|
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 |
|
|
23 %s TEXT_FCN |
|
|
24 %s MATRIX |
|
|
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> |
|
|
35 |
|
4093
|
36 #ifdef HAVE_UNISTD_H |
|
|
37 #ifdef HAVE_SYS_TYPES_H |
|
|
38 #include <sys/types.h> |
|
|
39 #endif |
|
|
40 #include <unistd.h> |
|
|
41 #endif |
|
|
42 |
|
1
|
43 #include "SLStack.h" |
|
|
44 |
|
2926
|
45 #include "cmd-edit.h" |
|
4153
|
46 #include "quit.h" |
|
4051
|
47 #include "lo-sstream.h" |
|
2926
|
48 |
|
1497
|
49 // These would be alphabetical, but y.tab.h must be included before |
|
|
50 // oct-gperf.h and y.tab.h must be included after token.h and the tree |
|
|
51 // class declarations. We can't include y.tab.h in oct-gperf.h |
|
|
52 // because it may not be protected to allow it to be included multiple |
|
|
53 // times. |
|
|
54 |
|
3665
|
55 #include "comment-list.h" |
|
2181
|
56 #include "defun.h" |
|
1355
|
57 #include "error.h" |
|
1351
|
58 #include "input.h" |
|
1355
|
59 #include "lex.h" |
|
2891
|
60 #include "ov.h" |
|
1355
|
61 #include "parse.h" |
|
2987
|
62 #include "pt-all.h" |
|
2891
|
63 #include "symtab.h" |
|
|
64 #include "token.h" |
|
|
65 #include "toplev.h" |
|
1355
|
66 #include "utils.h" |
|
|
67 #include "variables.h" |
|
2492
|
68 #include <y.tab.h> |
|
|
69 #include <oct-gperf.h> |
|
1
|
70 |
|
2716
|
71 #if ! (defined (FLEX_SCANNER) \ |
|
|
72 && defined (YY_FLEX_MAJOR_VERSION) && YY_FLEX_MAJOR_VERSION >= 2 \ |
|
|
73 && defined (YY_FLEX_MINOR_VERSION) && YY_FLEX_MINOR_VERSION >= 5) |
|
|
74 #error lex.l requires flex version 2.5.4 or later |
|
|
75 #endif |
|
|
76 |
|
3883
|
77 // TRUE means that we have encountered EOF on the input stream. |
|
|
78 bool parser_end_of_input = false; |
|
|
79 |
|
1826
|
80 // Flags that need to be shared between the lexer and parser. |
|
|
81 lexical_feedback lexer_flags; |
|
|
82 |
|
1351
|
83 // Stack to hold tokens so that we can delete them when the parser is |
|
|
84 // reset and avoid growing forever just because we are stashing some |
|
|
85 // information. This has to appear before lex.h is included, because |
|
|
86 // one of the macros defined there uses token_stack. |
|
2614
|
87 // |
|
|
88 // XXX FIXME XXX -- this should really be static, but that causes |
|
|
89 // problems on some systems. |
|
|
90 SLStack <token*> token_stack; |
|
1351
|
91 |
|
1826
|
92 // Did eat_whitespace() eat a space or tab, or a newline, or both? |
|
1
|
93 |
|
1826
|
94 typedef int yum_yum; |
|
1
|
95 |
|
1826
|
96 const yum_yum ATE_NOTHING = 0; |
|
|
97 const yum_yum ATE_SPACE_OR_TAB = 1; |
|
|
98 const yum_yum ATE_NEWLINE = 2; |
|
1088
|
99 |
|
3351
|
100 // Is the closest nesting level a square bracket, squiggly brace or a paren? |
|
1826
|
101 |
|
3351
|
102 class bracket_brace_paren_nesting_level : public SLStack <int> |
|
1826
|
103 { |
|
|
104 public: |
|
|
105 |
|
3351
|
106 bracket_brace_paren_nesting_level (void) : SLStack<int> () { } |
|
1826
|
107 |
|
3351
|
108 ~bracket_brace_paren_nesting_level (void) { } |
|
|
109 |
|
|
110 void bracket (void) { push (BRACKET); } |
|
|
111 bool is_bracket (void) { return ! empty () && top () == BRACKET; } |
|
1826
|
112 |
|
|
113 void brace (void) { push (BRACE); } |
|
|
114 bool is_brace (void) { return ! empty () && top () == BRACE; } |
|
|
115 |
|
|
116 void paren (void) { push (PAREN); } |
|
|
117 bool is_paren (void) { return ! empty () && top () == PAREN; } |
|
985
|
118 |
|
1826
|
119 bool none (void) { return empty (); } |
|
|
120 |
|
|
121 void remove (void) { if (! empty ()) SLStack<int>::pop (); } |
|
|
122 |
|
|
123 private: |
|
|
124 |
|
3351
|
125 enum { BRACKET = 1, BRACE = 2, PAREN = 3 }; |
|
1826
|
126 |
|
3351
|
127 bracket_brace_paren_nesting_level (const bracket_brace_paren_nesting_level&); |
|
1826
|
128 |
|
3351
|
129 bracket_brace_paren_nesting_level& |
|
|
130 operator = (const bracket_brace_paren_nesting_level&); |
|
1826
|
131 }; |
|
|
132 |
|
3351
|
133 static bracket_brace_paren_nesting_level nesting_level; |
|
1
|
134 |
|
2167
|
135 // Should whitespace in a literal matrix list be automatically |
|
|
136 // converted to commas and semicolons? |
|
|
137 // |
|
|
138 // user specifies value of var |
|
|
139 // -------------- ------------ |
|
|
140 // "ignore" 2 |
|
|
141 // "traditional" 1 |
|
|
142 // anything else 0 |
|
|
143 // |
|
|
144 // Octave will never insert a comma in a literal matrix list if the |
|
|
145 // user specifies "ignore". For example, the statement [1 2] will |
|
|
146 // result in an error instead of being treated the same as [1, 2], and |
|
|
147 // the statement |
|
|
148 // |
|
|
149 // [ 1, 2, |
|
|
150 // 3, 4 ] |
|
|
151 // |
|
|
152 // will result in the vector [1 2 3 4] instead of a matrix. |
|
|
153 // |
|
|
154 // Traditional behavior makes Octave convert spaces to a comma between |
|
|
155 // identifiers and `('. For example, the statement |
|
|
156 // |
|
|
157 // [eye (2)] |
|
|
158 // |
|
|
159 // will be parsed as |
|
|
160 // |
|
|
161 // [eye, (2)] |
|
|
162 // |
|
|
163 // and will result in an error since the `eye' function will be |
|
|
164 // called with no arguments. To get around this, you would have to |
|
|
165 // omit the space between `eye' and the `('. |
|
|
166 // |
|
|
167 // The default value is 0, which results in behavior that is the same |
|
|
168 // as traditional, except that Octave does not convert spaces to a |
|
|
169 // comma between identifiers and `('. For example, the statement |
|
|
170 // |
|
|
171 // [eye (2)] |
|
|
172 // |
|
|
173 // will result in a call to `eye' with the argument `2'. |
|
|
174 |
|
|
175 static int Vwhitespace_in_literal_matrix; |
|
|
176 |
|
4037
|
177 static bool Vwarn_matlab_incompatible = false; |
|
|
178 |
|
3388
|
179 static bool Vwarn_separator_insert = false; |
|
|
180 |
|
3400
|
181 static bool Vwarn_single_quote_string = false; |
|
|
182 |
|
146
|
183 // Forward declarations for functions defined at the bottom of this |
|
|
184 // file. |
|
|
185 |
|
1
|
186 static void fixup_column_count (char *s); |
|
146
|
187 static void do_comma_insert_check (void); |
|
3523
|
188 static int is_plot_keyword (const std::string& s); |
|
|
189 static int is_keyword (const std::string& s); |
|
|
190 static std::string plot_style_token (const std::string& s); |
|
|
191 static symbol_record *lookup_identifier (const std::string& s); |
|
1
|
192 static void grab_help_text (void); |
|
2857
|
193 static bool match_any (char c, const char *s); |
|
3263
|
194 static bool next_token_is_sep_op (void); |
|
3246
|
195 static bool next_token_is_bin_op (bool spc_prev); |
|
|
196 static bool next_token_is_postfix_unary_op (bool spc_prev); |
|
3523
|
197 static std::string strip_trailing_whitespace (char *s); |
|
3246
|
198 static void handle_number (void); |
|
975
|
199 static int handle_string (char delim, int text_style = 0); |
|
3351
|
200 static int handle_close_bracket (int spc_gobbled); |
|
3974
|
201 static int handle_identifier (void); |
|
3096
|
202 static bool have_continuation (bool trailing_comments_ok = true); |
|
|
203 static bool have_ellipsis_continuation (bool trailing_comments_ok = true); |
|
3665
|
204 static void scan_for_comments (const char *); |
|
1826
|
205 static yum_yum eat_whitespace (void); |
|
|
206 static yum_yum eat_continuation (void); |
|
3388
|
207 static void maybe_warn_separator_insert (char sep); |
