |
|
3400
|
187 static bool Vwarn_single_quote_string = false; |
|
|
188 |
|
146
|
189 // Forward declarations for functions defined at the bottom of this |
|
|
190 // file. |
|
|
191 |
|
1
|
192 static void fixup_column_count (char *s); |
|
146
|
193 static void do_comma_insert_check (void); |
|
3523
|
194 static int is_plot_keyword (const std::string& s); |
|
|
195 static int is_keyword (const std::string& s); |
|
|
196 static std::string plot_style_token (const std::string& s); |
|
|
197 static symbol_record *lookup_identifier (const std::string& s); |
|
1
|
198 static void grab_help_text (void); |
|
2857
|
199 static bool match_any (char c, const char *s); |
|
3263
|
200 static bool next_token_is_sep_op (void); |
|
3246
|
201 static bool next_token_is_bin_op (bool spc_prev); |
|
|
202 static bool next_token_is_postfix_unary_op (bool spc_prev); |
|
3523
|
203 static std::string strip_trailing_whitespace (char *s); |
|
3246
|
204 static void handle_number (void); |
|
975
|
205 static int handle_string (char delim, int text_style = 0); |
|
3351
|
206 static int handle_close_bracket (int spc_gobbled); |
|
3974
|
207 static int handle_identifier (void); |
|
3096
|
208 static bool have_continuation (bool trailing_comments_ok = true); |
|
|
209 static bool have_ellipsis_continuation (bool trailing_comments_ok = true); |
|
3665
|
210 static void scan_for_comments (const char *); |
|
1826
|
211 static yum_yum eat_whitespace (void); |
|
|
212 static yum_yum eat_continuation (void); |
|
3388
|
213 static void maybe_warn_separator_insert (char sep); |
|
3400
|
214 static void gripe_single_quote_string (void); |
|
4037
|
215 static void gripe_matlab_incompatible (const std::string& msg); |
|
|
216 static void maybe_gripe_matlab_incompatible_comment (char c); |
|
|
217 static void gripe_matlab_incompatible_continuation (void); |
|
|
218 static void gripe_matlab_incompatible_operator (const std::string& op); |
|
1
|
219 |
|
|
220 %} |
|
|
221 |
|
|
222 D [0-9] |
|
|
223 S [ \t] |
|
2042
|
224 NL ((\n)|(\r\n)) |
|
|
225 SNL ({S}|{NL}) |
|
1
|
226 EL (\.\.\.) |
|
967
|
227 BS (\\) |
|
|
228 CONT ({EL}|{BS}) |
|
1
|
229 Im [iIjJ] |
|
967
|
230 CCHAR [#%] |
|
|
231 COMMENT ({CCHAR}.*{NL}) |
|
|
232 SNLCMT ({SNL}|{COMMENT}) |
|
|
233 NOT ((\~)|(\!)) |
|
4037
|
234 POW ((\*\*)|(\^)) |
|
|
235 EPOW (\.{POW}) |
|
1
|
236 IDENT ([_a-zA-Z][_a-zA-Z0-9]*) |
|
|
237 EXPON ([DdEe][+-]?{D}+) |
|
3220
|
238 NUMBER (({D}+\.?{D}*{EXPON}?)|(\.{D}+{EXPON}?)|(0[xX][0-9a-fA-F]+)) |
|
1
|
239 %% |
|
|
240 |
|
968
|
241 %{ |
|
4208
|
242 // Help and other command-style functions are a pain in the ass. This |
|
968
|
243 // stuff needs to be simplified. May require some changes in the |
|
|
244 // parser too. |
|
|
245 %} |
|
|
246 |
|
4208
|
247 <COMMAND_START>{NL} { |
|
967
|
248 BEGIN 0; |
|
|
249 current_input_column = 1; |
|
2857
|
250 lexer_flags.quote_is_transpose = false; |
|
|
251 lexer_flags.cant_be_identifier = false; |
|
|
252 lexer_flags.convert_spaces_to_comma = true; |
|
967
|
253 return '\n'; |
|
|
254 } |
|
1
|
255 |
|
4208
|
256 <COMMAND_START>[\;\,] { |
|
1826
|
257 if (lexer_flags.doing_set && strcmp (yytext, ",") == 0) |
|
967
|
258 { |
|
4064
|
259 TOK_PUSH_AND_RETURN (yytext, STRING); |
|
967
|
260 } |
|
|
261 else |
|
|
262 { |
|
|
263 BEGIN 0; |
|
|
264 if (strcmp (yytext, ",") == 0) |
|
|
265 TOK_RETURN (','); |
|
|
266 else |
|
|
267 TOK_RETURN (';'); |
|
|
268 } |
|
|
269 } |
|
1
|
270 |
|
4208
|
271 <COMMAND_START>[\"\'] { |
|
975
|
272 current_input_column++; |
|
3246
|
273 return handle_string (yytext[0], true); |
|
975
|
274 } |
|
|
275 |
|
4208
|
276 <COMMAND_START>[^#% \t\n\;\,\"\'][^ \t\n\;\,]*{S}* { |
|
3523
|
277 std::string tok = strip_trailing_whitespace (yytext); |
|
4064
|
278 TOK_PUSH_AND_RETURN (tok, STRING); |
|
967
|
279 } |
|
1
|
280 |
|
968
|
281 %{ |
|
1
|
282 // For this and the next two rules, we're looking at ']', and we |
|
971
|
283 // need to know if the next token is `=' or `=='. |
|
1
|
284 // |
|
|
285 // It would have been so much easier if the delimiters were simply |
|
|
286 // different for the expression on the left hand side of the equals |
|
|
287 // operator. |
|
971
|
288 // |
|
|
289 // It's also a pain in the ass to decide whether to insert a comma |
|
|
290 // after seeing a ']' character... |
|
968
|
291 %} |
|
|
292 |
|
4208
|
293 <MATRIX_START>{SNLCMT}*\]{S}* { |
|
3665
|
294 scan_for_comments (yytext); |
|
1001
|
295 fixup_column_count (yytext); |
|
|
296 int c = yytext[yyleng-1]; |
|
|
297 int cont_is_spc = eat_continuation (); |
|
|
298 int spc_gobbled = (cont_is_spc || c == ' ' || c == '\t'); |
|
3351
|
299 return handle_close_bracket (spc_gobbled); |
|
967
|
300 } |
|
1
|
301 |
|
968
|
302 %{ |
|
1088
|
303 // Commas are element separators in matrix constants. If we don't |
|
|
304 // check for continuations here we can end up inserting too many |
|
|
305 // commas. |
|
968
|
306 %} |
|
|
307 |
|
4208
|
308 <MATRIX_START>{S}*\,{S}* { |
|
1088
|
309 current_input_column += yyleng; |
|
3388
|
310 |
|
1088
|
311 int tmp = eat_continuation (); |
|
3388
|
312 |
|
2857
|
313 lexer_flags.quote_is_transpose = false; |
|
|
314 lexer_flags.cant_be_identifier = false; |
|
|
315 lexer_flags.convert_spaces_to_comma = true; |
|
3388
|
316 |
|
|
317 if ((tmp & ATE_NEWLINE) == ATE_NEWLINE) |
|
|
318 { |
|
|
319 maybe_warn_separator_insert (';'); |
|
|
320 |
|
|
321 if (Vwhitespace_in_literal_matrix != 2) |
|
|
322 unput (';'); |
|
|
323 } |
|
|
324 |
|
1088
|
325 return (','); |
|
967
|
326 } |
|
1
|
327 |
|
968
|
328 %{ |
|
|
329 // In some cases, spaces in matrix constants can turn into commas. |
|
|
330 // If commas are required, spaces are not important in matrix |
|
1088
|
331 // constants so we just eat them. If we don't check for continuations |
|
|
332 // here we can end up inserting too many commas. |
|
968
|
333 %} |
|
430
|
334 |
|
4208
|
335 <MATRIX_START>{S}+ { |
|
1088
|
336 current_input_column += yyleng; |
|
3388
|
337 |
|
|
338 int tmp = eat_continuation (); |
|
|
339 int bin_op = next_token_is_bin_op (true); |
|
|
340 int postfix_un_op = next_token_is_postfix_unary_op (true); |
|
|
341 |
|
|
342 if (! (postfix_un_op || bin_op) |
|
|
343 && nesting_level.is_bracket () |
|
|
344 && lexer_flags.convert_spaces_to_comma) |
|
967
|
345 { |
|
3388
|
346 if ((tmp & ATE_NEWLINE) == ATE_NEWLINE) |
|
|
347 { |
|
|
348 maybe_warn_separator_insert (';'); |
|
967
|
349 |
|
3388
|
350 if (Vwhitespace_in_literal_matrix != 2) |
|
|
351 unput (';'); |
|
|
352 } |
|
|
353 |
|
|
354 if (Vwhitespace_in_literal_matrix != 2) |
|
1088
|
355 { |
|
2857
|
356 lexer_flags.quote_is_transpose = false; |
|
|
357 lexer_flags.cant_be_identifier = false; |
|
|
358 lexer_flags.convert_spaces_to_comma = true; |
|
3388
|
359 |
|
|
360 maybe_warn_separator_insert (','); |
|
|
361 |
|
1088
|
362 return (','); |
|
|
363 } |
|
967
|
364 } |
|
|
365 } |
|
430
|
366 |
|
968
|
367 %{ |
|
1088
|
368 // Semicolons are handled as row seprators in matrix constants. If we |
|
|
369 // don't eat whitespace here we can end up inserting too many |
|
|
370 // semicolons. |
|
968
|
371 %} |
|
|
372 |
|
4208
|
373 <MATRIX_START>{SNLCMT}*;{SNLCMT}* { |
|
3665
|
374 scan_for_comments (yytext); |
|
967
|
375 fixup_column_count (yytext); |
|
1001
|
376 eat_whitespace (); |
|
2857
|
377 lexer_flags.quote_is_transpose = false; |
|
|
378 lexer_flags.cant_be_identifier = false; |
|
|
379 lexer_flags.convert_spaces_to_comma = true; |
|
967
|
380 return ';'; |
|
|
381 } |
|
|
382 |
|
968
|
383 %{ |
|
1088
|
384 // In some cases, new lines can also become row separators. If we |
|
|
385 // don't eat whitespace here we can end up inserting too many |
|
|
386 // semicolons. |
|
985
|
387 %} |
|
|
388 |
|
4208
|
389 <MATRIX_START>{S}*{COMMENT}{SNLCMT}* | |
|
|
390 <MATRIX_START>{S}*{NL}{SNLCMT}* { |
|
3665
|
391 scan_for_comments (yytext); |
|
1082
|
392 fixup_column_count (yytext); |
|
1088
|
393 eat_whitespace (); |
|
3388
|
394 |
|
2167
|
395 if (Vwhitespace_in_literal_matrix != 2) |
|
985
|
396 { |
|
2857
|
397 lexer_flags.quote_is_transpose = false; |
|
|
398 lexer_flags.cant_be_identifier = false; |
|
|
399 lexer_flags.convert_spaces_to_comma = true; |
|
985
|
400 |
|
1826
|
401 if (nesting_level.none ()) |
|
985
|
402 return LEXICAL_ERROR; |
|
3388
|
403 } |
|
985
|
404 |
|
3388
|
405 if (nesting_level.is_bracket ()) |
|
|
406 { |
|
|
407 maybe_warn_separator_insert (';'); |
|
|
408 |
|
|
409 if (Vwhitespace_in_literal_matrix != 2) |
|
985
|
410 return ';'; |
|
|
411 } |
|
|
412 } |
|
|
413 |
|
|
414 %{ |
|
3351
|
415 // Open and close bracket are handled differently if we are in the range |
|
968
|
416 // part of a plot command. |
|
975
|
417 // |
|
968
|
418 %} |
|
1
|
419 |
|
967
|
420 \[{S}* { |
|
3351
|
421 nesting_level.bracket (); |
|
975
|
422 |
|
1082
|
423 current_input_column += yyleng; |
|
2857
|
424 lexer_flags.quote_is_transpose = false; |
|
|
425 lexer_flags.cant_be_identifier = false; |
|
|
426 lexer_flags.convert_spaces_to_comma = true; |
|
975
|
427 |
|
|
428 promptflag--; |
|
|
429 eat_whitespace (); |
|
|
430 |
|
1826
|
431 if (lexer_flags.plotting && ! lexer_flags.past_plot_range) |
|
967
|
432 { |
|
2857
|
433 lexer_flags.in_plot_range = true; |
|
1082
|
434 return OPEN_BRACE; |
|
967
|
435 } |
|
|
436 else |
|
|
437 { |
|
3351
|
438 lexer_flags.bracketflag++; |
|
4208
|
439 BEGIN MATRIX_START; |
|
1082
|
440 return '['; |
|
967
|
441 } |
|
|
442 } |
|
1
|
443 |
|
968
|
444 \] { |
|
1826
|
445 nesting_level.remove (); |
|
968
|
446 |
|
1826
|
447 if (lexer_flags.plotting && ! lexer_flags.past_plot_range) |
|
968
|
448 { |
|
2857
|
449 lexer_flags.in_plot_range = false; |
|
968
|
450 TOK_RETURN (CLOSE_BRACE); |
|
|
451 } |
|
|
452 else |
|
|
453 TOK_RETURN (']'); |
|
|
454 } |
|
|
455 |