|
3400
|
208 static void gripe_single_quote_string (void); |
|
4037
|
209 static void gripe_matlab_incompatible (const std::string& msg); |
|
|
210 static void maybe_gripe_matlab_incompatible_comment (char c); |
|
|
211 static void gripe_matlab_incompatible_continuation (void); |
|
|
212 static void gripe_matlab_incompatible_operator (const std::string& op); |
|
1
|
213 |
|
|
214 %} |
|
|
215 |
|
|
216 D [0-9] |
|
|
217 S [ \t] |
|
2042
|
218 NL ((\n)|(\r\n)) |
|
|
219 SNL ({S}|{NL}) |
|
1
|
220 EL (\.\.\.) |
|
967
|
221 BS (\\) |
|
|
222 CONT ({EL}|{BS}) |
|
1
|
223 Im [iIjJ] |
|
967
|
224 CCHAR [#%] |
|
|
225 COMMENT ({CCHAR}.*{NL}) |
|
|
226 SNLCMT ({SNL}|{COMMENT}) |
|
|
227 NOT ((\~)|(\!)) |
|
4037
|
228 POW ((\*\*)|(\^)) |
|
|
229 EPOW (\.{POW}) |
|
1
|
230 IDENT ([_a-zA-Z][_a-zA-Z0-9]*) |
|
|
231 EXPON ([DdEe][+-]?{D}+) |
|
3220
|
232 NUMBER (({D}+\.?{D}*{EXPON}?)|(\.{D}+{EXPON}?)|(0[xX][0-9a-fA-F]+)) |
|
1
|
233 %% |
|
|
234 |
|
968
|
235 %{ |
|
|
236 // Help and other text-style functions are a pain in the ass. This |
|
|
237 // stuff needs to be simplified. May require some changes in the |
|
|
238 // parser too. |
|
|
239 %} |
|
|
240 |
|
967
|
241 <TEXT_FCN>{NL} { |
|
|
242 BEGIN 0; |
|
|
243 current_input_column = 1; |
|
2857
|
244 lexer_flags.quote_is_transpose = false; |
|
|
245 lexer_flags.cant_be_identifier = false; |
|
|
246 lexer_flags.convert_spaces_to_comma = true; |
|
967
|
247 return '\n'; |
|
|
248 } |
|
1
|
249 |
|
967
|
250 <TEXT_FCN>[\;\,] { |
|
1826
|
251 if (lexer_flags.doing_set && strcmp (yytext, ",") == 0) |
|
967
|
252 { |
|
4064
|
253 TOK_PUSH_AND_RETURN (yytext, STRING); |
|
967
|
254 } |
|
|
255 else |
|
|
256 { |
|
|
257 BEGIN 0; |
|
|
258 if (strcmp (yytext, ",") == 0) |
|
|
259 TOK_RETURN (','); |
|
|
260 else |
|
|
261 TOK_RETURN (';'); |
|
|
262 } |
|
|
263 } |
|
1
|
264 |
|
975
|
265 <TEXT_FCN>[\"\'] { |
|
|
266 current_input_column++; |
|
3246
|
267 return handle_string (yytext[0], true); |
|
975
|
268 } |
|
|
269 |
|
3079
|
270 <TEXT_FCN>[^ \t\n\;\,\"\'][^ \t\n\;\,]*{S}* { |
|
3523
|
271 std::string tok = strip_trailing_whitespace (yytext); |
|
4064
|
272 TOK_PUSH_AND_RETURN (tok, STRING); |
|
967
|
273 } |
|
1
|
274 |
|
968
|
275 %{ |
|
1
|
276 // For this and the next two rules, we're looking at ']', and we |
|
971
|
277 // need to know if the next token is `=' or `=='. |
|
1
|
278 // |
|
|
279 // It would have been so much easier if the delimiters were simply |
|
|
280 // different for the expression on the left hand side of the equals |
|
|
281 // operator. |
|
971
|
282 // |
|
|
283 // It's also a pain in the ass to decide whether to insert a comma |
|
|
284 // after seeing a ']' character... |
|
968
|
285 %} |
|
|
286 |
|
2119
|
287 <MATRIX>{SNLCMT}*\]{S}* { |
|
3665
|
288 scan_for_comments (yytext); |
|
1001
|
289 fixup_column_count (yytext); |
|
|
290 int c = yytext[yyleng-1]; |
|
|
291 int cont_is_spc = eat_continuation (); |
|
|
292 int spc_gobbled = (cont_is_spc || c == ' ' || c == '\t'); |
|
3351
|
293 return handle_close_bracket (spc_gobbled); |
|
967
|
294 } |
|
1
|
295 |
|
968
|
296 %{ |
|
1088
|
297 // Commas are element separators in matrix constants. If we don't |
|
|
298 // check for continuations here we can end up inserting too many |
|
|
299 // commas. |
|
968
|
300 %} |
|
|
301 |
|
967
|
302 <MATRIX>{S}*\,{S}* { |
|
1088
|
303 current_input_column += yyleng; |
|
3388
|
304 |
|
1088
|
305 int tmp = eat_continuation (); |
|
3388
|
306 |
|
2857
|
307 lexer_flags.quote_is_transpose = false; |
|
|
308 lexer_flags.cant_be_identifier = false; |
|
|
309 lexer_flags.convert_spaces_to_comma = true; |
|
3388
|
310 |
|
|
311 if ((tmp & ATE_NEWLINE) == ATE_NEWLINE) |
|
|
312 { |
|
|
313 maybe_warn_separator_insert (';'); |
|
|
314 |
|
|
315 if (Vwhitespace_in_literal_matrix != 2) |
|
|
316 unput (';'); |
|
|
317 } |
|
|
318 |
|
1088
|
319 return (','); |
|
967
|
320 } |
|
1
|
321 |
|
968
|
322 %{ |
|
|
323 // In some cases, spaces in matrix constants can turn into commas. |
|
|
324 // If commas are required, spaces are not important in matrix |
|
1088
|
325 // constants so we just eat them. If we don't check for continuations |
|
|
326 // here we can end up inserting too many commas. |
|
968
|
327 %} |
|
430
|
328 |
|
968
|
329 <MATRIX>{S}+ { |
|
1088
|
330 current_input_column += yyleng; |
|
3388
|
331 |
|
|
332 int tmp = eat_continuation (); |
|
|
333 int bin_op = next_token_is_bin_op (true); |
|
|
334 int postfix_un_op = next_token_is_postfix_unary_op (true); |
|
|
335 |
|
|
336 if (! (postfix_un_op || bin_op) |
|
|
337 && nesting_level.is_bracket () |
|
|
338 && lexer_flags.convert_spaces_to_comma) |
|
967
|
339 { |
|
3388
|
340 if ((tmp & ATE_NEWLINE) == ATE_NEWLINE) |
|
|
341 { |
|
|
342 maybe_warn_separator_insert (';'); |
|
967
|
343 |
|
3388
|
344 if (Vwhitespace_in_literal_matrix != 2) |
|
|
345 unput (';'); |
|
|
346 } |
|
|
347 |
|
|
348 if (Vwhitespace_in_literal_matrix != 2) |
|
1088
|
349 { |
|
2857
|
350 lexer_flags.quote_is_transpose = false; |
|
|
351 lexer_flags.cant_be_identifier = false; |
|
|
352 lexer_flags.convert_spaces_to_comma = true; |
|
3388
|
353 |
|
|
354 maybe_warn_separator_insert (','); |
|
|
355 |
|
1088
|
356 return (','); |
|
|
357 } |
|
967
|
358 } |
|
|
359 } |
|
430
|
360 |
|
968
|
361 %{ |
|
1088
|
362 // Semicolons are handled as row seprators in matrix constants. If we |
|
|
363 // don't eat whitespace here we can end up inserting too many |
|
|
364 // semicolons. |
|
968
|
365 %} |
|
|
366 |
|
985
|
367 <MATRIX>{SNLCMT}*;{SNLCMT}* { |
|
3665
|
368 scan_for_comments (yytext); |
|
967
|
369 fixup_column_count (yytext); |
|
1001
|
370 eat_whitespace (); |
|
2857
|
371 lexer_flags.quote_is_transpose = false; |
|
|
372 lexer_flags.cant_be_identifier = false; |
|
|
373 lexer_flags.convert_spaces_to_comma = true; |
|
967
|
374 return ';'; |
|
|
375 } |
|
|
376 |
|
968
|
377 %{ |
|
1088
|
378 // In some cases, new lines can also become row separators. If we |
|
|
379 // don't eat whitespace here we can end up inserting too many |
|
|
380 // semicolons. |
|
985
|
381 %} |
|
|
382 |
|
3180
|
383 <MATRIX>{S}*{COMMENT}{SNLCMT}* | |
|
|
384 <MATRIX>{S}*{NL}{SNLCMT}* { |
|
3665
|
385 scan_for_comments (yytext); |
|
1082
|
386 fixup_column_count (yytext); |
|
1088
|
387 eat_whitespace (); |
|
3388
|
388 |
|
2167
|
389 if (Vwhitespace_in_literal_matrix != 2) |
|
985
|
390 { |
|
2857
|
391 lexer_flags.quote_is_transpose = false; |
|
|
392 lexer_flags.cant_be_identifier = false; |
|
|
393 lexer_flags.convert_spaces_to_comma = true; |
|
985
|
394 |
|
1826
|
395 if (nesting_level.none ()) |
|
985
|
396 return LEXICAL_ERROR; |
|
3388
|
397 } |
|
985
|
398 |
|
3388
|
399 if (nesting_level.is_bracket ()) |
|
|
400 { |
|
|
401 maybe_warn_separator_insert (';'); |
|
|
402 |
|
|
403 if (Vwhitespace_in_literal_matrix != 2) |
|
985
|
404 return ';'; |
|
|
405 } |
|
|
406 } |
|
|
407 |
|
|
408 %{ |
|
3351
|
409 // Open and close bracket are handled differently if we are in the range |
|
968
|
410 // part of a plot command. |
|
975
|
411 // |
|
968
|
412 %} |
|
1
|
413 |
|
967
|
414 \[{S}* { |
|
3351
|
415 nesting_level.bracket (); |
|
975
|
416 |
|
1082
|
417 current_input_column += yyleng; |
|
2857
|
418 lexer_flags.quote_is_transpose = false; |
|
|
419 lexer_flags.cant_be_identifier = false; |
|
|
420 lexer_flags.convert_spaces_to_comma = true; |
|
975