|
|
456 %{ |
|
|
457 // Imaginary numbers. |
|
|
458 %} |
|
|
459 |
|
|
460 {NUMBER}{Im} { |
|
3246
|
461 handle_number (); |
|
968
|
462 return IMAG_NUM; |
|
|
463 } |
|
|
464 |
|
|
465 %{ |
|
|
466 // Real numbers. Don't grab the `.' part of a dot operator as part of |
|
|
467 // the constant. |
|
|
468 %} |
|
|
469 |
|
|
470 {D}+/\.[\*/\\^'] | |
|
|
471 {NUMBER} { |
|
3246
|
472 handle_number (); |
|
968
|
473 return NUM; |
|
|
474 } |
|
|
475 |
|
|
476 %{ |
|
|
477 // Eat whitespace. Whitespace inside matrix constants is handled by |
|
4208
|
478 // the <MATRIX_START> start state code above. |
|
968
|
479 %} |
|
|
480 |
|
967
|
481 {S}* { |
|
|
482 current_input_column += yyleng; |
|
|
483 } |
|
|
484 |
|
968
|
485 %{ |
|
|
486 // Continuation lines. Allow comments after continuations. |
|
|
487 %} |
|
|
488 |
|
967
|
489 {CONT}{S}*{NL} | |
|
|
490 {CONT}{S}*{COMMENT} { |
|
4037
|
491 if (yytext[0] == '\\') |
|
|
492 gripe_matlab_incompatible_continuation (); |
|
3665
|
493 scan_for_comments (yytext); |
|
967
|
494 promptflag--; |
|
|
495 current_input_column = 1; |
|
|
496 } |
|
1
|
497 |
|
968
|
498 %{ |
|
|
499 // An ellipsis not at the end of a line is not a continuation, but |
|
|
500 // does have another meaning. |
|
|
501 %} |
|
|
502 |
|
967
|
503 {EL} { |
|
3973
|
504 if (lexer_flags.looking_at_parameter_list) |
|
3974
|
505 { |
|
|
506 warning ("`...' is deprecated; use varargin instead"); |
|
|
507 return VARARGIN; |
|
|
508 } |
|
3973
|
509 else if (lexer_flags.looking_at_return_list) |
|
3974
|
510 { |
|
|
511 warning ("`...' is deprecated; use varargout instead"); |
|
|
512 return VARARGOUT; |
|
|
513 } |
|
3973
|
514 else |
|
|
515 return LEXICAL_ERROR; |
|
967
|
516 } |
|
1
|
517 |
|
968
|
518 %{ |
|
|
519 // End of file. |
|
|
520 %} |
|
|
521 |
|
967
|
522 <<EOF>> { |
|
|
523 TOK_RETURN (END_OF_INPUT); |
|
|
524 } |
|
1
|
525 |
|
968
|
526 %{ |
|
970
|
527 // Identifiers. Truncate the token at the first space or tab but |
|
|
528 // don't write directly on yytext. |
|
968
|
529 %} |
|
|
530 |
|
967
|
531 {IDENT}{S}* { |
|
3974
|
532 return handle_identifier (); |
|
967
|
533 } |
|
1
|
534 |
|
968
|
535 %{ |
|
|
536 // A new line character. New line characters inside matrix constants |
|
4208
|
537 // are handled by the <MATRIX_START> start state code above. If closest |
|
985
|
538 // nesting is inside parentheses, don't return a row separator. |
|
968
|
539 %} |
|
|
540 |
|
967
|
541 {NL} { |
|
|
542 current_input_column = 1; |
|
2857
|
543 lexer_flags.quote_is_transpose = false; |
|
|
544 lexer_flags.cant_be_identifier = false; |
|
|
545 lexer_flags.convert_spaces_to_comma = true; |
|
985
|
546 |
|
1826
|
547 if (nesting_level.none ()) |
|
985
|
548 return '\n'; |
|
4037
|
549 else if (nesting_level.is_paren ()) |
|
|
550 gripe_matlab_incompatible ("bare newline inside parentheses"); |
|
|
551 else if (nesting_level.is_bracket ()) |
|
985
|
552 return LEXICAL_ERROR; |
|
967
|
553 } |
|
1
|
554 |
|
968
|
555 %{ |
|
|
556 // Single quote can either be the beginning of a string or a transpose |
|
|
557 // operator. |
|
|
558 %} |
|
|
559 |
|
967
|
560 "'" { |
|
|
561 current_input_column++; |
|
2857
|
562 lexer_flags.convert_spaces_to_comma = true; |
|
1
|
563 |
|
1826
|
564 if (lexer_flags.quote_is_transpose) |
|
967
|
565 { |
|
|
566 do_comma_insert_check (); |
|
|
567 return QUOTE; |
|
|
568 } |
|
|
569 else |
|
973
|
570 return handle_string ('\''); |
|
967
|
571 } |
|
1
|
572 |
|
968
|
573 %{ |
|
971
|
574 // Double quotes always begin strings. |
|
|
575 %} |
|
|
576 |
|
973
|
577 \" { |
|
|
578 current_input_column++; |
|
|
579 return handle_string ('"'); |
|
|
580 } |
|
971
|
581 |
|
|
582 %{ |
|
|
583 // The colon operator is handled differently if we are in the range |
|
|
584 // part of a plot command. |
|
968
|
585 %} |
|
|
586 |
|
967
|
587 ":" { |
|
1826
|
588 if (lexer_flags.plotting |
|
|
589 && (lexer_flags.in_plot_range || lexer_flags.in_plot_using)) |
|
2857
|
590 BIN_OP_RETURN (COLON, true); |
|
967
|
591 else |
|
2857
|
592 BIN_OP_RETURN (':', false); |
|
967
|
593 } |
|
1
|
594 |
|
968
|
595 %{ |
|
985
|
596 // Gobble comments. If closest nesting is inside parentheses, don't |
|
|
597 // return a new line. |
|
|
598 %} |
|
968
|
599 |
|
967
|
600 {CCHAR} { |
|
1826
|
601 if (help_buf.empty () |
|
|
602 && lexer_flags.beginning_of_function |
|
|
603 && nesting_level.none ()) |
|
967
|
604 { |
|
3665
|
605 lexer_flags.beginning_of_function = false; |
|
|
606 |
|
967
|
607 grab_help_text (); |
|
3665
|
608 |
|
|
609 octave_comment_buffer::append (help_buf); |
|
967
|
610 } |
|
|
611 else |
|
|
612 { |
|
3665
|
613 std::string buf; |
|
|
614 |
|
|
615 bool begin_comment = true; |
|
|
616 |
|
967
|
617 int c; |
|
|
618 while ((c = yyinput ()) != EOF && c != '\n') |
|
3665
|
619 { |
|
|
620 if (begin_comment && (c == '#' || c == '%')) |
|
|
621 ; /* Skip leading comment characters. */ |
|
|
622 else |
|
3802
|
623 buf += static_cast<char> (c); |
|
3665
|
624 } |
|
|
625 |
|
|
626 octave_comment_buffer::append (buf); |
|
967
|
627 } |
|
440
|
628 |
|
967
|
629 current_input_column = 1; |
|
2857
|
630 lexer_flags.quote_is_transpose = false; |
|
|
631 lexer_flags.cant_be_identifier = false; |
|
|
632 lexer_flags.convert_spaces_to_comma = true; |
|
985
|
633 |
|
4037
|
634 maybe_gripe_matlab_incompatible_comment (yytext[0]); |
|
|
635 |
|
1826
|
636 if (nesting_level.none ()) |
|
985
|
637 return '\n'; |
|
3351
|
638 else if (nesting_level.is_bracket ()) |
|
1566
|
639 return ';'; |
|
967
|
640 } |
|
440
|
641 |
|
968
|
642 %{ |
|
|
643 // Other operators. |
|
|
644 %} |
|
|
645 |
|
4037
|
646 ".+" { XBIN_OP_RETURN (EPLUS, false); } |
|
|
647 ".-" { XBIN_OP_RETURN (EMINUS, false); } |
|
2857
|
648 ".*" { BIN_OP_RETURN (EMUL, false); } |
|
|
649 "./" { BIN_OP_RETURN (EDIV, false); } |
|
|
650 ".\\" { BIN_OP_RETURN (ELEFTDIV, false); } |
|
4037
|
651 ".^" { BIN_OP_RETURN (EPOW, false); } |
|
|
652 ".**" { XBIN_OP_RETURN (EPOW, false); } |
|
2857
|
653 ".'" { do_comma_insert_check (); BIN_OP_RETURN (TRANSPOSE, true); } |
|
4037
|
654 "++" { do_comma_insert_check (); XBIN_OP_RETURN (PLUS_PLUS, true); } |
|
|
655 "--" { do_comma_insert_check (); XBIN_OP_RETURN (MINUS_MINUS, true); } |
|
2857
|
656 "<=" { BIN_OP_RETURN (EXPR_LE, false); } |
|
|
657 "==" { BIN_OP_RETURN (EXPR_EQ, false); } |
|
4037
|
658 "~=" { BIN_OP_RETURN (EXPR_NE, false); } |
|
|
659 "!=" { XBIN_OP_RETURN (EXPR_NE, false); } |
|
2857
|
660 ">=" { BIN_OP_RETURN (EXPR_GE, false); } |
|
2877
|
661 "&" { BIN_OP_RETURN (EXPR_AND, false); } |
|
2857
|
662 "|" { BIN_OP_RETURN (EXPR_OR, false); } |
|
|
663 "<" { BIN_OP_RETURN (EXPR_LT, false); } |
|
|
664 ">" { BIN_OP_RETURN (EXPR_GT, false); } |
|
|
665 "*" { BIN_OP_RETURN ('*', false); } |
|
|
666 "/" { BIN_OP_RETURN ('/', false); } |
|
|
667 "\\" { BIN_OP_RETURN (LEFTDIV, false); } |
|
|
668 ";" { BIN_OP_RETURN (';', true); } |
|
|
669 "," { BIN_OP_RETURN (',', true); } |
|
4037
|
670 "^" { BIN_OP_RETURN (POW, false); } |
|
|
671 "**" { XBIN_OP_RETURN (POW, false); } |
|
2857
|
672 "=" { BIN_OP_RETURN ('=', true); } |
|
2877
|
673 "&&" { BIN_OP_RETURN (EXPR_AND_AND, false); } |
|
2857
|
674 "||" { BIN_OP_RETURN (EXPR_OR_OR, false); } |
|
4037
|
675 "<<" { XBIN_OP_RETURN (LSHIFT, false); } |
|
|
676 ">>" { XBIN_OP_RETURN (RSHIFT, false); } |
|
967
|
677 |
|
|
678 {NOT} { |
|
1826
|
679 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
2857
|
680 lexer_flags.past_plot_range = true; |
|
4037
|
681 |
|
|
682 if (yytext[0] == '~') |
|
|
683 BIN_OP_RETURN (EXPR_NOT, false); |
|
|
684 else |
|
|
685 XBIN_OP_RETURN (EXPR_NOT, false); |
|
967
|
686 } |
|
1
|
687 |
|
1276
|
688 "+" { |
|
1826
|
689 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
2857
|
690 lexer_flags.past_plot_range = true; |
|
|
691 BIN_OP_RETURN ('+', false); |
|
967
|
692 } |
|
|
693 |
|
1276
|
694 "-" { |
|
1826
|
695 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
2857
|
696 lexer_flags.past_plot_range = true; |
|
|
697 BIN_OP_RETURN ('-', false); |
|
967
|
698 } |
|
|
699 |
|
|
700 "(" { |
|
4131
|
701 lexer_flags.looking_at_indirect_ref = false; |
|
1826
|
702 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
2857
|
703 lexer_flags.past_plot_range = true; |
|
1826
|
704 nesting_level.paren (); |
|
985
|
705 promptflag--; |
|
967
|
706 TOK_RETURN ('('); |
|
|
707 } |
|
|
708 |
|
|
709 ")" { |
|
1826
|
710 nesting_level.remove (); |
|
1001
|
711 |
|
967
|
712 current_input_column++; |
|
2857
|
713 lexer_flags.cant_be_identifier = true; |
|
|
714 lexer_flags.quote_is_transpose = true; |
|
3351
|
715 lexer_flags.convert_spaces_to_comma = nesting_level.is_bracket (); |
|
1001
|
716 do_comma_insert_check (); |
|
967
|
717 return ')'; |
|
|
718 } |
|
|
719 |
|
2066
|
720 "." { |
|
|
721 TOK_RETURN ('.'); |
|
|
722 } |
|
|
723 |
|
3988
|
724 "@" { |
|
|
725 current_input_column++; |
|
|
726 lexer_flags.cant_be_identifier = false; |
|
|
727 lexer_flags.quote_is_transpose = false; |
|
|
728 lexer_flags.convert_spaces_to_comma = false; |
|
|
729 return '@'; |
|
|
730 } |
|
|
731 |
|
4037
|
732 "+=" { XBIN_OP_RETURN (ADD_EQ, false); } |
|
|
733 "-=" { XBIN_OP_RETURN (SUB_EQ, false); } |
|
|
734 "*=" { XBIN_OP_RETURN (MUL_EQ, false); } |
|
|
735 "/=" { XBIN_OP_RETURN (DIV_EQ, false); } |
|
|
736 "\\=" { XBIN_OP_RETURN (LEFTDIV_EQ, false); } |
|
|
737 ".+=" { XBIN_OP_RETURN (ADD_EQ, false); } |
|
|
738 ".-=" { XBIN_OP_RETURN (SUB_EQ, false); } |
|
|
739 ".*=" { XBIN_OP_RETURN (EMUL_EQ, false); } |
|
|
740 "./=" { XBIN_OP_RETURN (EDIV_EQ, false); } |
|
|
741 ".\\=" { XBIN_OP_RETURN (ELEFTDIV_EQ, false); } |
|
|
742 {POW}= { XBIN_OP_RETURN (POW_EQ, false); } |
|
|
743 {EPOW}= { XBIN_OP_RETURN (EPOW_EQ, false); } |
|
|
744 "&=" { XBIN_OP_RETURN (AND_EQ, false); } |
|
|
745 "|=" { XBIN_OP_RETURN (OR_EQ, false); } |
|
|
746 "<<=" { XBIN_OP_RETURN (LSHIFT_EQ, false); } |
|
|
747 ">>=" { XBIN_OP_RETURN (RSHIFT_EQ, false); } |
|
2877
|
748 |
|
3351
|
749 "{" { |
|
|
750 nesting_level.brace (); |
|
|
751 promptflag--; |
|
|
752 TOK_RETURN ('{'); |
|
|
753 } |
|
|
754 |
|
|
755 "}" { |
|
|
756 nesting_level.remove (); |
|
|
757 |
|
|
758 current_input_column++; |
|
|
759 lexer_flags.cant_be_identifier = true; |
|
|
760 lexer_flags.quote_is_transpose = true; |
|
|
761 lexer_flags.convert_spaces_to_comma = nesting_level.is_bracket (); |
|
|
762 do_comma_insert_check (); // Is this really necessary? |
|
|
763 |
|
|
764 return '}'; |
|
|
765 } |
|
|
766 |
|
968
|
767 %{ |
|
2066
|
768 // Unrecognized input is a lexical error. |
|
968
|
769 %} |
|
1
|
770 |
|
2042
|
771 . { |
|
2066
|
772 current_input_column++; |
|
|
773 |
|
3867
|
774 // EOF can't happen here (we catch it above). |
|
|
775 |
|
|
776 error ("invalid character `%s' (ASCII %d) near line %d, column %d", |
|
|
777 undo_string_escape (yytext[0]), static_cast<int> (yytext[0]), |
|
|
778 input_line_number, current_input_column); |
|
2066
|
779 |
|
|
780 return LEXICAL_ERROR; |
|
|
781 } |
|
1
|
782 |
|
|
783 %% |
|
|
784 |
|
767
|
785 // GAG. |
|
|
786 // |
|
|
787 // If we're reading a matrix and the next character is '[', make sure |
|
|
788 // that we insert a comma ahead of it. |
|
|
789 |
|
146
|
790 void |
|
1
|
791 do_comma_insert_check (void) |
|
|
792 { |
|
1001
|
793 int spc_gobbled = eat_continuation (); |
|
2970
|
794 |
|
1
|
795 int c = yyinput (); |
|
2970
|
796 |
|
3246
|
797 unput (c); |
|
2970
|
798 |
|
1001
|
799 if (spc_gobbled) |
|
3246
|
800 unput (' '); |
|
2970
|
801 |
|
3351
|
802 lexer_flags.do_comma_insert = (lexer_flags.bracketflag && c == '['); |
|
1
|
803 } |
|
|
804 |
|
767
|
805 // Fix things up for errors or interrupts. The parser is never called |
|
|
806 // recursively, so it is always safe to reinitialize its state before |
|
|
807 // doing any parsing. |
|
|
808 |
|
1
|
809 void |
|
|
810 reset_parser (void) |
|
|
811 { |
|
1826
|
812 // Start off on the right foot. |
|
1
|
813 BEGIN 0; |
|
166
|
814 error_state = 0; |
|
3489
|
815 warning_state = 0; |
|
3883
|
816 parser_end_of_input = false; |
|
4186
|
817 symtab_context = 0; |
|
287
|
818 |
|
1826
|
819 // We do want a prompt by default. |
|
1
|
820 promptflag = 1; |
|
287
|
821 |
|
3351
|
822 // Error may have occurred inside some brackets, braces, or parentheses. |
|
985
|
823 nesting_level.clear (); |
|
287
|
824 |
|
1826
|
825 // Clear out the stack of token info used to track line and column |
|
|
826 // numbers. |
|
143
|
827 while (! token_stack.empty ()) |
|
4214
|
828 { |
|
|
829 delete token_stack.top (); |
|
|
830 token_stack.pop (); |
|
|
831 } |
|
287
|
832 |
|
1826
|
833 // Can be reset by defining a function. |
|
985
|
834 if (! (reading_script_file || reading_fcn_file)) |
|
|
835 { |
|
|
836 current_input_column = 1; |
|
2926
|
837 input_line_number = command_editor::current_command_number () - 1; |
|
985
|
838 } |
|
287
|
839 |
|
1826
|
840 // Only ask for input from stdin if we are expecting interactive |
|
|
841 // input. |
|
3174
|
842 if ((interactive || forced_interactive) |
|
3880
|
843 && ! (reading_fcn_file |
|
|
844 || reading_script_file |
|
|
845 || get_input_from_eval_string |
|
3174
|
846 || input_from_startup_file)) |
|
287
|
847 yyrestart (stdin); |
|
991
|
848 |
|
1826
|
849 // Clear the buffer for help text. |
|
|
850 help_buf.resize (0); |
|
1755
|
851 |
|
1826
|
852 // Reset other flags. |
|
|
853 lexer_flags.init (); |
|
1
|
854 } |
|
|
855 |
|
767
|
856 // If we read some newlines, we need figure out what column we're |
|
|
857 // really looking at. |
|
|
858 |
|
1
|
859 static void |
|
|
860 fixup_column_count (char *s) |
|
|
861 { |
|
|
862 char c; |
|
|
863 while ((c = *s++) != '\0') |
|
|
864 { |
|
|
865 if (c == '\n') |
|
143
|
866 current_input_column = 1; |
|
1
|
867 else |
|
|
868 current_input_column++; |
|
|
869 } |
|
|
870 } |
|
|
871 |
|
767
|
872 // Include these so that we don't have to link to libfl.a. |
|
246
|
873 |
|
1
|
874 #ifdef yywrap |
|
|
875 #undef yywrap |
|
|
876 #endif |
|
3332
|
877 int |
|
1
|
878 yywrap (void) |
|
|
879 { |
|
287
|
880 return 1; |
|
1
|
881 } |
|
|
882 |
|
767
|
883 // Tell us all what the current buffer is. |
|
|
884 |
|
1
|
885 YY_BUFFER_STATE |
|
|
886 current_buffer (void) |
|
|
887 { |
|
|
888 return YY_CURRENT_BUFFER; |
|
|
889 } |
|
|
890 |
|
767
|
891 // Create a new buffer. |
|
|
892 |
|
1
|
893 YY_BUFFER_STATE |
|
|
894 create_buffer (FILE *f) |
|
|
895 { |
|
|
896 return yy_create_buffer (f, YY_BUF_SIZE); |
|
|
897 } |
|
|
898 |
|
767
|
899 // Start reading a new buffer. |
|
|
900 |
|
1
|
901 void |
|
|
902 switch_to_buffer (YY_BUFFER_STATE buf) |
|
|
903 { |
|
|
904 yy_switch_to_buffer (buf); |
|
|
905 } |
|
|
906 |
|
767
|
907 // Delete a buffer. |
|
|
908 |
|
1
|
909 void |
|
|
910 delete_buffer (YY_BUFFER_STATE buf) |
|
|
911 { |
|
|
912 yy_delete_buffer (buf); |
|
|
913 } |
|
|
914 |
|
767
|
915 // Restore a buffer (for unwind-prot). |
|
|
916 |
|
1
|
917 void |
|
|
918 restore_input_buffer (void *buf) |
|
|
919 { |
|
2861
|
920 switch_to_buffer (static_cast<YY_BUFFER_STATE> (buf)); |
|
1
|
921 } |
|
|
922 |
|
767
|
923 // Delete a buffer (for unwind-prot). |
|
|
924 |
|
1
|
925 void |
|
|
926 delete_input_buffer (void *buf) |
|
|
927 { |
|
2861
|
928 delete_buffer (static_cast<YY_BUFFER_STATE> (buf)); |
|
1
|
929 } |
|
|
930 |
|
767
|
931 // Check to see if a character string matches any of the possible line |
|
|
932 // styles for plots. |
|
|
933 |
|
3536
|
934 static std::string |
|
3523
|
935 plot_style_token (const std::string& s) |
|
1
|
936 { |
|
3523
|
937 std::string retval; |
|
1823
|
938 |
|
3575
|
939 // XXX FIXME XXX -- specify minimum match length for these. |
|
2804
|
940 static const char *plot_styles[] = |
|
146
|
941 { |
|
925
|
942 "boxes", |
|
924
|
943 "boxerrorbars", |
|
2542
|
944 "boxxyerrorbars", |
|
|
945 "candlesticks", |
|
146
|
946 "dots", |
|
|
947 "errorbars", |
|
2542
|
948 "financebars", |
|
|
949 "fsteps", |
|
|
950 "histeps", |
|
146
|
951 "impulses", |
|
|
952 "lines", |
|
|
953 "linespoints", |
|
|
954 "points", |
|
926
|
955 "steps", |
|
2542
|
956 "vector", |
|
|
957 "xerrorbars", |
|
|
958 "xyerrorbars", |
|
|
959 "yerrorbars", |
|
522
|
960 0, |
|
146
|
961 }; |
|
|
962 |
|
2804
|
963 const char * const *tmp = plot_styles; |
|
522
|
964 while (*tmp) |
|
1
|
965 { |
|
1823
|
966 if (almost_match (*tmp, s.c_str ())) |
|
|
967 { |
|
|
968 retval = *tmp; |
|
|
969 break; |
|
|
970 } |
|
1
|
971 |
|
|
972 tmp++; |
|
|
973 } |
|
|
974 |
|
1823
|
975 return retval; |
|
1
|
976 } |
|
|
977 |
|
3165
|
978 // Check to see if a character string matches any of the possible axes |
|
|
979 // tags for plots. |
|
|
980 |
|
3536
|
981 static std::string |
|
3523
|
982 plot_axes_token (const std::string& s) |
|
3165
|
983 { |
|
3523
|
984 std::string retval; |
|
3165
|
985 |
|
3575
|
986 // XXX FIXME XXX -- specify minimum match length for these. |
|
3561
|
987 static const char *plot_axes[] = |
|
3165
|
988 { |
|
|
989 "x1y1", |
|
|
990 "x1y2", |
|
|
991 "x2y1", |
|
|
992 "x2y2", |
|
|
993 0, |
|
|
994 }; |
|
|
995 |
|
3561
|
996 const char **tmp = plot_axes; |
|
3165
|
997 while (*tmp) |
|
|
998 { |
|
|
999 if (almost_match (*tmp, s.c_str ())) |
|
|
1000 { |
|
|
1001 retval = *tmp; |
|
|
1002 break; |
|
|
1003 } |
|
|
1004 |
|
|
1005 tmp++; |
|
|
1006 } |
|
|
1007 |
|
|
1008 return retval; |
|
|
1009 } |
|
|
1010 |
|
767
|
1011 // Check to see if a character string matches any one of the plot |
|
941
|
1012 // option keywords. Don't match abbreviations for clear, since that's |
|
|
1013 // not a gnuplot keyword (users will probably only expect to be able |
|
|
1014 // to abbreviate actual gnuplot keywords). |
|
767
|
1015 |
|
1
|
1016 static int |
|
3523
|
1017 is_plot_keyword (const std::string& s) |
|
1
|
1018 { |
|
1823
|
1019 const char *t = s.c_str (); |
|
3575
|
1020 if (almost_match ("title", t, 1)) |
|
146
|
1021 { |
|
|
1022 return TITLE; |
|
|
1023 } |
|
3575
|
1024 else if (almost_match ("using", t, 1)) |
|
146
|
1025 { |
|
2857
|
1026 lexer_flags.in_plot_using = true; |
|
146
|
1027 return USING; |
|
|
1028 } |
|
3575
|
1029 else if (almost_match ("with", t, 1)) |
|
146
|
1030 { |
|
2857
|
1031 lexer_flags.in_plot_style = true; |
|
146
|
1032 return WITH; |
|
|
1033 } |
|
3575
|
1034 else if (almost_match ("axes", t, 2) || almost_match ("axis", t, 2)) |
|
3165
|
1035 { |
|
|
1036 lexer_flags.in_plot_axes = true; |
|
|
1037 return AXES; |
|
|
1038 } |
|
1823
|
1039 else if (strcmp ("clear", t) == 0) |
|
883
|
1040 { |
|
|
1041 return CLEAR; |
|
|
1042 } |
|
1
|
1043 else |
|
146
|
1044 { |
|
|
1045 return 0; |
|
|
1046 } |
|
1
|
1047 } |
|
|
1048 |
|
2877
|
1049 // Handle keywords. |
|
767
|
1050 |
|
1
|
1051 static int |
|
3523
|
1052 is_keyword (const std::string& s) |
|
1
|
1053 { |
|
3805
|
1054 int l = input_line_number; |
|
|
1055 int c = current_input_column; |
|
|
1056 |
|
3165
|
1057 if (lexer_flags.plotting) |
|
1
|
1058 { |
|
3165
|
1059 if (lexer_flags.in_plot_style) |
|
1
|
1060 { |
|
3523
|
1061 std::string sty = plot_style_token (s); |
|
3165
|
1062 |
|
|
1063 if (! sty.empty ()) |
|
|
1064 { |
|
|
1065 lexer_flags.in_plot_style = false; |
|
3805
|
1066 yylval.tok_val = new token (sty, l, c); |
|
3165
|
1067 token_stack.push (yylval.tok_val); |
|
|
1068 return STYLE; |
|
|
1069 } |
|
1
|
1070 } |
|
3165
|
1071 else if (lexer_flags.in_plot_axes) |
|
|
1072 { |
|
3523
|
1073 std::string axes = plot_axes_token (s); |
|
3165
|
1074 |
|
|
1075 if (! axes.empty ()) |
|
|
1076 { |
|
|
1077 lexer_flags.in_plot_axes = false; |
|
3805
|
1078 yylval.tok_val = new token (axes, l, c); |
|