|
421 |
|
|
422 promptflag--; |
|
|
423 eat_whitespace (); |
|
|
424 |
|
1826
|
425 if (lexer_flags.plotting && ! lexer_flags.past_plot_range) |
|
967
|
426 { |
|
2857
|
427 lexer_flags.in_plot_range = true; |
|
1082
|
428 return OPEN_BRACE; |
|
967
|
429 } |
|
|
430 else |
|
|
431 { |
|
3351
|
432 lexer_flags.bracketflag++; |
|
975
|
433 BEGIN MATRIX; |
|
1082
|
434 return '['; |
|
967
|
435 } |
|
|
436 } |
|
1
|
437 |
|
968
|
438 \] { |
|
1826
|
439 nesting_level.remove (); |
|
968
|
440 |
|
1826
|
441 if (lexer_flags.plotting && ! lexer_flags.past_plot_range) |
|
968
|
442 { |
|
2857
|
443 lexer_flags.in_plot_range = false; |
|
968
|
444 TOK_RETURN (CLOSE_BRACE); |
|
|
445 } |
|
|
446 else |
|
|
447 TOK_RETURN (']'); |
|
|
448 } |
|
|
449 |
|
|
450 %{ |
|
|
451 // Imaginary numbers. |
|
|
452 %} |
|
|
453 |
|
|
454 {NUMBER}{Im} { |
|
3246
|
455 handle_number (); |
|
968
|
456 return IMAG_NUM; |
|
|
457 } |
|
|
458 |
|
|
459 %{ |
|
|
460 // Real numbers. Don't grab the `.' part of a dot operator as part of |
|
|
461 // the constant. |
|
|
462 %} |
|
|
463 |
|
|
464 {D}+/\.[\*/\\^'] | |
|
|
465 {NUMBER} { |
|
3246
|
466 handle_number (); |
|
968
|
467 return NUM; |
|
|
468 } |
|
|
469 |
|
|
470 %{ |
|
|
471 // Eat whitespace. Whitespace inside matrix constants is handled by |
|
|
472 // the <MATRIX> start state code above. |
|
|
473 %} |
|
|
474 |
|
967
|
475 {S}* { |
|
|
476 current_input_column += yyleng; |
|
|
477 } |
|
|
478 |
|
968
|
479 %{ |
|
|
480 // Continuation lines. Allow comments after continuations. |
|
|
481 %} |
|
|
482 |
|
967
|
483 {CONT}{S}*{NL} | |
|
|
484 {CONT}{S}*{COMMENT} { |
|
4037
|
485 if (yytext[0] == '\\') |
|
|
486 gripe_matlab_incompatible_continuation (); |
|
3665
|
487 scan_for_comments (yytext); |
|
967
|
488 promptflag--; |
|
|
489 current_input_column = 1; |
|
|
490 } |
|
1
|
491 |
|
968
|
492 %{ |
|
|
493 // An ellipsis not at the end of a line is not a continuation, but |
|
|
494 // does have another meaning. |
|
|
495 %} |
|
|
496 |
|
967
|
497 {EL} { |
|
3973
|
498 if (lexer_flags.looking_at_parameter_list) |
|
3974
|
499 { |
|
|
500 warning ("`...' is deprecated; use varargin instead"); |
|
|
501 return VARARGIN; |
|
|
502 } |
|
3973
|
503 else if (lexer_flags.looking_at_return_list) |
|
3974
|
504 { |
|
|
505 warning ("`...' is deprecated; use varargout instead"); |
|
|
506 return VARARGOUT; |
|
|
507 } |
|
3973
|
508 else |
|
|
509 return LEXICAL_ERROR; |
|
967
|
510 } |
|
1
|
511 |
|
968
|
512 %{ |
|
|
513 // End of file. |
|
|
514 %} |
|
|
515 |
|
967
|
516 <<EOF>> { |
|
|
517 TOK_RETURN (END_OF_INPUT); |
|
|
518 } |
|
1
|
519 |
|
968
|
520 %{ |
|
970
|
521 // Identifiers. Truncate the token at the first space or tab but |
|
|
522 // don't write directly on yytext. |
|
968
|
523 %} |
|
|
524 |
|
967
|
525 {IDENT}{S}* { |
|
3974
|
526 return handle_identifier (); |
|
967
|
527 } |
|
1
|
528 |
|
968
|
529 %{ |
|
|
530 // A new line character. New line characters inside matrix constants |
|
985
|
531 // are handled by the <MATRIX> start state code above. If closest |
|
|
532 // nesting is inside parentheses, don't return a row separator. |
|
968
|
533 %} |
|
|
534 |
|
967
|
535 {NL} { |
|
|
536 current_input_column = 1; |
|
2857
|
537 lexer_flags.quote_is_transpose = false; |
|
|
538 lexer_flags.cant_be_identifier = false; |
|
|
539 lexer_flags.convert_spaces_to_comma = true; |
|
985
|
540 |
|
1826
|
541 if (nesting_level.none ()) |
|
985
|
542 return '\n'; |
|
4037
|
543 else if (nesting_level.is_paren ()) |
|
|
544 gripe_matlab_incompatible ("bare newline inside parentheses"); |
|
|
545 else if (nesting_level.is_bracket ()) |
|
985
|
546 return LEXICAL_ERROR; |
|
967
|
547 } |
|
1
|
548 |
|
968
|
549 %{ |
|
|
550 // Single quote can either be the beginning of a string or a transpose |
|
|
551 // operator. |
|
|
552 %} |
|
|
553 |
|
967
|
554 "'" { |
|
|
555 current_input_column++; |
|
2857
|
556 lexer_flags.convert_spaces_to_comma = true; |
|
1
|
557 |
|
1826
|
558 if (lexer_flags.quote_is_transpose) |
|
967
|
559 { |
|
|
560 do_comma_insert_check (); |
|
|
561 return QUOTE; |
|
|
562 } |
|
|
563 else |
|
973
|
564 return handle_string ('\''); |
|
967
|
565 } |
|
1
|
566 |
|
968
|
567 %{ |
|
971
|
568 // Double quotes always begin strings. |
|
|
569 %} |
|
|
570 |
|
973
|
571 \" { |
|
|
572 current_input_column++; |
|
|
573 return handle_string ('"'); |
|
|
574 } |
|
971
|
575 |
|
|
576 %{ |
|
|
577 // The colon operator is handled differently if we are in the range |
|
|
578 // part of a plot command. |
|
968
|
579 %} |
|
|
580 |
|
967
|
581 ":" { |
|
1826
|
582 if (lexer_flags.plotting |
|
|
583 && (lexer_flags.in_plot_range || lexer_flags.in_plot_using)) |
|
2857
|
584 BIN_OP_RETURN (COLON, true); |
|
967
|
585 else |
|
2857
|
586 BIN_OP_RETURN (':', false); |
|
967
|
587 } |
|
1
|
588 |
|
968
|
589 %{ |
|
985
|
590 // Gobble comments. If closest nesting is inside parentheses, don't |
|
|
591 // return a new line. |
|
|
592 %} |
|
968
|
593 |
|
967
|
594 {CCHAR} { |
|
1826
|
595 if (help_buf.empty () |
|
|
596 && lexer_flags.beginning_of_function |
|
|
597 && nesting_level.none ()) |
|
967
|
598 { |
|
3665
|
599 lexer_flags.beginning_of_function = false; |
|
|
600 |
|
967
|
601 grab_help_text (); |
|
3665
|
602 |
|
|
603 octave_comment_buffer::append (help_buf); |
|
967
|
604 } |
|
|
605 else |
|
|
606 { |
|
3665
|
607 std::string buf; |
|
|
608 |
|
|
609 bool begin_comment = true; |
|
|
610 |
|
967
|
611 int c; |
|
|
612 while ((c = yyinput ()) != EOF && c != '\n') |
|
3665
|
613 { |
|
|
614 if (begin_comment && (c == '#' || c == '%')) |
|
|
615 ; /* Skip leading comment characters. */ |
|
|
616 else |
|
3802
|
617 buf += static_cast<char> (c); |
|
3665
|
618 } |
|
|
619 |
|
|
620 octave_comment_buffer::append (buf); |
|
967
|
621 } |
|
440
|
622 |
|
967
|
623 current_input_column = 1; |
|
2857
|
624 lexer_flags.quote_is_transpose = false; |
|
|
625 lexer_flags.cant_be_identifier = false; |
|
|
626 lexer_flags.convert_spaces_to_comma = true; |
|
985
|
627 |
|
4037
|
628 maybe_gripe_matlab_incompatible_comment (yytext[0]); |
|
|
629 |
|
1826
|
630 if (nesting_level.none ()) |
|
985
|
631 return '\n'; |
|
3351
|
632 else if (nesting_level.is_bracket ()) |
|
1566
|
633 return ';'; |
|
967
|
634 } |
|
440
|
635 |
|
968
|
636 %{ |
|
|
637 // Other operators. |
|
|
638 %} |
|
|
639 |
|
4037
|
640 ".+" { XBIN_OP_RETURN (EPLUS, false); } |
|
|
641 ".-" { XBIN_OP_RETURN (EMINUS, false); } |
|
2857
|
642 ".*" { BIN_OP_RETURN (EMUL, false); } |
|
|
643 "./" { BIN_OP_RETURN (EDIV, false); } |
|
|
644 ".\\" { BIN_OP_RETURN (ELEFTDIV, false); } |
|
4037
|
645 ".^" { BIN_OP_RETURN (EPOW, false); } |
|
|
646 ".**" { XBIN_OP_RETURN (EPOW, false); } |
|
2857
|
647 ".'" { do_comma_insert_check (); BIN_OP_RETURN (TRANSPOSE, true); } |
|
4037
|
648 "++" { do_comma_insert_check (); XBIN_OP_RETURN (PLUS_PLUS, true); } |
|
|
649 "--" { do_comma_insert_check (); XBIN_OP_RETURN (MINUS_MINUS, true); } |
|
2857
|
650 "<=" { BIN_OP_RETURN (EXPR_LE, false); } |
|
|
651 "==" { BIN_OP_RETURN (EXPR_EQ, false); } |
|
4037
|
652 "~=" { BIN_OP_RETURN (EXPR_NE, false); } |
|
|
653 "!=" { XBIN_OP_RETURN (EXPR_NE, false); } |
|
2857
|
654 ">=" { BIN_OP_RETURN (EXPR_GE, false); } |
|
2877
|
655 "&" { BIN_OP_RETURN (EXPR_AND, false); } |
|
2857
|
656 "|" { BIN_OP_RETURN (EXPR_OR, false); } |
|
|
657 "<" { BIN_OP_RETURN (EXPR_LT, false); } |
|
|
658 ">" { BIN_OP_RETURN (EXPR_GT, false); } |
|
|
659 "*" { BIN_OP_RETURN ('*', false); } |
|
|
660 "/" { BIN_OP_RETURN ('/', false); } |
|
|
661 "\\" { BIN_OP_RETURN (LEFTDIV, false); } |
|
|
662 ";" { BIN_OP_RETURN (';', true); } |
|
|
663 "," { BIN_OP_RETURN (',', true); } |
|
4037
|
664 "^" { BIN_OP_RETURN (POW, false); } |
|
|
665 "**" { XBIN_OP_RETURN (POW, false); } |
|
2857
|
666 "=" { BIN_OP_RETURN ('=', true); } |
|
2877
|
667 "&&" { BIN_OP_RETURN (EXPR_AND_AND, false); } |
|
2857
|
668 "||" { BIN_OP_RETURN (EXPR_OR_OR, false); } |
|
4037
|
669 "<<" { XBIN_OP_RETURN (LSHIFT, false); } |
|
|
670 ">>" { XBIN_OP_RETURN (RSHIFT, false); } |
|
967
|
671 |
|
|
672 {NOT} { |
|
1826
|
673 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
2857
|
674 lexer_flags.past_plot_range = true; |
|
4037
|
675 |
|
|
676 if (yytext[0] == '~') |
|
|
677 BIN_OP_RETURN (EXPR_NOT, false); |
|
|
678 else |
|
|
679 XBIN_OP_RETURN (EXPR_NOT, false); |
|
967
|
680 } |
|
1
|
681 |
|
1276
|
682 "+" { |
|
1826
|
683 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
2857
|
684 lexer_flags.past_plot_range = true; |
|
|
685 BIN_OP_RETURN ('+', false); |
|
967
|
686 } |
|
|
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 |
|
|
694 "(" { |
|
4131
|
695 lexer_flags.looking_at_indirect_ref = false; |
|
1826
|
696 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
2857
|
697 lexer_flags.past_plot_range = true; |
|
1826
|
698 nesting_level.paren (); |
|
985
|
699 promptflag--; |
|
967
|
700 TOK_RETURN ('('); |
|
|
701 } |
|
|
702 |
|
|
703 ")" { |
|
1826
|
704 nesting_level.remove (); |
|
1001
|
705 |
|
967
|
706 current_input_column++; |
|
2857
|
707 lexer_flags.cant_be_identifier = true; |
|
|
708 lexer_flags.quote_is_transpose = true; |
|
3351
|
709 lexer_flags.convert_spaces_to_comma = nesting_level.is_bracket (); |
|
1001
|
710 do_comma_insert_check (); |
|
967
|
711 return ')'; |
|
|
712 } |
|
|
713 |
|
2066
|
714 "." { |
|
|
715 TOK_RETURN ('.'); |
|
|
716 } |
|
|
717 |
|
3988
|
718 "@" { |
|
|
719 current_input_column++; |
|
|
720 lexer_flags.cant_be_identifier = false; |
|
|
721 lexer_flags.quote_is_transpose = false; |
|
|
722 lexer_flags.convert_spaces_to_comma = false; |
|
|
723 return '@'; |
|
|
724 } |
|
|
725 |
|
4037
|
726 "+=" { XBIN_OP_RETURN (ADD_EQ, false); } |
|
|
727 "-=" { XBIN_OP_RETURN (SUB_EQ, false); } |
|
|
728 "*=" { XBIN_OP_RETURN (MUL_EQ, false); } |
|
|
729 "/=" { XBIN_OP_RETURN (DIV_EQ, false); } |
|
|
730 "\\=" { XBIN_OP_RETURN (LEFTDIV_EQ, false); } |
|
|
731 ".+=" { XBIN_OP_RETURN (ADD_EQ, false); } |
|
|
732 ".-=" { XBIN_OP_RETURN (SUB_EQ, false); } |
|
|
733 ".*=" { XBIN_OP_RETURN (EMUL_EQ, false); } |
|
|
734 "./=" { XBIN_OP_RETURN (EDIV_EQ, false); } |
|
|
735 ".\\=" { XBIN_OP_RETURN (ELEFTDIV_EQ, false); } |
|
|
736 {POW}= { XBIN_OP_RETURN (POW_EQ, false); } |
|
|
737 {EPOW}= { XBIN_OP_RETURN (EPOW_EQ, false); } |
|
|
738 "&=" { XBIN_OP_RETURN (AND_EQ, false); } |
|
|
739 "|=" { XBIN_OP_RETURN (OR_EQ, false); } |
|
|
740 "<<=" { XBIN_OP_RETURN (LSHIFT_EQ, false); } |
|
|
741 ">>=" { XBIN_OP_RETURN (RSHIFT_EQ, false); } |
|
2877
|
742 |
|
3351
|
743 "{" { |
|
|
744 nesting_level.brace (); |
|
|
745 promptflag--; |
|
|
746 TOK_RETURN ('{'); |
|
|
747 } |
|
|
748 |
|
|
749 "}" { |
|
|
750 nesting_level.remove (); |
|
|
751 |
|
|
752 current_input_column++; |
|
|
753 lexer_flags.cant_be_identifier = true; |
|
|
754 lexer_flags.quote_is_transpose = true; |
|
|
755 lexer_flags.convert_spaces_to_comma = nesting_level.is_bracket (); |
|
|
756 do_comma_insert_check (); // Is this really necessary? |
|
|
757 |
|
|
758 return '}'; |
|
|
759 } |
|
|
760 |
|
968
|
761 %{ |
|
2066
|
762 // Unrecognized input is a lexical error. |
|
968
|
763 %} |
|
1
|
764 |
|
2042
|
765 . { |
|
2066
|
766 current_input_column++; |
|
|
767 |
|
3867
|
768 // EOF can't happen here (we catch it above). |
|
|
769 |
|
|
770 error ("invalid character `%s' (ASCII %d) near line %d, column %d", |
|
|
771 undo_string_escape (yytext[0]), static_cast<int> (yytext[0]), |
|
|
772 input_line_number, current_input_column); |
|
2066
|
773 |
|
|
774 return LEXICAL_ERROR; |
|
|
775 } |
|
1
|
776 |
|
|
777 %% |
|
|
778 |
|
767
|
779 // GAG. |
|
|
780 // |
|
|
781 // If we're reading a matrix and the next character is '[', make sure |
|
|
782 // that we insert a comma ahead of it. |
|
|
783 |
|
146
|
784 void |
|
1
|
785 do_comma_insert_check (void) |
|
|
786 { |
|
1001
|
787 int spc_gobbled = eat_continuation (); |
|
2970
|
788 |
|
1
|
789 int c = yyinput (); |
|
2970
|
790 |
|
3246
|
791 unput (c); |
|
2970
|
792 |
|
1001
|
793 if (spc_gobbled) |
|
3246
|
794 unput (' '); |
|
2970
|
795 |
|
3351
|
796 lexer_flags.do_comma_insert = (lexer_flags.bracketflag && c == '['); |
|
1
|
797 } |
|
|
798 |
|
767
|
799 // Fix things up for errors or interrupts. The parser is never called |
|
|
800 // recursively, so it is always safe to reinitialize its state before |
|
|
801 // doing any parsing. |
|
|
802 |
|
1
|
803 void |
|
|
804 reset_parser (void) |
|
|
805 { |
|
1826
|
806 // Start off on the right foot. |
|
1
|
807 BEGIN 0; |
|
166
|
808 error_state = 0; |
|
3489
|
809 warning_state = 0; |
|
3883
|
810 parser_end_of_input = false; |
|
287
|
811 |
|
1826
|
812 // We do want a prompt by default. |
|
1
|
813 promptflag = 1; |
|
287
|
814 |
|
3351
|
815 // Error may have occurred inside some brackets, braces, or parentheses. |
|
985
|
816 nesting_level.clear (); |
|
287
|
817 |
|
1826
|
818 // Clear out the stack of token info used to track line and column |
|
|
819 // numbers. |
|
143
|
820 while (! token_stack.empty ()) |
|
|
821 delete token_stack.pop (); |
|
287
|
822 |
|
1826
|
823 // Can be reset by defining a function. |
|
985
|
824 if (! (reading_script_file || reading_fcn_file)) |
|
|
825 { |
|
|
826 current_input_column = 1; |
|
2926
|
827 input_line_number = command_editor::current_command_number () - 1; |
|
985
|
828 } |
|
287
|
829 |
|
1826
|
830 // Only ask for input from stdin if we are expecting interactive |
|
|
831 // input. |
|
3174
|
832 if ((interactive || forced_interactive) |
|
3880
|
833 && ! (reading_fcn_file |
|
|
834 || reading_script_file |
|
|
835 || get_input_from_eval_string |
|
3174
|
836 || input_from_startup_file)) |
|
287
|
837 yyrestart (stdin); |
|
991
|
838 |
|
1826
|
839 // Clear the buffer for help text. |
|
|
840 help_buf.resize (0); |
|
1755
|
841 |
|
1826
|
842 // Reset other flags. |
|
|
843 lexer_flags.init (); |
|
1
|
844 } |
|
|
845 |
|
767
|
846 // If we read some newlines, we need figure out what column we're |
|
|
847 // really looking at. |
|
|
848 |
|
1
|
849 static void |
|
|
850 fixup_column_count (char *s) |
|
|
851 { |
|
|
852 char c; |
|
|
853 while ((c = *s++) != '\0') |
|
|
854 { |
|
|
855 if (c == '\n') |
|
143
|
856 current_input_column = 1; |
|
1
|
857 else |
|
|
858 current_input_column++; |
|
|
859 } |
|
|
860 } |
|
|
861 |
|
767
|
862 // Include these so that we don't have to link to libfl.a. |
|
246
|
863 |
|
1
|
864 #ifdef yywrap |
|
|
865 #undef yywrap |
|
|
866 #endif |
|
3332
|
867 int |
|
1
|
868 yywrap (void) |
|
|
869 { |
|
287
|
870 return 1; |
|
1
|
871 } |
|
|
872 |
|
767
|
873 // Tell us all what the current buffer is. |
|
|
874 |
|
1
|
875 YY_BUFFER_STATE |
|
|
876 current_buffer (void) |
|
|
877 { |
|
|
878 return YY_CURRENT_BUFFER; |
|
|
879 } |
|
|