3165
|
1079 token_stack.push (yylval.tok_val); |
|
|
1080 return AXES_TAG; |
|
|
1081 } |
|
|
1082 } |
|
1
|
1083 } |
|
|
1084 |
|
1823
|
1085 int len = s.length (); |
|
922
|
1086 |
|
1823
|
1087 const octave_kw *kw = octave_kw_lookup (s.c_str (), len); |
|
191
|
1088 |
|
1497
|
1089 if (kw) |
|
143
|
1090 { |
|
1497
|
1091 yylval.tok_val = 0; |
|
|
1092 |
|
|
1093 switch (kw->kw_id) |
|
|
1094 { |
|
|
1095 case all_va_args_kw: |
|
|
1096 case break_kw: |
|
2764
|
1097 case case_kw: |
|
1497
|
1098 case catch_kw: |
|
|
1099 case continue_kw: |
|
|
1100 case else_kw: |
|
|
1101 case elseif_kw: |
|
|
1102 case global_kw: |
|
2764
|
1103 case otherwise_kw: |
|
1497
|
1104 case return_kw: |
|
2846
|
1105 case static_kw: |
|
3484
|
1106 case until_kw: |
|
1497
|
1107 case unwind_protect_cleanup_kw: |
|
|
1108 break; |
|
|
1109 |
|
|
1110 case end_kw: |
|
|
1111 yylval.tok_val = new token (token::simple_end, l, c); |
|
|
1112 break; |
|
|
1113 |
|
|
1114 case end_try_catch_kw: |
|
|
1115 yylval.tok_val = new token (token::try_catch_end, l, c); |
|
|
1116 break; |
|
|
1117 |
|
|
1118 case end_unwind_protect_kw: |
|
|
1119 yylval.tok_val = new token (token::unwind_protect_end, l, c); |
|
|
1120 break; |
|
|
1121 |
|
|
1122 case endfor_kw: |
|
|
1123 yylval.tok_val = new token (token::for_end, l, c); |
|
|
1124 break; |
|
|
1125 |
|
|
1126 case endfunction_kw: |
|
|
1127 yylval.tok_val = new token (token::function_end, l, c); |
|
|
1128 break; |
|
|
1129 |
|
|
1130 case endif_kw: |
|
|
1131 yylval.tok_val = new token (token::if_end, l, c); |
|
|
1132 break; |
|
|
1133 |
|
2764
|
1134 case endswitch_kw: |
|
|
1135 yylval.tok_val = new token (token::switch_end, l, c); |
|
|
1136 break; |
|
|
1137 |
|
1497
|
1138 case endwhile_kw: |
|
|
1139 yylval.tok_val = new token (token::while_end, l, c); |
|
|
1140 break; |
|
|
1141 |
|
3484
|
1142 case do_kw: |
|
1497
|
1143 case for_kw: |
|
|
1144 case while_kw: |
|
|
1145 promptflag--; |
|
1826
|
1146 lexer_flags.looping++; |
|
1497
|
1147 break; |
|
|
1148 |
|
|
1149 case if_kw: |
|
|
1150 case try_kw: |
|
2764
|
1151 case switch_kw: |
|
1497
|
1152 case unwind_protect_kw: |
|
|
1153 promptflag--; |
|
|
1154 break; |
|
|
1155 |
|
|
1156 case gplot_kw: |
|
2857
|
1157 lexer_flags.plotting = true; |
|
1497
|
1158 yylval.tok_val = new token (token::two_dee, l, c); |
|
|
1159 break; |
|
|
1160 |
|
|
1161 case gsplot_kw: |
|
2857
|
1162 lexer_flags.plotting = true; |
|
1497
|
1163 yylval.tok_val = new token (token::three_dee, l, c); |
|
|
1164 break; |
|
|
1165 |
|
|
1166 case replot_kw: |
|
2857
|
1167 lexer_flags.plotting = true; |
|
1497
|
1168 yylval.tok_val = new token (token::replot, l, c); |
|
|
1169 break; |
|
|
1170 |
|
|
1171 case function_kw: |
|
1826
|
1172 if (lexer_flags.defining_func) |
|
1497
|
1173 { |
|
|
1174 error ("function keyword invalid within a function body"); |
|
|
1175 |
|
|
1176 if ((reading_fcn_file || reading_script_file) |
|
1755
|
1177 && ! curr_fcn_file_name.empty ()) |
|
1497
|
1178 error ("defining new function near line %d of file `%s.m'", |
|
1755
|
1179 input_line_number, curr_fcn_file_name.c_str ()); |
|
1497
|
1180 else |
|
1755
|
1181 error ("defining new function near line %d", |
|
|
1182 input_line_number); |
|
1497
|
1183 |
|
|
1184 return LEXICAL_ERROR; |
|
|
1185 } |
|
|
1186 else |
|
|
1187 { |
|
2877
|
1188 // Prepare for local symbols. |
|
|
1189 |
|
1497
|
1190 tmp_local_sym_tab = new symbol_table (); |
|
2877
|
1191 |
|
|
1192 promptflag--; |
|
|
1193 |
|
2857
|
1194 lexer_flags.defining_func = true; |
|
2877
|
1195 lexer_flags.parsed_function_name = false; |
|
2857
|
1196 lexer_flags.beginning_of_function = true; |
|
2877
|
1197 |
|
1497
|
1198 if (! (reading_fcn_file || reading_script_file)) |
|
|
1199 input_line_number = 1; |
|
|
1200 } |
|
|
1201 break; |
|
|
1202 |
|
3174
|
1203 case magic_file_kw: |
|
|
1204 { |
|
|
1205 if ((reading_fcn_file || reading_script_file) |
|
|
1206 && ! curr_fcn_file_full_name.empty ()) |
|
|
1207 yylval.tok_val = new token (curr_fcn_file_full_name, l, c); |
|
|
1208 else |
|
|
1209 yylval.tok_val = new token ("stdin", l, c); |
|
|
1210 } |
|
|
1211 break; |
|
|
1212 |
|
|
1213 case magic_line_kw: |
|
|
1214 yylval.tok_val = new token (static_cast<double> (l), "", l, c); |
|
|
1215 break; |
|
|
1216 |
|
3974
|
1217 case varargin_kw: |
|
|
1218 if (! lexer_flags.looking_at_parameter_list) |
|
|
1219 return 0; |
|
|
1220 break; |
|
|
1221 |
|
|
1222 case varargout_kw: |
|
|
1223 if (! lexer_flags.looking_at_return_list) |
|
|
1224 return 0; |
|
|
1225 break; |
|
|
1226 |
|
1497
|
1227 default: |
|
|
1228 panic_impossible (); |
|
|
1229 } |
|
|
1230 |
|
|
1231 if (! yylval.tok_val) |
|
|
1232 yylval.tok_val = new token (l, c); |
|
|
1233 |
|
476
|
1234 token_stack.push (yylval.tok_val); |
|
1497
|
1235 |
|
|
1236 return kw->tok; |
|
143
|
1237 } |
|
1
|
1238 |
|
|
1239 return 0; |
|
|
1240 } |
|
|
1241 |
|
767
|
1242 // Try to find an identifier. All binding to global or builtin |
|
|
1243 // variables occurs when expressions are evaluated. |
|
|
1244 |
|
1
|
1245 static symbol_record * |
|
3523
|
1246 lookup_identifier (const std::string& name) |
|
1
|
1247 { |
|
2856
|
1248 return curr_sym_tab->lookup (name, true); |
|
1
|
1249 } |
|
|
1250 |
|
2702
|
1251 static bool |
|
3523
|
1252 is_variable (const std::string& name) |
|
2702
|
1253 { |
|
2856
|
1254 symbol_record *sr = curr_sym_tab->lookup (name); |
|
2702
|
1255 |
|
|
1256 return sr && sr->is_variable (); |
|
|
1257 } |
|
|
1258 |
|
|
1259 static void |
|
3523
|
1260 force_local_variable (const std::string& name) |
|
2702
|
1261 { |
|
4101
|
1262 if (! is_variable (name)) |
|
|
1263 curr_sym_tab->clear (name); |
|
|
1264 |
|
2856
|
1265 symbol_record *sr = curr_sym_tab->lookup (name, true); |
|
2702
|
1266 |
|
|
1267 if (sr) |
|
|
1268 sr->define (octave_value ()); |
|
|
1269 } |
|
|
1270 |
|
1019
|
1271 // Grab the help text from an function file. Always overwrites the |
|
|
1272 // current contents of help_buf. |
|
767
|
1273 |
|
3427
|
1274 // XXX FIXME XXX -- gobble_leading_white_space() in parse.y |
|
2300
|
1275 // duplicates some of this code! |
|
|
1276 |
|
1
|
1277 static void |
|
|
1278 grab_help_text (void) |
|
|
1279 { |
|
1755
|
1280 help_buf.resize (0); |
|
1019
|
1281 |
|
2300
|
1282 bool begin_comment = true; |
|
|
1283 bool in_comment = true; |
|
3427
|
1284 bool discard_space = true; |
|
3665
|
1285 |
|
1019
|
1286 int c = 0; |
|
1
|
1287 |
|
1019
|
1288 while ((c = yyinput ()) != EOF) |
|
|
1289 { |
|
2300
|
1290 if (begin_comment) |
|
|
1291 { |
|
|
1292 if (c == '%' || c == '#') |
|
|
1293 continue; |
|
3427
|
1294 else if (discard_space && c == ' ') |
|
|
1295 { |
|
|
1296 discard_space = false; |
|
|
1297 continue; |
|
|
1298 } |
|
2300
|
1299 else |
|
|
1300 begin_comment = false; |
|
|
1301 } |
|
|
1302 |
|
1019
|
1303 if (in_comment) |
|
1
|
1304 { |
|
3802
|
1305 help_buf += static_cast<char> (c); |
|
1755
|
1306 |
|
1019
|
1307 if (c == '\n') |
|
3427
|
1308 { |
|
|
1309 in_comment = false; |
|
|
1310 discard_space = true; |
|
|
1311 } |
|
1019
|
1312 } |
|
|
1313 else |
|
|
1314 { |
|
|
1315 switch (c) |
|
991
|
1316 { |
|
4037
|
1317 case '#': |
|
1019
|
1318 case '%': |
|
4037
|
1319 maybe_gripe_matlab_incompatible_comment (yytext[0]); |
|
2300
|
1320 in_comment = true; |
|
|
1321 begin_comment = true; |
|
1019
|
1322 break; |
|
777
|
1323 |
|
1019
|
1324 case ' ': |
|
|
1325 case '\t': |
|
|
1326 break; |
|
777
|
1327 |
|
1019
|
1328 default: |
|
|
1329 goto done; |
|
1
|
1330 } |
|
|
1331 } |
|
1019
|
1332 } |
|
991
|
1333 |
|
1019
|
1334 done: |
|
991
|
1335 |
|
1019
|
1336 if (c) |
|
3246
|
1337 unput (c); |
|
1
|
1338 } |
|
|
1339 |
|
767
|
1340 // Return 1 if the given character matches any character in the given |
|
|
1341 // string. |
|
|
1342 |
|
2857
|
1343 static bool |
|
2804
|
1344 match_any (char c, const char *s) |
|
1
|
1345 { |
|
|
1346 char tmp; |
|
|
1347 while ((tmp = *s++) != '\0') |
|
|
1348 { |
|
|
1349 if (c == tmp) |
|
2857
|
1350 return true; |
|
1
|
1351 } |
|
2857
|
1352 return false; |
|
1
|
1353 } |
|
|
1354 |
|
767
|
1355 // Given information about the spacing surrounding an operator, |
|
|
1356 // return 1 if it looks like it should be treated as a binary |
|
|
1357 // operator. For example, |
|
|
1358 // |
|
3774
|
1359 // [ 1 + 2 ] or [ 1+ 2] or [ 1+2 ] ==> binary |
|
|
1360 // |
|
|
1361 // [ 1 +2 ] ==> unary |
|
767
|
1362 |
|
2857
|
1363 static bool |
|
3246
|
1364 looks_like_bin_op (bool spc_prev, int next_char) |
|
1
|
1365 { |
|
3246
|
1366 bool spc_next = (next_char == ' ' || next_char == '\t'); |
|
|
1367 |
|
608
|
1368 return ((spc_prev && spc_next) || ! spc_prev); |
|
1
|
1369 } |
|
|
1370 |
|
3263
|
1371 // Recognize separators. If the separator is a CRLF pair, it is |
|
|
1372 // replaced by a single LF. |
|
|
1373 |
|
|
1374 static bool |
|
|
1375 next_token_is_sep_op (void) |
|
|
1376 { |
|
|
1377 bool retval = false; |
|
|
1378 |
|
|
1379 int c1 = yyinput (); |
|
|
1380 |
|
|
1381 if (c1 == '\r') |
|
|
1382 { |
|
|
1383 int c2 = yyinput (); |
|
|
1384 |
|
|
1385 if (c2 == '\n') |
|
|
1386 { |
|
|
1387 c1 = '\n'; |
|
|
1388 |
|
|
1389 retval = true; |
|
|
1390 } |
|
|
1391 else |
|
|
1392 unput (c2); |
|
|
1393 } |
|
|
1394 else |
|
|
1395 retval = match_any (c1, ",;\n]"); |
|
|
1396 |
|
|
1397 unput (c1); |
|
|
1398 |
|
|
1399 return retval; |
|
|
1400 } |
|
|
1401 |
|
767
|
1402 // Try to determine if the next token should be treated as a postfix |
|
|
1403 // unary operator. This is ugly, but it seems to do the right thing. |
|
|
1404 |
|
2857
|
1405 static bool |
|
3246
|
1406 next_token_is_postfix_unary_op (bool spc_prev) |
|
1
|
1407 { |
|
2857
|
1408 bool un_op = false; |
|
1
|
1409 |
|
|
1410 int c0 = yyinput (); |
|
|
1411 |
|
3246
|