880 |
|
767
|
881 // Create a new buffer. |
|
|
882 |
|
1
|
883 YY_BUFFER_STATE |
|
|
884 create_buffer (FILE *f) |
|
|
885 { |
|
|
886 return yy_create_buffer (f, YY_BUF_SIZE); |
|
|
887 } |
|
|
888 |
|
767
|
889 // Start reading a new buffer. |
|
|
890 |
|
1
|
891 void |
|
|
892 switch_to_buffer (YY_BUFFER_STATE buf) |
|
|
893 { |
|
|
894 yy_switch_to_buffer (buf); |
|
|
895 } |
|
|
896 |
|
767
|
897 // Delete a buffer. |
|
|
898 |
|
1
|
899 void |
|
|
900 delete_buffer (YY_BUFFER_STATE buf) |
|
|
901 { |
|
|
902 yy_delete_buffer (buf); |
|
|
903 } |
|
|
904 |
|
767
|
905 // Restore a buffer (for unwind-prot). |
|
|
906 |
|
1
|
907 void |
|
|
908 restore_input_buffer (void *buf) |
|
|
909 { |
|
2861
|
910 switch_to_buffer (static_cast<YY_BUFFER_STATE> (buf)); |
|
1
|
911 } |
|
|
912 |
|
767
|
913 // Delete a buffer (for unwind-prot). |
|
|
914 |
|
1
|
915 void |
|
|
916 delete_input_buffer (void *buf) |
|
|
917 { |
|
2861
|
918 delete_buffer (static_cast<YY_BUFFER_STATE> (buf)); |
|
1
|
919 } |
|
|
920 |
|
767
|
921 // Check to see if a character string matches any of the possible line |
|
|
922 // styles for plots. |
|
|
923 |
|
3536
|
924 static std::string |
|
3523
|
925 plot_style_token (const std::string& s) |
|
1
|
926 { |
|
3523
|
927 std::string retval; |
|
1823
|
928 |
|
3575
|
929 // XXX FIXME XXX -- specify minimum match length for these. |
|
2804
|
930 static const char *plot_styles[] = |
|
146
|
931 { |
|
925
|
932 "boxes", |
|
924
|
933 "boxerrorbars", |
|
2542
|
934 "boxxyerrorbars", |
|
|
935 "candlesticks", |
|
146
|
936 "dots", |
|
|
937 "errorbars", |
|
2542
|
938 "financebars", |
|
|
939 "fsteps", |
|
|
940 "histeps", |
|
146
|
941 "impulses", |
|
|
942 "lines", |
|
|
943 "linespoints", |
|
|
944 "points", |
|
926
|
945 "steps", |
|
2542
|
946 "vector", |
|
|
947 "xerrorbars", |
|
|
948 "xyerrorbars", |
|
|
949 "yerrorbars", |
|
522
|
950 0, |
|
146
|
951 }; |
|
|
952 |
|
2804
|
953 const char * const *tmp = plot_styles; |
|
522
|
954 while (*tmp) |
|
1
|
955 { |
|
1823
|
956 if (almost_match (*tmp, s.c_str ())) |
|
|
957 { |
|
|
958 retval = *tmp; |
|
|
959 break; |
|
|
960 } |
|
1
|
961 |
|
|
962 tmp++; |
|
|
963 } |
|
|
964 |
|
1823
|
965 return retval; |
|
1
|
966 } |
|
|
967 |
|
3165
|
968 // Check to see if a character string matches any of the possible axes |
|
|
969 // tags for plots. |
|
|
970 |
|
3536
|
971 static std::string |
|
3523
|
972 plot_axes_token (const std::string& s) |
|
3165
|
973 { |
|
3523
|
974 std::string retval; |
|
3165
|
975 |
|
3575
|
976 // XXX FIXME XXX -- specify minimum match length for these. |
|
3561
|
977 static const char *plot_axes[] = |
|
3165
|
978 { |
|
|
979 "x1y1", |
|
|
980 "x1y2", |
|
|
981 "x2y1", |
|
|
982 "x2y2", |
|
|
983 0, |
|
|
984 }; |
|
|
985 |
|
3561
|
986 const char **tmp = plot_axes; |
|
3165
|
987 while (*tmp) |
|
|
988 { |
|
|
989 if (almost_match (*tmp, s.c_str ())) |
|
|
990 { |
|
|
991 retval = *tmp; |
|
|
992 break; |
|
|
993 } |
|
|
994 |
|
|
995 tmp++; |
|
|
996 } |
|
|
997 |
|
|
998 return retval; |
|
|
999 } |
|
|
1000 |
|
767
|
1001 // Check to see if a character string matches any one of the plot |
|
941
|
1002 // option keywords. Don't match abbreviations for clear, since that's |
|
|
1003 // not a gnuplot keyword (users will probably only expect to be able |
|
|
1004 // to abbreviate actual gnuplot keywords). |
|
767
|
1005 |
|
1
|
1006 static int |
|
3523
|
1007 is_plot_keyword (const std::string& s) |
|
1
|
1008 { |
|
1823
|
1009 const char *t = s.c_str (); |
|
3575
|
1010 if (almost_match ("title", t, 1)) |
|
146
|
1011 { |
|
|
1012 return TITLE; |
|
|
1013 } |
|
3575
|
1014 else if (almost_match ("using", t, 1)) |
|
146
|
1015 { |
|
2857
|
1016 lexer_flags.in_plot_using = true; |
|
146
|
1017 return USING; |
|
|
1018 } |
|
3575
|
1019 else if (almost_match ("with", t, 1)) |
|
146
|
1020 { |
|
2857
|
1021 lexer_flags.in_plot_style = true; |
|
146
|
1022 return WITH; |
|
|
1023 } |
|
3575
|
1024 else if (almost_match ("axes", t, 2) || almost_match ("axis", t, 2)) |
|
3165
|
1025 { |
|
|
1026 lexer_flags.in_plot_axes = true; |
|
|
1027 return AXES; |
|
|
1028 } |
|
1823
|
1029 else if (strcmp ("clear", t) == 0) |
|
883
|
1030 { |
|
|
1031 return CLEAR; |
|
|
1032 } |
|
1
|
1033 else |
|
146
|
1034 { |
|
|
1035 return 0; |
|
|
1036 } |
|
1
|
1037 } |
|
|
1038 |
|
2877
|
1039 // Handle keywords. |
|
767
|
1040 |
|
1
|
1041 static int |
|
3523
|
1042 is_keyword (const std::string& s) |
|
1
|
1043 { |
|
3805
|
1044 int l = input_line_number; |
|
|
1045 int c = current_input_column; |
|
|
1046 |
|
3165
|
1047 if (lexer_flags.plotting) |
|
1
|
1048 { |
|
3165
|
1049 if (lexer_flags.in_plot_style) |
|
1
|
1050 { |
|
3523
|
1051 std::string sty = plot_style_token (s); |
|
3165
|
1052 |
|
|
1053 if (! sty.empty ()) |
|
|
1054 { |
|
|
1055 lexer_flags.in_plot_style = false; |
|
3805
|
1056 yylval.tok_val = new token (sty, l, c); |
|
3165
|
1057 token_stack.push (yylval.tok_val); |
|
|
1058 return STYLE; |
|
|
1059 } |
|
1
|
1060 } |
|
3165
|
1061 else if (lexer_flags.in_plot_axes) |
|
|
1062 { |
|
3523
|
1063 std::string axes = plot_axes_token (s); |
|
3165
|
1064 |
|
|
1065 if (! axes.empty ()) |
|
|
1066 { |
|
|
1067 lexer_flags.in_plot_axes = false; |
|
3805
|
1068 yylval.tok_val = new token (axes, l, c); |
|
3165
|
1069 token_stack.push (yylval.tok_val); |
|
|
1070 return AXES_TAG; |
|
|
1071 } |
|
|
1072 } |
|
1
|
1073 } |
|
|
1074 |
|
1823
|
1075 int len = s.length (); |
|
922
|
1076 |
|
1823
|
1077 const octave_kw *kw = octave_kw_lookup (s.c_str (), len); |
|
191
|
1078 |
|
1497
|
1079 if (kw) |
|
143
|
1080 { |
|
1497
|
1081 yylval.tok_val = 0; |
|
|
1082 |
|
|
1083 switch (kw->kw_id) |
|
|
1084 { |
|
|
1085 case all_va_args_kw: |
|
|
1086 case break_kw: |
|
2764
|
1087 case case_kw: |
|
1497
|
1088 case catch_kw: |
|
|
1089 case continue_kw: |
|
|
1090 case else_kw: |
|
|
1091 case elseif_kw: |
|
|
1092 case global_kw: |
|
2764
|
1093 case otherwise_kw: |
|
1497
|
1094 case return_kw: |
|
2846
|
1095 case static_kw: |
|
3484
|
1096 case until_kw: |
|
1497
|
1097 case unwind_protect_cleanup_kw: |
|
|
1098 break; |
|
|
1099 |
|
|
1100 case end_kw: |
|
|
1101 yylval.tok_val = new token (token::simple_end, l, c); |
|
|
1102 break; |
|
|
1103 |
|
|
1104 case end_try_catch_kw: |
|
|
1105 yylval.tok_val = new token (token::try_catch_end, l, c); |
|
|
1106 break; |
|
|
1107 |
|
|
1108 case end_unwind_protect_kw: |
|
|
1109 yylval.tok_val = new token (token::unwind_protect_end, l, c); |
|
|
1110 break; |
|
|
1111 |
|
|
1112 case endfor_kw: |
|
|
1113 yylval.tok_val = new token (token::for_end, l, c); |
|
|
1114 break; |
|
|
1115 |
|
|
1116 case endfunction_kw: |
|
|
1117 yylval.tok_val = new token (token::function_end, l, c); |
|
|
1118 break; |
|
|
1119 |
|
|
1120 case endif_kw: |
|
|
1121 yylval.tok_val = new token (token::if_end, l, c); |
|
|
1122 break; |
|
|
1123 |
|
2764
|
1124 case endswitch_kw: |
|
|
1125 yylval.tok_val = new token (token::switch_end, l, c); |
|
|
1126 break; |
|
|
1127 |
|
1497
|
1128 case endwhile_kw: |
|
|
1129 yylval.tok_val = new token (token::while_end, l, c); |
|
|
1130 break; |
|
|
1131 |
|
3484
|
1132 case do_kw: |
|
1497
|
1133 case for_kw: |
|
|
1134 case while_kw: |
|
|
1135 promptflag--; |
|
1826
|
1136 lexer_flags.looping++; |
|
1497
|
1137 break; |
|
|
1138 |
|
|
1139 case if_kw: |
|
|
1140 case try_kw: |