1412 if (c0 == '\'' && ! spc_prev) |
|
|
1413 { |
|
|
1414 un_op = true; |
|
|
1415 } |
|
|
1416 else if (c0 == '.') |
|
|
1417 { |
|
|
1418 int c1 = yyinput (); |
|
|
1419 un_op = (c1 == '\''); |
|
|
1420 unput (c1); |
|
|
1421 } |
|
1
|
1422 |
|
3246
|
1423 unput (c0); |
|
1
|
1424 |
|
|
1425 return un_op; |
|
|
1426 } |
|
|
1427 |
|
767
|
1428 // Try to determine if the next token should be treated as a binary |
|
3246
|
1429 // operator. |
|
1521
|
1430 // |
|
3246
|
1431 // This kluge exists because whitespace is not always ignored inside |
|
3774
|
1432 // the square brackets that are used to create matrix objects (though |
|
|
1433 // spacing only really matters in the cases that can be interpreted |
|
|
1434 // either as binary ops or prefix unary ops: currently just +, -). |
|
|
1435 // |
|
3779
|
1436 // Note that a line continuation directly following a + or - operator |
|
|
1437 // (e.g., the characters '[' 'a' ' ' '+' '\' LFD 'b' ']') will be |
|
|
1438 // parsed as a binary operator. |
|
767
|
1439 |
|
2857
|
1440 static bool |
|
3246
|
1441 next_token_is_bin_op (bool spc_prev) |
|
1
|
1442 { |
|
2857
|
1443 bool bin_op = false; |
|
1
|
1444 |
|
|
1445 int c0 = yyinput (); |
|
|
1446 |
|
|
1447 switch (c0) |
|
|
1448 { |
|
777
|
1449 case '+': |
|
|
1450 case '-': |
|
3774
|
1451 { |
|
|
1452 int c1 = yyinput (); |
|
|
1453 |
|
|
1454 switch (c1) |
|
|
1455 { |
|
|
1456 case '+': |
|
|
1457 case '-': |
|
|
1458 // Unary ops, spacing doesn't matter. |
|
|
1459 break; |
|
|
1460 |
|
|
1461 case '=': |
|
|
1462 // Binary ops, spacing doesn't matter. |
|
|
1463 bin_op = true; |
|
|
1464 break; |
|
|
1465 |
|
|
1466 default: |
|
|
1467 // Could be either, spacing matters. |
|
|
1468 bin_op = looks_like_bin_op (spc_prev, c1); |
|
|
1469 break; |
|
|
1470 } |
|
|
1471 |
|
|
1472 unput (c1); |
|
|
1473 } |
|
|
1474 break; |
|
|
1475 |
|
|
1476 case ':': |
|
3246
|
1477 case '/': |
|
|
1478 case '\\': |
|
|
1479 case '^': |
|
3774
|
1480 // Always a binary op (may also include /=, \=, and ^=). |
|
|
1481 bin_op = true; |
|
1276
|
1482 break; |
|
|
1483 |
|
3246
|
1484 // .+ .- ./ .\ .^ .* .** |
|
1554
|
1485 case '.': |
|
|
1486 { |
|
|
1487 int c1 = yyinput (); |
|
3246
|
1488 |
|
3774
|
1489 if (match_any (c1, "+-/\\^*")) |
|
|
1490 // Always a binary op (may also include .+=, .-=, ./=, ...). |
|
|
1491 bin_op = true; |
|
3698
|
1492 else if (! isdigit (c1) && c1 != ' ' && c1 != '\t' && c1 != '.') |
|
3774
|
1493 // A structure element reference is a binary op. |
|
|
1494 bin_op = true; |
|
3246
|
1495 |
|
|
1496 unput (c1); |
|
1554
|
1497 } |
|
|
1498 break; |
|
|
1499 |
|
3246
|
1500 // = == & && | || * ** |
|
|
1501 case '=': |
|
1
|
1502 case '&': |
|
3246
|
1503 case '|': |
|
1
|
1504 case '*': |
|
3774
|
1505 // Always a binary op (may also include ==, &&, ||, **). |
|
|
1506 bin_op = true; |
|
3246
|
1507 break; |
|
|
1508 |
|
3774
|
1509 // < <= <> > >= |
|
1
|
1510 case '<': |
|
|
1511 case '>': |
|
3774
|
1512 // Always a binary op (may also include <=, <>, >=). |
|
|
1513 bin_op = true; |
|
|
1514 break; |
|
|
1515 |
|
|
1516 // ~= != |
|
777
|
1517 case '~': |
|
|
1518 case '!': |
|
3246
|
1519 { |
|
|
1520 int c1 = yyinput (); |
|
|
1521 |
|
3774
|
1522 // ~ and ! can be unary ops, so require following =. |
|
|
1523 if (c1 == '=') |
|
|
1524 bin_op = true; |
|
3246
|
1525 |
|
|
1526 unput (c1); |
|
|
1527 } |
|
1
|
1528 break; |
|
|
1529 |
|
|
1530 default: |
|
1276
|
1531 break; |
|
1
|
1532 } |
|
|
1533 |
|
3246
|
1534 unput (c0); |
|
1
|
1535 |
|
|
1536 return bin_op; |
|
|
1537 } |
|
|
1538 |
|
767
|
1539 // Used to delete trailing white space from tokens. |
|
|
1540 |
|
3536
|
1541 static std::string |
|
1
|
1542 strip_trailing_whitespace (char *s) |
|
|
1543 { |
|
3523
|
1544 std::string retval = s; |
|
1
|
1545 |
|
1823
|
1546 size_t pos = retval.find_first_of (" \t"); |
|
1
|
1547 |
|
1823
|
1548 if (pos != NPOS) |
|
|
1549 retval.resize (pos); |
|
1
|
1550 |
|
|
1551 return retval; |
|
|
1552 } |
|
|
1553 |
|
3665
|
1554 static void |
|
|
1555 scan_for_comments (const char *text) |
|
|
1556 { |
|
|
1557 std::string comment_buf; |
|
|
1558 |
|
|
1559 bool in_comment = false; |
|
|
1560 bool beginning_of_comment = false; |
|
|
1561 |
|
|
1562 int len = strlen (text); |
|
|
1563 int i = 0; |
|
|
1564 |
|
|
1565 while (i < len) |
|
|
1566 { |
|
|
1567 char c = text[i++]; |
|
|
1568 |
|
|
1569 switch (c) |
|
|
1570 { |
|
|
1571 case '%': |
|
|
1572 case '#': |
|
|
1573 if (in_comment) |
|
|
1574 { |
|
|
1575 if (! beginning_of_comment) |
|
3802
|
1576 comment_buf += static_cast<char> (c); |
|
3665
|
1577 } |
|
|
1578 else |
|
|
1579 { |
|
4037
|
1580 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
1581 in_comment = true; |
|
|
1582 beginning_of_comment = true; |
|
|
1583 } |
|
|
1584 break; |
|
|
1585 |
|
|
1586 case '\n': |
|
|
1587 if (in_comment) |
|
|
1588 { |
|
3802
|
1589 comment_buf += static_cast<char> (c); |
|
3665
|
1590 octave_comment_buffer::append (comment_buf); |
|
|
1591 comment_buf.resize (0); |
|
|
1592 in_comment = false; |
|
|
1593 beginning_of_comment = false; |
|
|
1594 } |
|
|
1595 break; |
|
|
1596 |
|
|
1597 case '\r': |
|
|
1598 if (in_comment) |
|
3802
|
1599 comment_buf += static_cast<char> (c); |
|
3665
|
1600 if (i < len) |
|
|
1601 { |
|
|
1602 c = text[i++]; |
|
|
1603 |
|
|
1604 if (c == '\n') |
|
|
1605 { |
|
|
1606 if (in_comment) |
|
|
1607 { |
|
3802
|
1608 comment_buf += static_cast<char> (c); |
|
3665
|
1609 octave_comment_buffer::append (comment_buf); |
|
|
1610 in_comment = false; |
|
|
1611 beginning_of_comment = false; |
|
|
1612 } |
|
|
1613 } |
|
|
1614 } |
|
|
1615 |
|
|
1616 default: |
|
|
1617 if (in_comment) |
|
|
1618 { |
|
3802
|
1619 comment_buf += static_cast<char> (c); |
|
3665
|
1620 beginning_of_comment = false; |
|
|
1621 } |
|
|
1622 break; |
|
|
1623 } |
|
|
1624 } |
|
|
1625 |
|
|
1626 if (! comment_buf.empty ()) |
|
|
1627 octave_comment_buffer::append (comment_buf); |
|
|
1628 } |
|
|
1629 |
|
1001
|
1630 // Discard whitespace, including comments and continuations. |
|
1088
|
1631 // |
|
|
1632 // Return value is logical OR of the following values: |
|
|
1633 // |
|
1826
|
1634 // ATE_NOTHING : no spaces to eat |
|
1088
|
1635 // ATE_SPACE_OR_TAB : space or tab in input |
|
|
1636 // ATE_NEWLINE : bare new line in input |
|
1001
|
1637 |
|
1826
|
1638 static yum_yum |
|
975
|
1639 eat_whitespace (void) |
|
|
1640 { |
|
1826
|
1641 yum_yum retval = ATE_NOTHING; |
|
3665
|
1642 |
|
|
1643 std::string comment_buf; |
|
|
1644 |
|
2857
|
1645 bool in_comment = false; |
|
3665
|
1646 bool beginning_of_comment = false; |
|
|
1647 |
|
|
1648 int c = 0; |
|
|
1649 |
|
975
|
1650 while ((c = yyinput ()) != EOF) |
|
|
1651 { |
|
|
1652 current_input_column++; |
|
|
1653 |
|
|
1654 switch (c) |
|
|
1655 { |
|
|
1656 case ' ': |
|
|
1657 case '\t': |
|
3665
|
1658 if (in_comment) |
|
|
1659 { |
|
3802
|
1660 comment_buf += static_cast<char> (c); |
|
3665
|
1661 beginning_of_comment = false; |
|
|
1662 } |
|
1088
|
1663 retval |= ATE_SPACE_OR_TAB; |
|
975
|
1664 break; |
|
|
1665 |
|
|
1666 case '\n': |
|
1088
|
1667 retval |= ATE_NEWLINE; |
|
3665
|
1668 if (in_comment) |
|
|
1669 { |
|
3802
|
1670 comment_buf += static_cast<char> (c); |
|
3665
|
1671 octave_comment_buffer::append (comment_buf); |
|
|
1672 comment_buf.resize (0); |
|
|
1673 in_comment = false; |
|
|
1674 beginning_of_comment = false; |
|
|
1675 } |
|
975
|
1676 current_input_column = 0; |
|
|
1677 break; |
|
|
1678 |
|
|
1679 case '#': |
|
|
1680 case '%': |
|
3665
|
1681 if (in_comment) |
|
|
1682 { |
|
|
1683 if (! beginning_of_comment) |
|
3802
|
1684 comment_buf += static_cast<char> (c); |
|
3665
|
1685 } |
|
|
1686 else |
|
|
1687 { |
|
4037
|
1688 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
1689 in_comment = true; |
|
|
1690 beginning_of_comment = true; |
|
|
1691 } |
|
975
|
1692 break; |
|
|
1693 |
|
1001
|
1694 case '.': |
|
|
1695 if (in_comment) |
|
3665
|
1696 { |
|
3802
|
1697 comment_buf += static_cast<char> (c); |
|
3665
|
1698 beginning_of_comment = false; |
|
|
1699 break; |
|
|
1700 } |
|
1001
|
1701 else |
|
|
1702 { |
|
|
1703 if (have_ellipsis_continuation ()) |
|
|
1704 break; |
|
|
1705 else |
|
|
1706 goto done; |
|
|
1707 } |
|
|
1708 |
|
|
1709 case '\\': |
|
|
1710 if (in_comment) |
|
3665
|
1711 { |
|
3802
|
1712 comment_buf += static_cast<char> (c); |
|
3665
|
1713 beginning_of_comment = false; |
|
|
1714 break; |
|
|
1715 } |
|
1001
|
1716 else |
|
|
1717 { |
|
3105
|
1718 if (have_continuation ()) |
|
1001
|
1719 break; |
|
|
1720 else |
|
|
1721 goto done; |
|
|
1722 } |
|
|
1723 |
|
975
|
1724 default: |
|
|
1725 if (in_comment) |
|
3665
|
1726 { |
|
3802
|
1727 comment_buf += static_cast<char> (c); |
|
3665
|
1728 beginning_of_comment = false; |
|
|
1729 break; |
|
|
1730 } |
|
975
|
1731 else |
|
|
1732 goto done; |
|
|
1733 } |
|
|
1734 } |
|
|
1735 |
|
3665
|
1736 if (! comment_buf.empty ()) |
|
|
1737 octave_comment_buffer::append (comment_buf); |
|
|
1738 |
|
975
|
1739 done: |
|
3246
|
1740 unput (c); |
|
1082
|
1741 current_input_column--; |
|
1001
|
1742 return retval; |
|
975
|
1743 } |
|
|
1744 |
|
3220
|
1745 static inline bool |
|
|
1746 looks_like_hex (const char *s, int len) |
|
|
1747 { |
|
|
1748 return (len > 2 && s[0] == '0' && (s[1] == 'x' || s[1] == 'X')); |
|
|
1749 } |
|
|
1750 |
|
975
|
1751 static void |
|
3246
|
1752 handle_number (void) |
|
972
|
1753 { |
|
3220
|
1754 double value = 0.0; |
|
|
1755 int nread = 0; |
|
|
1756 |
|
3598
|
1757 if (looks_like_hex (yytext, strlen (yytext))) |
|
3220
|
1758 { |
|
|
1759 unsigned long ival; |
|
3598
|
1760 |
|
|
1761 nread = sscanf (yytext, "%lx", &ival); |
|
|
1762 |
|
3220
|
1763 value = static_cast<double> (ival); |
|
|
1764 } |
|
|
1765 else |
|
3598
|
1766 { |
|
|
1767 char *tmp = strsave (yytext); |
|
|
1768 |
|
|
1769 char *idx = strpbrk (tmp, "Dd"); |
|
2621
|
1770 |
|
3598
|
1771 if (idx) |
|
|
1772 *idx = 'e'; |
|
|
1773 |
|
|
1774 nread = sscanf (tmp, "%lf", &value); |
|
|
1775 |
|
|
1776 delete [] tmp; |
|
|
1777 } |
|
972
|
1778 |
|
1826
|
1779 // If yytext doesn't contain a valid number, we are in deep doo doo. |
|
985
|
1780 |
|
972
|
1781 assert (nread == 1); |
|
|
1782 |
|
3988
|
1783 lexer_flags.quote_is_transpose = true; |
|
|
1784 lexer_flags.cant_be_identifier = true; |
|
|
1785 lexer_flags.convert_spaces_to_comma = true; |
|
972
|
1786 |
|
1826
|
1787 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
3575
|
1788 lexer_flags.past_plot_range = true; |
|
972
|
1789 |
|
|
1790 yylval.tok_val = new token (value, yytext, input_line_number, |
|
|
1791 current_input_column); |
|
|
1792 |
|
|
1793 token_stack.push (yylval.tok_val); |
|
|
1794 |
|
|
1795 current_input_column += yyleng; |
|
|
1796 |
|
|
1797 do_comma_insert_check (); |
|
|
1798 } |
|
|
1799 |
|
1001
|
1800 // We have seen a backslash and need to find out if it should be |
|
|
1801 // treated as a continuation character. If so, this eats it, up to |
|
|
1802 // and including the new line character. |
|
|
1803 // |
|
973
|
1804 // Match whitespace only, followed by a comment character or newline. |
|
|
1805 // Once a comment character is found, discard all input until newline. |
|
|
1806 // If non-whitespace characters are found before comment |
|
|
1807 // characters, return 0. Otherwise, return 1. |
|
|
1808 |
|
2857
|
1809 static bool |
|
3096
|
1810 have_continuation (bool trailing_comments_ok) |
|
973
|
1811 { |
|
4051
|
1812 OSSTREAM buf; |
|
973
|
1813 |
|
3665
|
1814 std::string comment_buf; |
|
|
1815 |
|
2857
|
1816 bool in_comment = false; |
|
3665
|
1817 bool beginning_of_comment = false; |
|
|
1818 |
|
|
1819 int c = 0; |
|
|
1820 |
|
973
|
1821 while ((c = yyinput ()) != EOF) |
|
|
1822 { |
|
3802
|
1823 buf << static_cast<char> (c); |
|
973
|
1824 |
|
|
1825 switch (c) |
|
|
1826 { |
|
|
1827 case ' ': |
|
|
1828 case '\t': |
|
3665
|
1829 if (in_comment) |
|
|
1830 { |
|
3802
|
1831 comment_buf += static_cast<char> (c); |
|
3665
|
1832 beginning_of_comment = false; |
|
|
1833 } |
|
973
|
1834 break; |
|
|
1835 |
|
|
1836 case '%': |
|
|
1837 case '#': |
|
1091
|
1838 if (trailing_comments_ok) |
|
3665
|
1839 { |
|
|
1840 if (in_comment) |
|
|
1841 { |
|
|
1842 if (! beginning_of_comment) |
|
3802
|
1843 comment_buf += static_cast<char> (c); |
|
3665
|
1844 } |
|
|
1845 else |
|
|
1846 { |
|
4037
|
1847 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
1848 in_comment = true; |
|
|
1849 beginning_of_comment = true; |
|
|
1850 } |
|
|
1851 } |
|
1091
|
1852 else |
|
|
1853 goto cleanup; |
|
973
|
1854 break; |
|
|
1855 |
|
|
1856 case '\n': |
|
3665
|
1857 if (in_comment) |
|
|
1858 { |
|
3802
|
1859 comment_buf += static_cast<char> (c); |
|
3665
|
1860 octave_comment_buffer::append (comment_buf); |
|
|
1861 } |
|
975
|
1862 current_input_column = 0; |
|
1001
|
1863 promptflag--; |
|
4037
|
1864 gripe_matlab_incompatible_continuation (); |
|
2857
|
1865 return true; |
|
973
|
1866 |
|
3263
|
1867 case '\r': |
|
3665
|
1868 if (in_comment) |
|
3802
|
1869 comment_buf += static_cast<char> (c); |
|
3263
|
1870 c = yyinput (); |
|
|
1871 if (c == EOF) |
|
|
1872 break; |
|
|
1873 else if (c == '\n') |
|
|
1874 { |
|
3665
|
1875 if (in_comment) |
|
|
1876 { |
|
3802
|
1877 comment_buf += static_cast<char> (c); |
|
3665
|
1878 octave_comment_buffer::append (comment_buf); |
|
|
1879 } |
|
3263
|
1880 current_input_column = 0; |
|
|
1881 promptflag--; |
|
4037
|
1882 gripe_matlab_incompatible_continuation (); |
|
3263
|
1883 return true; |
|
3665
|
1884 } |
|
3263
|
1885 |
|
3802
|
1886 // Fall through... |
|
|
1887 |
|
973
|
1888 default: |
|
3665
|
1889 if (in_comment) |
|
|
1890 { |
|
3802
|
1891 comment_buf += static_cast<char> (c); |
|
3665
|
1892 beginning_of_comment = false; |
|
|
1893 } |
|
|
1894 else |
|
1091
|
1895 goto cleanup; |
|
|
1896 break; |
|
973
|
1897 } |
|
|
1898 } |
|
|
1899 |
|
3246
|
1900 unput (c); |
|
2857
|
1901 return false; |
|
973
|
1902 |
|
3096
|
1903 cleanup: |
|
4051
|
1904 |
|
|
1905 buf << OSSTREAM_ENDS; |
|
|
1906 std::string s = OSSTREAM_STR (buf); |
|
|
1907 OSSTREAM_FREEZE (buf); |
|
|
1908 |
|
|
1909 int len = s.length (); |
|
|
1910 while (len--) |
|
|
1911 unput (s[len]); |
|
3096
|
1912 |
|
2857
|
1913 return false; |
|
973
|
1914 } |
|
|
1915 |
|
1001
|
1916 // We have seen a `.' and need to see if it is the start of a |
|
|
1917 // continuation. If so, this eats it, up to and including the new |
|
|
1918 // line character. |
|
|
1919 |
|
2857
|
1920 static bool |
|
3096
|
1921 have_ellipsis_continuation (bool trailing_comments_ok) |
|
973
|
1922 { |
|
|
1923 char c1 = yyinput (); |
|
|
1924 if (c1 == '.') |
|
|
1925 { |
|
|
1926 char c2 = yyinput (); |
|
1091
|
1927 if (c2 == '.' && have_continuation (trailing_comments_ok)) |
|
2857
|
1928 return true; |
|
973
|
1929 else |
|
|
1930 { |
|
3246
|
1931 unput (c2); |
|
|
1932 unput (c1); |
|
973
|
1933 } |
|
|
1934 } |
|
|
1935 else |
|
3246
|
1936 unput (c1); |
|
973
|
1937 |
|
2857
|
1938 return false; |
|
973
|
1939 } |
|
|
1940 |
|
1001
|
1941 // See if we have a continuation line. If so, eat it and the leading |
|
|
1942 // whitespace on the next line. |
|
1088
|
1943 // |
|
|
1944 // Return value is the same as described for eat_whitespace(). |
|
1001
|
1945 |
|
1826
|
1946 static yum_yum |
|
1001
|
1947 eat_continuation (void) |
|
|
1948 { |
|
1826
|
1949 int retval = ATE_NOTHING; |
|
3665
|
1950 |
|
1001
|
1951 int c = yyinput (); |
|
3665
|
1952 |
|
1001
|
1953 if ((c == '.' && have_ellipsis_continuation ()) |
|
3105
|
1954 || (c == '\\' && have_continuation ())) |
|
1001
|
1955 retval = eat_whitespace (); |
|
|
1956 else |
|
3246
|
1957 unput (c); |
|
1001
|
1958 |
|
|
1959 return retval; |
|
|
1960 } |
|
|
1961 |
|
973
|
1962 static int |
|
975
|
1963 handle_string (char delim, int text_style) |
|
973
|
1964 { |
|
4051
|
1965 OSSTREAM buf; |
|
973
|
1966 |
|
3805
|
1967 int bos_line = input_line_number; |
|
|
1968 int bos_col = current_input_column; |
|
|
1969 |
|
973
|
1970 int c; |
|
1031
|
1971 int escape_pending = 0; |
|
973
|
1972 |
|
|
1973 while ((c = yyinput ()) != EOF) |
|
|
1974 { |
|
|
1975 current_input_column++; |
|
|
1976 |
|
3105
|
1977 if (c == '\\') |
|
973
|
1978 { |
|
1053
|
1979 if (escape_pending) |
|
|
1980 { |
|
3802
|
1981 buf << static_cast<char> (c); |
|
1053
|
1982 escape_pending = 0; |
|
|
1983 } |
|
|
1984 else |
|
|
1985 { |
|
3096
|
1986 if (have_continuation (false)) |
|
1053
|
1987 escape_pending = 0; |
|
|
1988 else |
|
|
1989 { |
|
3802
|
1990 buf << static_cast<char> (c); |
|
1053
|
1991 escape_pending = 1; |
|
|
1992 } |
|
|
1993 } |
|
1031
|
1994 continue; |
|
973
|
1995 } |
|
|
1996 else if (c == '.') |
|
|
1997 { |
|
3096
|
1998 if (! have_ellipsis_continuation (false)) |
|
3802
|
1999 buf << static_cast<char> (c); |
|
973
|
2000 } |
|
|
2001 else if (c == '\n') |
|
|
2002 { |
|
1053
|
2003 error ("unterminated string constant"); |
|
973
|
2004 break; |
|
|
2005 } |
|
|
2006 else if (c == delim) |
|
|
2007 { |
|
1031
|
2008 if (escape_pending) |
|
3802
|
2009 buf << static_cast<char> (c); |
|
973
|
2010 else |
|
|
2011 { |
|
|
2012 c = yyinput (); |
|
|
2013 if (c == delim) |
|
3802
|
2014 buf << static_cast<char> (c); |
|
973
|
2015 else |
|
|
2016 { |
|
3246
|
2017 unput (c); |
|
4051
|
2018 buf << OSSTREAM_ENDS; |
|
|
2019 std::string s = do_string_escapes (OSSTREAM_STR (buf)); |
|
|
2020 OSSTREAM_FREEZE (buf); |
|
975
|
2021 |
|
1826
|
2022 if (text_style && lexer_flags.doing_set) |
|
975
|
2023 { |
|
3523
|
2024 s = std::string (1, delim) + s + std::string (1, delim); |
|
975
|
2025 } |
|
|
2026 else |
|
|
2027 { |
|
2857
|
2028 lexer_flags.quote_is_transpose = true; |
|
|
2029 lexer_flags.cant_be_identifier = true; |
|
|
2030 lexer_flags.convert_spaces_to_comma = true; |
|
975
|
2031 } |
|
|
2032 |
|
3805
|
2033 yylval.tok_val = new token (s, bos_line, bos_col); |
|
973
|
2034 token_stack.push (yylval.tok_val); |
|
3400
|
2035 |
|
4037
|
2036 if (delim == '"') |
|
|
2037 gripe_matlab_incompatible ("\" used as string delimiter"); |
|
|
2038 else if (delim == '\'') |
|
3400
|
2039 gripe_single_quote_string (); |
|
|
2040 |
|
4064
|
2041 return STRING; |
|
973
|
2042 } |
|
|
2043 } |
|
|
2044 } |
|
|
2045 else |
|
|
2046 { |
|
3802
|
2047 buf << static_cast<char> (c); |
|
973
|
2048 } |
|
|
2049 |
|
1031
|
2050 escape_pending = 0; |
|
973
|
2051 } |
|
|
2052 |
|
|
2053 return LEXICAL_ERROR; |
|
|
2054 } |
|
|
2055 |
|
3208
|
2056 static bool |
|
|
2057 next_token_is_assign_op (void) |
|
|
2058 { |
|
|
2059 bool retval = false; |
|
|
2060 |
|
|
2061 int c0 = yyinput (); |
|
|
2062 |
|
|
2063 switch (c0) |
|
|
2064 { |
|
|
2065 case '=': |
|
|
2066 { |
|
|
2067 int c1 = yyinput (); |
|
|
2068 unput (c1); |
|
|
2069 if (c1 != '=') |
|
|
2070 retval = true; |
|
|
2071 } |
|
|
2072 break; |
|
|
2073 |
|
|
2074 case '+': |
|
|
2075 case '-': |
|
|
2076 case '*': |
|
|
2077 case '/': |
|
|
2078 case '\\': |
|
|
2079 case '&': |
|
|
2080 case '|': |
|
|
2081 { |
|
|
2082 int c1 = yyinput (); |
|
|
2083 unput (c1); |
|
|
2084 if (c1 == '=') |
|
|
2085 retval = true; |
|
|
2086 } |
|
|
2087 break; |
|
|
2088 |
|
|
2089 case '.': |
|
|
2090 { |
|
|
2091 int c1 = yyinput (); |
|
|
2092 if (match_any (c1, "+-*/\\")) |
|
|
2093 { |
|
|
2094 int c2 = yyinput (); |
|
|
2095 unput (c2); |
|
|
2096 if (c2 == '=') |
|
|
2097 retval = true; |
|
|
2098 } |
|
|
2099 unput (c1); |
|
|
2100 } |
|
|
2101 break; |
|
|
2102 |
|
|
2103 case '>': |
|
|
2104 { |
|
|
2105 int c1 = yyinput (); |
|
|
2106 if (c1 == '>') |