|
2764
|
1141 case switch_kw: |
|
1497
|
1142 case unwind_protect_kw: |
|
|
1143 promptflag--; |
|
|
1144 break; |
|
|
1145 |
|
|
1146 case gplot_kw: |
|
2857
|
1147 lexer_flags.plotting = true; |
|
1497
|
1148 yylval.tok_val = new token (token::two_dee, l, c); |
|
|
1149 break; |
|
|
1150 |
|
|
1151 case gsplot_kw: |
|
2857
|
1152 lexer_flags.plotting = true; |
|
1497
|
1153 yylval.tok_val = new token (token::three_dee, l, c); |
|
|
1154 break; |
|
|
1155 |
|
|
1156 case replot_kw: |
|
2857
|
1157 lexer_flags.plotting = true; |
|
1497
|
1158 yylval.tok_val = new token (token::replot, l, c); |
|
|
1159 break; |
|
|
1160 |
|
|
1161 case function_kw: |
|
1826
|
1162 if (lexer_flags.defining_func) |
|
1497
|
1163 { |
|
|
1164 error ("function keyword invalid within a function body"); |
|
|
1165 |
|
|
1166 if ((reading_fcn_file || reading_script_file) |
|
1755
|
1167 && ! curr_fcn_file_name.empty ()) |
|
1497
|
1168 error ("defining new function near line %d of file `%s.m'", |
|
1755
|
1169 input_line_number, curr_fcn_file_name.c_str ()); |
|
1497
|
1170 else |
|
1755
|
1171 error ("defining new function near line %d", |
|
|
1172 input_line_number); |
|
1497
|
1173 |
|
|
1174 return LEXICAL_ERROR; |
|
|
1175 } |
|
|
1176 else |
|
|
1177 { |
|
2877
|
1178 // Prepare for local symbols. |
|
|
1179 |
|
1497
|
1180 tmp_local_sym_tab = new symbol_table (); |
|
2877
|
1181 |
|
|
1182 promptflag--; |
|
|
1183 |
|
2857
|
1184 lexer_flags.defining_func = true; |
|
2877
|
1185 lexer_flags.parsed_function_name = false; |
|
2857
|
1186 lexer_flags.beginning_of_function = true; |
|
2877
|
1187 |
|
1497
|
1188 if (! (reading_fcn_file || reading_script_file)) |
|
|
1189 input_line_number = 1; |
|
|
1190 } |
|
|
1191 break; |
|
|
1192 |
|
3174
|
1193 case magic_file_kw: |
|
|
1194 { |
|
|
1195 if ((reading_fcn_file || reading_script_file) |
|
|
1196 && ! curr_fcn_file_full_name.empty ()) |
|
|
1197 yylval.tok_val = new token (curr_fcn_file_full_name, l, c); |
|
|
1198 else |
|
|
1199 yylval.tok_val = new token ("stdin", l, c); |
|
|
1200 } |
|
|
1201 break; |
|
|
1202 |
|
|
1203 case magic_line_kw: |
|
|
1204 yylval.tok_val = new token (static_cast<double> (l), "", l, c); |
|
|
1205 break; |
|
|
1206 |
|
3974
|
1207 case varargin_kw: |
|
|
1208 if (! lexer_flags.looking_at_parameter_list) |
|
|
1209 return 0; |
|
|
1210 break; |
|
|
1211 |
|
|
1212 case varargout_kw: |
|
|
1213 if (! lexer_flags.looking_at_return_list) |
|
|
1214 return 0; |
|
|
1215 break; |
|
|
1216 |
|
1497
|
1217 default: |
|
|
1218 panic_impossible (); |
|
|
1219 } |
|
|
1220 |
|
|
1221 if (! yylval.tok_val) |
|
|
1222 yylval.tok_val = new token (l, c); |
|
|
1223 |
|
476
|
1224 token_stack.push (yylval.tok_val); |
|
1497
|
1225 |
|
|
1226 return kw->tok; |
|
143
|
1227 } |
|
1
|
1228 |
|
|
1229 return 0; |
|
|
1230 } |
|
|
1231 |
|
767
|
1232 // Try to find an identifier. All binding to global or builtin |
|
|
1233 // variables occurs when expressions are evaluated. |
|
|
1234 |
|
1
|
1235 static symbol_record * |
|
3523
|
1236 lookup_identifier (const std::string& name) |
|
1
|
1237 { |
|
2856
|
1238 return curr_sym_tab->lookup (name, true); |
|
1
|
1239 } |
|
|
1240 |
|
2702
|
1241 static bool |
|
3523
|
1242 is_variable (const std::string& name) |
|
2702
|
1243 { |
|
2856
|
1244 symbol_record *sr = curr_sym_tab->lookup (name); |
|
2702
|
1245 |
|
|
1246 return sr && sr->is_variable (); |
|
|
1247 } |
|
|
1248 |
|
|
1249 static void |
|
3523
|
1250 force_local_variable (const std::string& name) |
|
2702
|
1251 { |
|
4101
|
1252 if (! is_variable (name)) |
|
|
1253 curr_sym_tab->clear (name); |
|
|
1254 |
|
2856
|
1255 symbol_record *sr = curr_sym_tab->lookup (name, true); |
|
2702
|
1256 |
|
|
1257 if (sr) |
|
|
1258 sr->define (octave_value ()); |
|
|
1259 } |
|
|
1260 |
|
1019
|
1261 // Grab the help text from an function file. Always overwrites the |
|
|
1262 // current contents of help_buf. |
|
767
|
1263 |
|
3427
|
1264 // XXX FIXME XXX -- gobble_leading_white_space() in parse.y |
|
2300
|
1265 // duplicates some of this code! |
|
|
1266 |
|
1
|
1267 static void |
|
|
1268 grab_help_text (void) |
|
|
1269 { |
|
1755
|
1270 help_buf.resize (0); |
|
1019
|
1271 |
|
2300
|
1272 bool begin_comment = true; |
|
|
1273 bool in_comment = true; |
|
3427
|
1274 bool discard_space = true; |
|
3665
|
1275 |
|
1019
|
1276 int c = 0; |
|
1
|
1277 |
|
1019
|
1278 while ((c = yyinput ()) != EOF) |
|
|
1279 { |
|
2300
|
1280 if (begin_comment) |
|
|
1281 { |
|
|
1282 if (c == '%' || c == '#') |
|
|
1283 continue; |
|
3427
|
1284 else if (discard_space && c == ' ') |
|
|
1285 { |
|
|
1286 discard_space = false; |
|
|
1287 continue; |
|
|
1288 } |
|
2300
|
1289 else |
|
|
1290 begin_comment = false; |
|
|
1291 } |
|
|
1292 |
|
1019
|
1293 if (in_comment) |
|
1
|
1294 { |
|
3802
|
1295 help_buf += static_cast<char> (c); |
|
1755
|
1296 |
|
1019
|
1297 if (c == '\n') |
|
3427
|
1298 { |
|
|
1299 in_comment = false; |
|
|
1300 discard_space = true; |
|
|
1301 } |
|
1019
|
1302 } |
|
|
1303 else |
|
|
1304 { |
|
|
1305 switch (c) |
|
991
|
1306 { |
|
4037
|
1307 case '#': |
|
1019
|
1308 case '%': |
|
4037
|
1309 maybe_gripe_matlab_incompatible_comment (yytext[0]); |
|
2300
|
1310 in_comment = true; |
|
|
1311 begin_comment = true; |
|
1019
|
1312 break; |
|
777
|
1313 |
|
1019
|
1314 case ' ': |
|
|
1315 case '\t': |
|
|
1316 break; |
|
777
|
1317 |
|
1019
|
1318 default: |
|
|
1319 goto done; |
|
1
|
1320 } |
|
|
1321 } |
|
1019
|
1322 } |
|
991
|
1323 |
|
1019
|
1324 done: |
|
991
|
1325 |
|
1019
|
1326 if (c) |
|
3246
|
1327 unput (c); |
|
1
|
1328 } |
|
|
1329 |
|
767
|
1330 // Return 1 if the given character matches any character in the given |
|
|
1331 // string. |
|
|
1332 |
|
2857
|
1333 static bool |
|
2804
|
1334 match_any (char c, const char *s) |
|
1
|
1335 { |
|
|
1336 char tmp; |
|
|
1337 while ((tmp = *s++) != '\0') |
|
|
1338 { |
|
|
1339 if (c == tmp) |
|
2857
|
1340 return true; |
|
1
|
1341 } |
|
2857
|
1342 return false; |
|
1
|
1343 } |
|
|
1344 |
|
767
|
1345 // Given information about the spacing surrounding an operator, |
|
|
1346 // return 1 if it looks like it should be treated as a binary |
|
|
1347 // operator. For example, |
|
|
1348 // |
|
3774
|
1349 // [ 1 + 2 ] or [ 1+ 2] or [ 1+2 ] ==> binary |
|
|
1350 // |
|
|
1351 // [ 1 +2 ] ==> unary |
|
767
|
1352 |
|
2857
|
1353 static bool |
|
3246
|
1354 looks_like_bin_op (bool spc_prev, int next_char) |
|
1
|
1355 { |
|
3246
|
1356 bool spc_next = (next_char == ' ' || next_char == '\t'); |
|
|
1357 |
|
608
|
1358 return ((spc_prev && spc_next) || ! spc_prev); |
|
1
|
1359 } |
|
|
1360 |
|
3263
|
1361 // Recognize separators. If the separator is a CRLF pair, it is |
|
|
1362 // replaced by a single LF. |
|
|
1363 |
|
|
1364 static bool |
|
|
1365 next_token_is_sep_op (void) |
|
|
1366 { |
|
|
1367 bool retval = false; |
|
|
1368 |
|
|
1369 int c1 = yyinput (); |
|
|
1370 |
|
|
1371 if (c1 == '\r') |
|
|
1372 { |
|
|
1373 int c2 = yyinput (); |
|
|
1374 |
|
|
1375 if (c2 == '\n') |
|
|
1376 { |
|
|
1377 c1 = '\n'; |
|
|
1378 |
|
|
1379 retval = true; |
|
|
1380 } |
|
|
1381 else |
|
|
1382 unput (c2); |
|
|
1383 } |
|
|
1384 else |
|
|
1385 retval = match_any (c1, ",;\n]"); |
|
|
1386 |
|
|
1387 unput (c1); |
|
|
1388 |
|
|
1389 return retval; |
|
|
1390 } |
|
|
1391 |
|
767
|