|
|
2107 { |
|
|
2108 int c2 = yyinput (); |
|
|
2109 unput (c2); |
|
|
2110 if (c2 == '=') |
|
|
2111 retval = true; |
|
|
2112 } |
|
|
2113 unput (c1); |
|
|
2114 } |
|
|
2115 break; |
|
|
2116 |
|
|
2117 case '<': |
|
|
2118 { |
|
|
2119 int c1 = yyinput (); |
|
|
2120 if (c1 == '<') |
|
|
2121 { |
|
|
2122 int c2 = yyinput (); |
|
|
2123 unput (c2); |
|
|
2124 if (c2 == '=') |
|
|
2125 retval = true; |
|
|
2126 } |
|
|
2127 unput (c1); |
|
|
2128 } |
|
|
2129 break; |
|
|
2130 |
|
|
2131 default: |
|
|
2132 break; |
|
|
2133 } |
|
|
2134 |
|
|
2135 unput (c0); |
|
|
2136 |
|
|
2137 return retval; |
|
|
2138 } |
|
|
2139 |
|
971
|
2140 static int |
|
3351
|
2141 handle_close_bracket (int spc_gobbled) |
|
971
|
2142 { |
|
3208
|
2143 int retval = ']'; |
|
|
2144 |
|
1826
|
2145 if (! nesting_level.none ()) |
|
971
|
2146 { |
|
1826
|
2147 nesting_level.remove (); |
|
3351
|
2148 lexer_flags.bracketflag--; |
|
971
|
2149 } |
|
|
2150 |
|
3351
|
2151 if (lexer_flags.bracketflag == 0) |
|
1001
|
2152 BEGIN 0; |
|
|
2153 |
|
3208
|
2154 if (next_token_is_assign_op () && ! lexer_flags.looking_at_return_list) |
|
971
|
2155 { |
|
3208
|
2156 retval = CLOSE_BRACE; |
|
971
|
2157 } |
|
|
2158 else |
|
|
2159 { |
|
3208
|
2160 int c1 = yyinput (); |
|
971
|
2161 unput (c1); |
|
|
2162 |
|
3388
|
2163 if (lexer_flags.bracketflag) |
|
971
|
2164 { |
|
3246
|
2165 int bin_op = next_token_is_bin_op (spc_gobbled); |
|
3263
|
2166 |
|
3246
|
2167 int postfix_un_op = next_token_is_postfix_unary_op (spc_gobbled); |
|
971
|
2168 |
|
3263
|
2169 int sep_op = next_token_is_sep_op (); |
|
971
|
2170 |
|
3263
|
2171 if (! (postfix_un_op || bin_op || sep_op) |
|
3351
|
2172 && nesting_level.is_bracket () |
|
1826
|
2173 && lexer_flags.convert_spaces_to_comma) |
|
971
|
2174 { |
|
3388
|
2175 maybe_warn_separator_insert (','); |
|
|
2176 |
|
|
2177 if (Vwhitespace_in_literal_matrix != 2) |
|
|
2178 { |
|
|
2179 unput (','); |
|
|
2180 return ']'; |
|
|
2181 } |
|
971
|
2182 } |
|
|
2183 } |
|
|
2184 } |
|
|
2185 |
|
2857
|
2186 lexer_flags.quote_is_transpose = true; |
|
|
2187 lexer_flags.cant_be_identifier = false; |
|
|
2188 lexer_flags.convert_spaces_to_comma = true; |
|
3208
|
2189 |
|
|
2190 return retval; |
|
971
|
2191 } |
|
|
2192 |
|
1072
|
2193 static void |
|
|
2194 maybe_unput_comma (int spc_gobbled) |
|
|
2195 { |
|
3388
|
2196 if (nesting_level.is_bracket ()) |
|
1072
|
2197 { |
|
3246
|
2198 int bin_op = next_token_is_bin_op (spc_gobbled); |
|
1072
|
2199 |
|
3246
|
2200 int postfix_un_op = next_token_is_postfix_unary_op (spc_gobbled); |
|
1072
|
2201 |
|
|
2202 int c1 = yyinput (); |
|
|
2203 int c2 = yyinput (); |
|
2970
|
2204 |
|
1072
|
2205 unput (c2); |
|
|
2206 unput (c1); |
|
2970
|
2207 |
|
3263
|
2208 int sep_op = next_token_is_sep_op (); |
|
2970
|
2209 |
|
1072
|
2210 int dot_op = (c1 == '.' |
|
|
2211 && (isalpha (c2) || isspace (c2) || c2 == '_')); |
|
2970
|
2212 |
|
3388
|
2213 if (postfix_un_op || bin_op || sep_op || dot_op) |
|
|
2214 return; |
|
|
2215 |
|
3985
|
2216 int index_op = (c1 == '(' || c1 == '{'); |
|
3388
|
2217 |
|
|
2218 if (index_op) |
|
|
2219 { |
|
3985
|
2220 // If there is no space before the indexing op, we don't |
|
|
2221 // insert a comma. |
|
3388
|
2222 if (! spc_gobbled) |
|
|
2223 return; |
|
|
2224 |
|
|
2225 maybe_warn_separator_insert (','); |
|
1072
|
2226 |
|
3388
|
2227 // If there is a space, we only insert a comma if we are |
|
|
2228 // trying to be Matlab-like. |
|
|
2229 if (Vwhitespace_in_literal_matrix == 1) |
|
|
2230 unput (','); |
|
|
2231 } |
|
|
2232 else |
|
|
2233 { |
|
|
2234 maybe_warn_separator_insert (','); |
|
|
2235 |
|
|
2236 if (Vwhitespace_in_literal_matrix != 2) |
|
|
2237 unput (','); |
|
|
2238 } |
|
1072
|
2239 } |
|
|
2240 } |
|
|
2241 |
|
767
|
2242 // Figure out exactly what kind of token to return when we have seen |
|
|
2243 // an identifier. Handles keywords. |
|
|
2244 |
|
146
|
2245 static int |
|
3974
|
2246 handle_identifier (void) |
|
146
|
2247 { |
|
3974
|
2248 std::string tok = strip_trailing_whitespace (yytext); |
|
|
2249 |
|
|
2250 int c = yytext[yyleng-1]; |
|
|
2251 |
|
|
2252 int cont_is_spc = eat_continuation (); |
|
|
2253 |
|
|
2254 int spc_gobbled = (cont_is_spc || c == ' ' || c == '\t'); |
|
|
2255 |
|
1826
|
2256 // It is almost always an error for an identifier to be followed |
|
|
2257 // directly by another identifier. Special cases are handled |
|
|
2258 // below. |
|
883
|
2259 |
|
2857
|
2260 lexer_flags.cant_be_identifier = true; |
|
883
|
2261 |
|
2970
|
2262 // If we are expecting a structure element, avoid recognizing |
|
|
2263 // keywords and other special names and return STRUCT_ELT, which is |
|
|
2264 // a string that is also a valid identifier. But first, we have to |
|
|
2265 // decide whether to insert a comma. |
|
747
|
2266 |
|
1826
|
2267 if (lexer_flags.looking_at_indirect_ref) |
|
1072
|
2268 { |
|
2970
|
2269 do_comma_insert_check (); |
|
|
2270 |
|
1072
|
2271 maybe_unput_comma (spc_gobbled); |
|
2819
|
2272 |
|
|
2273 yylval.tok_val = new token (tok, input_line_number, |
|
|
2274 current_input_column); |
|
|
2275 |
|
|
2276 token_stack.push (yylval.tok_val); |
|
|
2277 |
|
2857
|
2278 lexer_flags.cant_be_identifier = false; |
|
|
2279 lexer_flags.quote_is_transpose = true; |
|
|
2280 lexer_flags.convert_spaces_to_comma = true; |
|
2819
|
2281 |
|
|
2282 current_input_column += yyleng; |
|
|
2283 |
|
2970
|
2284 return STRUCT_ELT; |
|
1072
|
2285 } |
|
747
|
2286 |
|
1826
|
2287 // If we have a regular keyword, or a plot STYLE, return it. |
|
|
2288 // Keywords can be followed by identifiers (TOK_RETURN handles |
|
|
2289 // that). |
|
146
|
2290 |
|
|
2291 int kw_token = is_keyword (tok); |
|
2970
|
2292 |
|
146
|
2293 if (kw_token) |
|
|
2294 { |
|
|
2295 if (kw_token == STYLE) |
|
|
2296 { |
|
1060
|
2297 current_input_column += yyleng; |
|
2857
|
2298 lexer_flags.quote_is_transpose = false; |
|
|
2299 lexer_flags.convert_spaces_to_comma = true; |
|
146
|
2300 return kw_token; |
|
|
2301 } |
|
|
2302 else |
|
|
2303 TOK_RETURN (kw_token); |
|
|
2304 } |
|
|
2305 |
|
1826
|
2306 // See if we have a plot keyword (title, using, with, or clear). |
|
146
|
2307 |
|
1826
|
2308 if (lexer_flags.plotting) |
|
941
|
2309 { |
|
1826
|
2310 // Yes, we really do need both of these plot_range variables. |
|
|
2311 // One is used to mark when we are past all possiblity of a plot |
|
|
2312 // range, the other is used to mark when we are actually between |
|
|
2313 // the square brackets that surround the range. |
|
146
|
2314 |
|
1826
|
2315 if (! lexer_flags.in_plot_range) |
|
2857
|
2316 lexer_flags.past_plot_range = true; |
|
941
|
2317 |
|
3351
|
2318 // Option keywords can't appear in brackets, braces, or parentheses. |
|
1273
|
2319 |
|
|
2320 int plot_option_kw = 0; |
|
2970
|
2321 |
|
1826
|
2322 if (nesting_level.none ()) |
|
1273
|
2323 plot_option_kw = is_plot_keyword (tok); |
|
3575
|
2324 |
|
1826
|
2325 if (lexer_flags.cant_be_identifier && plot_option_kw) |
|
941
|
2326 TOK_RETURN (plot_option_kw); |
|
|
2327 } |
|
146
|
2328 |
|
3480
|
2329 int c1 = yyinput (); |
|
|
2330 |
|
|
2331 bool next_tok_is_paren = (c1 == '('); |
|
|
2332 |
|
|
2333 bool next_tok_is_eq = false; |
|
|
2334 if (c1 == '=') |
|
|
2335 { |
|
|
2336 int c2 = yyinput (); |
|
|
2337 unput (c2); |
|
|
2338 |
|
|
2339 if (c2 != '=') |
|
|
2340 next_tok_is_eq = true; |
|
|
2341 } |
|
|
2342 |
|
|
2343 unput (c1); |
|
1001
|
2344 |
|
1826
|
2345 // Make sure we put the return values of a function in the symbol |
|
|
2346 // table that is local to the function. |
|
146
|
2347 |
|
2877
|
2348 // If we are defining a function and we have not seen the function |
|
|
2349 // name yet and the next token is `=', then this identifier must be |
|
|
2350 // the only return value for the function and it belongs in the |
|
|
2351 // local symbol table. |
|
|
2352 |
|
1826
|
2353 if (next_tok_is_eq |
|
2877
|
2354 && lexer_flags.defining_func |
|
|
2355 && ! lexer_flags.parsed_function_name) |
|
146
|
2356 curr_sym_tab = tmp_local_sym_tab; |
|
|
2357 |
|
2702
|
2358 // Kluge alert. |
|
|
2359 // |
|
|
2360 // If we are looking at a text style function, set up to gobble its |
|
2745
|
2361 // arguments. |
|
|
2362 // |
|
|
2363 // If the following token is `=', or if we are parsing a function |
|
3189
|
2364 // return list or function parameter list, or if we are looking at |
|
|
2365 // something like [ab,cd] = foo (), force the symbol to be inserted |
|
|
2366 // as a variable in the current symbol table. |
|
2702
|
2367 |
|
4208
|
2368 if (is_command_name (tok) && ! is_variable (tok)) |
|
2702
|
2369 { |
|
2745
|
2370 if (next_tok_is_eq |
|
|
2371 || lexer_flags.looking_at_return_list |
|
3189
|
2372 || lexer_flags.looking_at_parameter_list |
|
4101
|
2373 || lexer_flags.looking_at_matrix_or_assign_lhs) |
|
2745
|
2374 { |
|
|
2375 force_local_variable (tok); |
|
|
2376 } |
|
2702
|
2377 else if (! next_tok_is_paren) |
|
|
2378 { |
|
|
2379 if (tok == "gset") |
|
2857
|
2380 lexer_flags.doing_set = true; |
|
2702
|
2381 |
|
4208
|
2382 BEGIN COMMAND_START; |
|
2702
|
2383 } |
|
|
2384 } |
|
|
2385 |
|
1826
|
2386 // Find the token in the symbol table. |
|
146
|
2387 |
|
|
2388 yylval.tok_val = new token (lookup_identifier (tok), |
|
|
2389 input_line_number, |
|
|
2390 current_input_column); |
|
|
2391 |
|
|
2392 token_stack.push (yylval.tok_val); |
|
|
2393 |
|
1826
|
2394 // After seeing an identifer, it is ok to convert spaces to a comma |
|
|
2395 // (if needed). |
|
146
|
2396 |
|
2857
|
2397 lexer_flags.convert_spaces_to_comma = true; |
|
146
|
2398 |
|
2877
|
2399 if (! next_tok_is_eq) |
|
|
2400 { |
|
|
2401 lexer_flags.quote_is_transpose = true; |
|
146
|
2402 |
|
2877
|
2403 do_comma_insert_check (); |