1392 // Try to determine if the next token should be treated as a postfix |
|
|
1393 // unary operator. This is ugly, but it seems to do the right thing. |
|
|
1394 |
|
2857
|
1395 static bool |
|
3246
|
1396 next_token_is_postfix_unary_op (bool spc_prev) |
|
1
|
1397 { |
|
2857
|
1398 bool un_op = false; |
|
1
|
1399 |
|
|
1400 int c0 = yyinput (); |
|
|
1401 |
|
3246
|
1402 if (c0 == '\'' && ! spc_prev) |
|
|
1403 { |
|
|
1404 un_op = true; |
|
|
1405 } |
|
|
1406 else if (c0 == '.') |
|
|
1407 { |
|
|
1408 int c1 = yyinput (); |
|
|
1409 un_op = (c1 == '\''); |
|
|
1410 unput (c1); |
|
|
1411 } |
|
1
|
1412 |
|
3246
|
1413 unput (c0); |
|
1
|
1414 |
|
|
1415 return un_op; |
|
|
1416 } |
|
|
1417 |
|
767
|
1418 // Try to determine if the next token should be treated as a binary |
|
3246
|
1419 // operator. |
|
1521
|
1420 // |
|
3246
|
1421 // This kluge exists because whitespace is not always ignored inside |
|
3774
|
1422 // the square brackets that are used to create matrix objects (though |
|
|
1423 // spacing only really matters in the cases that can be interpreted |
|
|
1424 // either as binary ops or prefix unary ops: currently just +, -). |
|
|
1425 // |
|
3779
|
1426 // Note that a line continuation directly following a + or - operator |
|
|
1427 // (e.g., the characters '[' 'a' ' ' '+' '\' LFD 'b' ']') will be |
|
|
1428 // parsed as a binary operator. |
|
767
|
1429 |
|
2857
|
1430 static bool |
|
3246
|
1431 next_token_is_bin_op (bool spc_prev) |
|
1
|
1432 { |
|
2857
|
1433 bool bin_op = false; |
|
1
|
1434 |
|
|
1435 int c0 = yyinput (); |
|
|
1436 |
|
|
1437 switch (c0) |
|
|
1438 { |
|
777
|
1439 case '+': |
|
|
1440 case '-': |
|
3774
|
1441 { |
|
|
1442 int c1 = yyinput (); |
|
|
1443 |
|
|
1444 switch (c1) |
|
|
1445 { |
|
|
1446 case '+': |
|
|
1447 case '-': |
|
|
1448 // Unary ops, spacing doesn't matter. |
|
|
1449 break; |
|
|
1450 |
|
|
1451 case '=': |
|
|
1452 // Binary ops, spacing doesn't matter. |
|
|
1453 bin_op = true; |
|
|
1454 break; |
|
|
1455 |
|
|
1456 default: |
|
|
1457 // Could be either, spacing matters. |
|
|
1458 bin_op = looks_like_bin_op (spc_prev, c1); |
|
|
1459 break; |
|
|
1460 } |
|
|
1461 |
|
|
1462 unput (c1); |
|
|
1463 } |
|
|
1464 break; |
|
|
1465 |
|
|
1466 case ':': |
|
3246
|
1467 case '/': |
|
|
1468 case '\\': |
|
|
1469 case '^': |
|
3774
|
1470 // Always a binary op (may also include /=, \=, and ^=). |
|
|
1471 bin_op = true; |
|
1276
|
1472 break; |
|
|
1473 |
|
3246
|
1474 // .+ .- ./ .\ .^ .* .** |
|
1554
|
1475 case '.': |
|
|
1476 { |
|
|
1477 int c1 = yyinput (); |
|
3246
|
1478 |
|
3774
|
1479 if (match_any (c1, "+-/\\^*")) |
|
|
1480 // Always a binary op (may also include .+=, .-=, ./=, ...). |
|
|
1481 bin_op = true; |
|
3698
|
1482 else if (! isdigit (c1) && c1 != ' ' && c1 != '\t' && c1 != '.') |
|
3774
|
1483 // A structure element reference is a binary op. |
|
|
1484 bin_op = true; |
|
3246
|
1485 |
|
|
1486 unput (c1); |
|
1554
|
1487 } |
|
|
1488 break; |
|
|
1489 |
|
3246
|
1490 // = == & && | || * ** |
|
|
1491 case '=': |
|
1
|
1492 case '&': |
|
3246
|
1493 case '|': |
|
1
|
1494 case '*': |
|
3774
|
1495 // Always a binary op (may also include ==, &&, ||, **). |
|
|
1496 bin_op = true; |
|
3246
|
1497 break; |
|
|
1498 |
|
3774
|
1499 // < <= <> > >= |
|
1
|
1500 case '<': |
|
|
1501 case '>': |
|
3774
|
1502 // Always a binary op (may also include <=, <>, >=). |
|
|
1503 bin_op = true; |
|
|
1504 break; |
|
|
1505 |
|
|
1506 // ~= != |
|
777
|
1507 case '~': |
|
|
1508 case '!': |
|
3246
|
1509 { |
|
|
1510 int c1 = yyinput (); |
|
|
1511 |
|
3774
|
1512 // ~ and ! can be unary ops, so require following =. |
|
|
1513 if (c1 == '=') |
|
|
1514 bin_op = true; |
|
3246
|
1515 |
|
|
1516 unput (c1); |
|
|
1517 } |
|
1
|
1518 break; |
|
|
1519 |
|
|
1520 default: |
|
1276
|
1521 break; |
|
1
|
1522 } |
|
|
1523 |
|
3246
|
1524 unput (c0); |
|
1
|
1525 |
|
|
1526 return bin_op; |
|
|
1527 } |
|
|
1528 |
|
767
|
1529 // Used to delete trailing white space from tokens. |
|
|
1530 |
|
3536
|
1531 static std::string |
|
1
|
1532 strip_trailing_whitespace (char *s) |
|
|
1533 { |
|
3523
|
1534 std::string retval = s; |
|
1
|
1535 |
|
1823
|
1536 size_t pos = retval.find_first_of (" \t"); |
|
1
|
1537 |
|
1823
|
1538 if (pos != NPOS) |
|
|
1539 retval.resize (pos); |
|
1
|
1540 |
|
|
1541 return retval; |
|
|
1542 } |
|
|
1543 |
|
3665
|
1544 static void |
|
|
1545 scan_for_comments (const char *text) |
|
|
1546 { |
|
|
1547 std::string comment_buf; |
|
|
1548 |
|
|
1549 bool in_comment = false; |
|
|
1550 bool beginning_of_comment = false; |
|
|
1551 |
|
|
1552 int len = strlen (text); |
|
|
1553 int i = 0; |
|
|
1554 |
|
|
1555 while (i < len) |
|
|
1556 { |
|
|
1557 char c = text[i++]; |
|
|
1558 |
|
|
1559 switch (c) |
|
|
1560 { |
|
|
1561 case '%': |
|
|
1562 case '#': |
|
|
1563 if (in_comment) |
|
|
1564 { |
|
|
1565 if (! beginning_of_comment) |
|
3802
|
1566 comment_buf += static_cast<char> (c); |
|
3665
|
1567 } |
|
|
1568 else |
|
|
1569 { |
|
4037
|
1570 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
1571 in_comment = true; |
|
|
1572 beginning_of_comment = true; |
|
|
1573 } |
|
|
1574 break; |
|
|
1575 |
|
|
1576 case '\n': |
|
|
1577 if (in_comment) |
|
|
1578 { |
|
3802
|
1579 comment_buf += static_cast<char> (c); |
|
3665
|
1580 octave_comment_buffer::append (comment_buf); |
|
|
1581 comment_buf.resize (0); |
|
|
1582 in_comment = false; |
|
|
1583 beginning_of_comment = false; |
|
|
1584 } |
|
|
1585 break; |
|
|
1586 |
|
|
1587 case '\r': |
|
|
1588 if (in_comment) |
|
3802
|
1589 comment_buf += static_cast<char> (c); |
|
3665
|
1590 if (i < len) |
|
|
1591 { |
|
|
1592 c = text[i++]; |
|
|
1593 |
|
|
1594 if (c == '\n') |
|
|
1595 { |
|
|
1596 if (in_comment) |
|
|
1597 { |
|
3802
|
1598 comment_buf += static_cast<char> (c); |
|
3665
|
1599 octave_comment_buffer::append (comment_buf); |
|
|
1600 in_comment = false; |
|
|
1601 beginning_of_comment = false; |
|
|
1602 } |
|
|
1603 } |
|
|
1604 } |
|
|
1605 |
|
|
1606 default: |
|
|
1607 if (in_comment) |
|
|
1608 { |
|
3802
|
1609 comment_buf += static_cast<char> (c); |
|
3665
|
1610 beginning_of_comment = false; |
|
|
1611 } |
|
|
1612 break; |
|
|
1613 } |
|
|
1614 } |
|
|
1615 |
|
|
1616 if (! comment_buf.empty ()) |
|
|
1617 octave_comment_buffer::append (comment_buf); |
|
|
1618 } |
|
|
1619 |
|
1001
|
1620 // Discard whitespace, including comments and continuations. |
|
1088
|
1621 // |
|
|
1622 // Return value is logical OR of the following values: |
|
|
1623 // |
|
1826
|
1624 // ATE_NOTHING : no spaces to eat |
|
1088
|
1625 // ATE_SPACE_OR_TAB : space or tab in input |
|
|
1626 // ATE_NEWLINE : bare new line in input |
|
1001
|
1627 |
|
1826
|
1628 static yum_yum |
|
975
|
1629 eat_whitespace (void) |
|
|
1630 { |
|
1826
|
1631 yum_yum retval = ATE_NOTHING; |
|
3665
|
1632 |
|
|
1633 std::string comment_buf; |
|
|
1634 |
|
2857
|
1635 bool in_comment = false; |
|
3665
|
1636 bool beginning_of_comment = false; |
|
|
1637 |
|
|
1638 int c = 0; |
|
|
1639 |
|
975
|
1640 while ((c = yyinput ()) != EOF) |
|
|
1641 { |
|
|
1642 current_input_column++; |
|
|
1643 |
|
|
1644 switch (c) |
|
|
1645 { |
|
|
1646 case ' ': |
|
|
1647 case '\t': |
|
3665
|
1648 if (in_comment) |
|
|
1649 { |
|
3802
|
1650 comment_buf += static_cast<char> (c); |
|
3665
|
1651 beginning_of_comment = false; |
|
|
1652 } |
|
1088
|