|
|
2404 |
|
|
2405 maybe_unput_comma (spc_gobbled); |
|
146
|
2406 } |
|
|
2407 |
|
2877
|
2408 current_input_column += yyleng; |
|
146
|
2409 |
|
|
2410 return NAME; |
|
|
2411 } |
|
|
2412 |
|
767
|
2413 // Print a warning if a function file that defines a function has |
|
|
2414 // anything other than comments and whitespace following the END token |
|
|
2415 // that matches the FUNCTION statement. |
|
|
2416 |
|
1
|
2417 void |
|
|
2418 check_for_garbage_after_fcn_def (void) |
|
|
2419 { |
|
1826
|
2420 // By making a newline be the next character to be read, we will |
|
|
2421 // force the parser to return after reading the function. Calling |
|
3246
|
2422 // unput with EOF does not work. |
|
1
|
2423 |
|
3665
|
2424 std::string comment_buf; |
|
|
2425 |
|
2857
|
2426 bool in_comment = false; |
|
3665
|
2427 bool beginning_of_comment = true; |
|
|
2428 |
|
1
|
2429 int lineno = input_line_number; |
|
3665
|
2430 |
|
|
2431 int c = 0; |
|
|
2432 |
|
1
|
2433 while ((c = yyinput ()) != EOF) |
|
|
2434 { |
|
|
2435 switch (c) |
|
|
2436 { |
|
|
2437 case ' ': |
|
|
2438 case '\t': |
|
|
2439 case ';': |
|
|
2440 case ',': |
|
3665
|
2441 if (in_comment) |
|
|
2442 { |
|
3802
|
2443 comment_buf += static_cast<char> (c); |
|
3665
|
2444 beginning_of_comment = false; |
|
|
2445 } |
|
1
|
2446 break; |
|
777
|
2447 |
|
1
|
2448 case '%': |
|
|
2449 case '#': |
|
3665
|
2450 if (in_comment) |
|
|
2451 { |
|
|
2452 if (! beginning_of_comment) |
|
3802
|
2453 comment_buf += static_cast<char> (c); |
|
3665
|
2454 } |
|
|
2455 else |
|
|
2456 { |
|
4037
|
2457 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
2458 in_comment = true; |
|
|
2459 beginning_of_comment = true; |
|
|
2460 } |
|
1
|
2461 break; |
|
777
|
2462 |
|
3802
|
2463 case '\n': |
|
|
2464 if (in_comment) |
|
|
2465 { |
|
|
2466 comment_buf += static_cast<char> (c); |
|
|
2467 octave_comment_buffer::append (comment_buf); |
|
|
2468 comment_buf.resize (0); |
|
|
2469 in_comment = false; |
|
|
2470 beginning_of_comment = false; |
|
|
2471 } |
|
|
2472 break; |
|
|
2473 |
|
|
2474 case '\r': |
|
|
2475 if (in_comment) |
|
|
2476 comment_buf += static_cast<char> (c); |
|
|
2477 c = yyinput (); |
|
|
2478 if (c == EOF) |
|
|
2479 break; |
|
|
2480 else if (c == '\n') |
|
|
2481 { |
|
|
2482 if (in_comment) |
|
|
2483 { |
|
|
2484 comment_buf += static_cast<char> (c); |
|
|
2485 octave_comment_buffer::append (comment_buf); |
|
|
2486 comment_buf.resize (0); |
|
|
2487 in_comment = false; |
|
|
2488 beginning_of_comment = false; |
|
|
2489 } |
|
|
2490 break; |
|
|
2491 } |
|
|
2492 |
|
|
2493 // Fall through... |
|
|
2494 |
|
1
|
2495 default: |
|
|
2496 if (in_comment) |
|
3665
|
2497 { |
|
3802
|
2498 comment_buf += static_cast<char> (c); |
|
3665
|
2499 beginning_of_comment = false; |
|
|
2500 break; |
|
|
2501 } |
|
1
|
2502 else |
|
|
2503 { |
|
|
2504 warning ("ignoring trailing garbage after end of function\n\ |
|
1755
|
2505 near line %d of file `%s.m'", lineno, curr_fcn_file_name.c_str ()); |
|
1
|
2506 |
|
3246
|
2507 unput ('\n'); |
|
1
|
2508 return; |
|
|
2509 } |
|
|
2510 } |
|
|
2511 } |
|
3665
|
2512 |
|
|
2513 if (! comment_buf.empty ()) |
|
|
2514 octave_comment_buffer::append (comment_buf); |
|
|
2515 |
|
3246
|
2516 unput ('\n'); |
|
1
|
2517 } |
|
|
2518 |
|
1826
|
2519 void |
|
|
2520 lexical_feedback::init (void) |
|
|
2521 { |
|
|
2522 // Not initially defining a matrix list. |
|
3351
|
2523 bracketflag = 0; |
|
1826
|
2524 |
|
|
2525 // Not initially inside a loop or if statement. |
|
|
2526 looping = 0; |
|
|
2527 |
|
2857
|
2528 // Not initially defining a function. |
|
|
2529 beginning_of_function = false; |
|
|
2530 defining_func = false; |
|
2877
|
2531 parsed_function_name = false; |
|
2857
|
2532 |
|
|
2533 // Not parsing a function return or parameter list. |
|
|
2534 looking_at_return_list = false; |
|
|
2535 looking_at_parameter_list = false; |
|
|
2536 |
|
3796
|
2537 // Not parsing a matrix or the left hand side of multi-value |
|
|
2538 // assignment statement. |
|
|
2539 looking_at_matrix_or_assign_lhs = false; |
|
|
2540 |
|
2857
|
2541 // Next token can be identifier. |
|
|
2542 cant_be_identifier = false; |
|
|
2543 |
|
|
2544 // No need to do comma insert or convert spaces to comma at |
|
|
2545 // beginning of input. |
|
|
2546 convert_spaces_to_comma = true; |
|
|
2547 do_comma_insert = false; |
|
|
2548 |
|
|
2549 // Not initially doing any plotting or setting of plot attributes. |
|
|
2550 doing_set = false; |
|
|
2551 in_plot_range = false; |
|
|
2552 in_plot_style = false; |
|
3165
|
2553 in_plot_axes = false; |
|
2857
|
2554 in_plot_using = false; |
|
|
2555 past_plot_range = false; |
|
|
2556 plotting = false; |
|
|
2557 |
|
1826
|
2558 // Not initially looking at indirect references. |
|
2857
|
2559 looking_at_indirect_ref = false; |
|
1826
|
2560 |
|
|
2561 // Quote marks strings intially. |
|
2857
|
2562 quote_is_transpose = false; |
|
1826
|
2563 } |
|
|
2564 |
|
3388
|
2565 static void |
|
|
2566 maybe_warn_separator_insert (char sep) |
|
|
2567 { |
|
3523
|
2568 std::string nm = curr_fcn_file_full_name; |
|
3388
|
2569 |
|
|
2570 if (Vwarn_separator_insert) |
|
|
2571 { |
|
|
2572 if (nm.empty ()) |
|
|
2573 warning ("potential auto-insertion of `%c' near line %d", |
|
|
2574 sep, input_line_number); |
|
|
2575 else |
|
|
2576 warning ("potential auto-insertion of `%c' near line %d of file %s", |
|
|
2577 sep, input_line_number, nm.c_str ()); |
|
|
2578 } |
|
|
2579 } |
|
|
2580 |
|
3400
|
2581 static void |
|
|
2582 gripe_single_quote_string (void) |
|
|
2583 { |
|
3523
|
2584 std::string nm = curr_fcn_file_full_name; |
|
3400
|
2585 |
|
|
2586 if (Vwarn_single_quote_string) |
|
|
2587 { |
|
|
2588 if (nm.empty ()) |
|
|
2589 warning ("single quote delimited string near line %d", |
|
|
2590 input_line_number); |
|
|
2591 else |
|
|
2592 warning ("single quote delimited string near line %d of file %s", |
|
|
2593 input_line_number, nm.c_str ()); |
|
|
2594 } |
|
|
2595 } |
|
|
2596 |
|
4037
|
2597 static void |
|
|
2598 gripe_matlab_incompatible (const std::string& msg) |
|
|
2599 { |
|
|
2600 if (Vwarn_matlab_incompatible) |
|
|
2601 warning ("potential Matlab compatibility problem: %s", msg.c_str ()); |
|
|
2602 } |
|
|
2603 |
|
|
2604 static void |
|
|
2605 maybe_gripe_matlab_incompatible_comment (char c) |
|
|
2606 { |
|
|
2607 if (c == '#') |
|
|
2608 gripe_matlab_incompatible ("# used as comment character"); |
|
|
2609 } |
|
|
2610 |
|
|
2611 static void |
|
|
2612 gripe_matlab_incompatible_continuation (void) |
|
|
2613 { |
|
|
2614 gripe_matlab_incompatible ("\\ used as line continuation marker"); |
|
|
2615 } |
|
|
2616 |
|
|
2617 static void |
|
|
2618 gripe_matlab_incompatible_operator (const std::string& op) |
|
|
2619 { |
|
|
2620 std::string t = op; |
|
|
2621 int n = t.length (); |
|
|
2622 if (t[n-1] == '\n') |
|
|
2623 t.resize (n-1); |
|
|
2624 gripe_matlab_incompatible (t + " used as operator"); |
|
|
2625 } |
|
|
2626 |
|
|
2627 static int |
|
|
2628 warn_matlab_incompatible (void) |
|
|
2629 { |
|
|
2630 Vwarn_matlab_incompatible = check_preference ("warn_matlab_incompatible"); |
|
|
2631 |
|
|
2632 return 0; |
|
|
2633 } |
|
|
2634 |
|
3388
|
2635 static int |
|
|
2636 warn_separator_insert (void) |
|
|
2637 { |
|
|
2638 Vwarn_separator_insert = check_preference ("warn_separator_insert"); |
|
|
2639 |
|
|
2640 return 0; |
|
|
2641 } |
|
|
2642 |
|
3246
|
2643 static int |
|
3400
|
2644 warn_single_quote_string (void) |
|
|
2645 { |
|
|
2646 Vwarn_single_quote_string = check_preference ("warn_single_quote_string"); |
|
|
2647 |
|
|
2648 return 0; |
|
|
2649 } |
|
|
2650 |
|
|
2651 static int |
|
2167
|
2652 whitespace_in_literal_matrix (void) |
|
|
2653 { |
|
|
2654 int pref = 0; |
|
3009
|
2655 |
|
3523
|
2656 std::string val = builtin_string_variable ("whitespace_in_literal_matrix"); |
|
3009
|
2657 |
|
2167
|
2658 if (! val.empty ()) |
|
|
2659 { |
|
3565
|
2660 if (val == "ignore") |
|
2167
|
2661 pref = 2; |
|
3565
|
2662 else if (val == "traditional") |
|
2167
|
2663 pref = 1; |
|
|
2664 } |
|
3009
|
2665 |
|
2167
|
2666 Vwhitespace_in_literal_matrix = pref; |
|
3009
|
2667 |
|
2167
|
2668 return 0; |
|
|
2669 } |
|
|
2670 |
|
|
2671 void |
|
|
2672 symbols_of_lex (void) |
|
|
2673 { |
|
4037
|
2674 DEFVAR (warn_matlab_incompatible, 0.0, warn_matlab_incompatible, |
|
|
2675 "-*- texinfo -*-\n\ |
|
|
2676 @defvr {Built-in Variable} warn_matlab_incompatible\n\ |
|
|
2677 Print warnings for Octave language features that may cause\n\ |
|
|
2678 compatibility problems with Matlab.\n\ |
|
|
2679 @end defvr"); |
|
|
2680 |
|
3388
|
2681 DEFVAR (warn_separator_insert, 0.0, warn_separator_insert, |
|
3428
|
2682 "-*- texinfo -*-\n\ |
|
|
2683 @defvr {Built-in Variable} warn_separator_insert\n\ |
|
3567
|
2684 Print warning if commas or semicolons might be inserted\n\ |
|
3428
|
2685 automatically in literal matrices.\n\ |
|
|
2686 @end defvr"); |
|
3388
|
2687 |
|
3400
|
2688 DEFVAR (warn_single_quote_string, 0.0, warn_single_quote_string, |
|
3428
|
2689 "-*- texinfo -*-\n\ |
|
|
2690 @defvr {Built-in Variable} warn_single_quote_string\n\ |
|
|
2691 Print warning if a signle quote character is used to introduce a\n\ |
|
|
2692 string constant.\n\ |
|
|
2693 @end defvr"); |
|
3400
|
2694 |
|
3258
|
2695 DEFVAR (whitespace_in_literal_matrix, "", whitespace_in_literal_matrix, |
|
3428
|
2696 "-*- texinfo -*-\n\ |
|
|
2697 @defvr {Built-in Variable} whitespace_in_literal_matrix\n\ |
|
|
2698 Control auto-insertion of commas and semicolons in literal matrices.\n\ |
|
|
2699 @end defvr"); |
|
|
2700 |
|
2167
|
2701 } |
|
|
2702 |
|
1994
|
2703 /* |
|
|
2704 ;;; Local Variables: *** |
|
|
2705 ;;; mode: C++ *** |
|
|
2706 ;;; End: *** |
|
|
2707 */ |