1653 retval |= ATE_SPACE_OR_TAB; |
|
975
|
1654 break; |
|
|
1655 |
|
|
1656 case '\n': |
|
1088
|
1657 retval |= ATE_NEWLINE; |
|
3665
|
1658 if (in_comment) |
|
|
1659 { |
|
3802
|
1660 comment_buf += static_cast<char> (c); |
|
3665
|
1661 octave_comment_buffer::append (comment_buf); |
|
|
1662 comment_buf.resize (0); |
|
|
1663 in_comment = false; |
|
|
1664 beginning_of_comment = false; |
|
|
1665 } |
|
975
|
1666 current_input_column = 0; |
|
|
1667 break; |
|
|
1668 |
|
|
1669 case '#': |
|
|
1670 case '%': |
|
3665
|
1671 if (in_comment) |
|
|
1672 { |
|
|
1673 if (! beginning_of_comment) |
|
3802
|
1674 comment_buf += static_cast<char> (c); |
|
3665
|
1675 } |
|
|
1676 else |
|
|
1677 { |
|
4037
|
1678 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
1679 in_comment = true; |
|
|
1680 beginning_of_comment = true; |
|
|
1681 } |
|
975
|
1682 break; |
|
|
1683 |
|
1001
|
1684 case '.': |
|
|
1685 if (in_comment) |
|
3665
|
1686 { |
|
3802
|
1687 comment_buf += static_cast<char> (c); |
|
3665
|
1688 beginning_of_comment = false; |
|
|
1689 break; |
|
|
1690 } |
|
1001
|
1691 else |
|
|
1692 { |
|
|
1693 if (have_ellipsis_continuation ()) |
|
|
1694 break; |
|
|
1695 else |
|
|
1696 goto done; |
|
|
1697 } |
|
|
1698 |
|
|
1699 case '\\': |
|
|
1700 if (in_comment) |
|
3665
|
1701 { |
|
3802
|
1702 comment_buf += static_cast<char> (c); |
|
3665
|
1703 beginning_of_comment = false; |
|
|
1704 break; |
|
|
1705 } |
|
1001
|
1706 else |
|
|
1707 { |
|
3105
|
1708 if (have_continuation ()) |
|
1001
|
1709 break; |
|
|
1710 else |
|
|
1711 goto done; |
|
|
1712 } |
|
|
1713 |
|
975
|
1714 default: |
|
|
1715 if (in_comment) |
|
3665
|
1716 { |
|
3802
|
1717 comment_buf += static_cast<char> (c); |
|
3665
|
1718 beginning_of_comment = false; |
|
|
1719 break; |
|
|
1720 } |
|
975
|
1721 else |
|
|
1722 goto done; |
|
|
1723 } |
|
|
1724 } |
|
|
1725 |
|
3665
|
1726 if (! comment_buf.empty ()) |
|
|
1727 octave_comment_buffer::append (comment_buf); |
|
|
1728 |
|
975
|
1729 done: |
|
3246
|
1730 unput (c); |
|
1082
|
1731 current_input_column--; |
|
1001
|
1732 return retval; |
|
975
|
1733 } |
|
|
1734 |
|
3220
|
1735 static inline bool |
|
|
1736 looks_like_hex (const char *s, int len) |
|
|
1737 { |
|
|
1738 return (len > 2 && s[0] == '0' && (s[1] == 'x' || s[1] == 'X')); |
|
|
1739 } |
|
|
1740 |
|
975
|
1741 static void |
|
3246
|
1742 handle_number (void) |
|
972
|
1743 { |
|
3220
|
1744 double value = 0.0; |
|
|
1745 int nread = 0; |
|
|
1746 |
|
3598
|
1747 if (looks_like_hex (yytext, strlen (yytext))) |
|
3220
|
1748 { |
|
|
1749 unsigned long ival; |
|
3598
|
1750 |
|
|
1751 nread = sscanf (yytext, "%lx", &ival); |
|
|
1752 |
|
3220
|
1753 value = static_cast<double> (ival); |
|
|
1754 } |
|
|
1755 else |
|
3598
|
1756 { |
|
|
1757 char *tmp = strsave (yytext); |
|
|
1758 |
|
|
1759 char *idx = strpbrk (tmp, "Dd"); |
|
2621
|
1760 |
|
3598
|
1761 if (idx) |
|
|
1762 *idx = 'e'; |
|
|
1763 |
|
|
1764 nread = sscanf (tmp, "%lf", &value); |
|
|
1765 |
|
|
1766 delete [] tmp; |
|
|
1767 } |
|
972
|
1768 |
|
1826
|
1769 // If yytext doesn't contain a valid number, we are in deep doo doo. |
|
985
|
1770 |
|
972
|
1771 assert (nread == 1); |
|
|
1772 |
|
3988
|
1773 lexer_flags.quote_is_transpose = true; |
|
|
1774 lexer_flags.cant_be_identifier = true; |
|
|
1775 lexer_flags.convert_spaces_to_comma = true; |
|
972
|
1776 |
|
1826
|
1777 if (lexer_flags.plotting && ! lexer_flags.in_plot_range) |
|
3575
|
1778 lexer_flags.past_plot_range = true; |
|
972
|
1779 |
|
|
1780 yylval.tok_val = new token (value, yytext, input_line_number, |
|
|
1781 current_input_column); |
|
|
1782 |
|
|
1783 token_stack.push (yylval.tok_val); |
|
|
1784 |
|
|
1785 current_input_column += yyleng; |
|
|
1786 |
|
|
1787 do_comma_insert_check (); |
|
|
1788 } |
|
|
1789 |
|
1001
|
1790 // We have seen a backslash and need to find out if it should be |
|
|
1791 // treated as a continuation character. If so, this eats it, up to |
|
|
1792 // and including the new line character. |
|
|
1793 // |
|
973
|
1794 // Match whitespace only, followed by a comment character or newline. |
|
|
1795 // Once a comment character is found, discard all input until newline. |
|
|
1796 // If non-whitespace characters are found before comment |
|
|
1797 // characters, return 0. Otherwise, return 1. |
|
|
1798 |
|
2857
|
1799 static bool |
|
3096
|
1800 have_continuation (bool trailing_comments_ok) |
|
973
|
1801 { |
|
4051
|
1802 OSSTREAM buf; |
|
973
|
1803 |
|
3665
|
1804 std::string comment_buf; |
|
|
1805 |
|
2857
|
1806 bool in_comment = false; |
|
3665
|
1807 bool beginning_of_comment = false; |
|
|
1808 |
|
|
1809 int c = 0; |
|
|
1810 |
|
973
|
1811 while ((c = yyinput ()) != EOF) |
|
|
1812 { |
|
3802
|
1813 buf << static_cast<char> (c); |
|
973
|
1814 |
|
|
1815 switch (c) |
|
|
1816 { |
|
|
1817 case ' ': |
|
|
1818 case '\t': |
|
3665
|
1819 if (in_comment) |
|
|
1820 { |
|
3802
|
1821 comment_buf += static_cast<char> (c); |
|
3665
|
1822 beginning_of_comment = false; |
|
|
1823 } |
|
973
|
1824 break; |
|
|
1825 |
|
|
1826 case '%': |
|
|
1827 case '#': |
|
1091
|
1828 if (trailing_comments_ok) |
|
3665
|
1829 { |
|
|
1830 if (in_comment) |
|
|
1831 { |
|
|
1832 if (! beginning_of_comment) |
|
3802
|
1833 comment_buf += static_cast<char> (c); |
|
3665
|
1834 } |
|
|
1835 else |
|
|
1836 { |
|
4037
|
1837 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
1838 in_comment = true; |
|
|
1839 beginning_of_comment = true; |
|
|
1840 } |
|
|
1841 } |
|
1091
|
1842 else |
|
|
1843 goto cleanup; |
|
973
|
1844 break; |
|
|
1845 |
|
|
1846 case '\n': |
|
3665
|
1847 if (in_comment) |
|
|
1848 { |
|
3802
|
1849 comment_buf += static_cast<char> (c); |
|
3665
|
1850 octave_comment_buffer::append (comment_buf); |
|
|
1851 } |
|
975
|
1852 current_input_column = 0; |
|
1001
|
1853 promptflag--; |
|
4037
|
1854 gripe_matlab_incompatible_continuation (); |
|
2857
|
1855 return true; |
|
973
|
1856 |
|
3263
|
1857 case '\r': |
|
3665
|
1858 if (in_comment) |
|
3802
|
1859 comment_buf += static_cast<char> (c); |
|
3263
|
1860 c = yyinput (); |
|
|
1861 if (c == EOF) |
|
|
1862 break; |
|
|
1863 else if (c == '\n') |
|
|
1864 { |
|
3665
|
1865 if (in_comment) |
|
|
1866 { |
|
3802
|
1867 comment_buf += static_cast<char> (c); |
|
3665
|
1868 octave_comment_buffer::append (comment_buf); |
|
|
1869 } |
|
3263
|
1870 current_input_column = 0; |
|
|
1871 promptflag--; |
|
4037
|
1872 gripe_matlab_incompatible_continuation (); |
|
3263
|
1873 return true; |
|
3665
|
1874 } |
|
3263
|
1875 |
|
3802
|
1876 // Fall through... |
|
|
1877 |
|
973
|
1878 default: |
|
3665
|
1879 if (in_comment) |
|
|
1880 { |
|
3802
|
1881 comment_buf += static_cast<char> (c); |
|
3665
|
1882 beginning_of_comment = false; |
|
|
1883 } |
|
|
1884 else |
|
1091
|
1885 goto cleanup; |
|
|
1886 break; |
|
973
|
1887 } |
|
|
1888 } |
|
|
1889 |
|
3246
|
1890 unput (c); |
|
2857
|
1891 return false; |
|
973
|
1892 |
|
3096
|
1893 cleanup: |
|
4051
|
1894 |
|
|
1895 buf << OSSTREAM_ENDS; |
|
|
1896 std::string s = OSSTREAM_STR (buf); |
|
|
1897 OSSTREAM_FREEZE (buf); |
|
|
1898 |
|
|
1899 int len = s.length (); |
|
|
1900 while (len--) |
|
|
1901 unput (s[len]); |
|
3096
|
1902 |
|
2857
|
1903 return false; |
|
973
|
1904 } |
|
|
1905 |
|
1001
|
1906 // We have seen a `.' and need to see if it is the start of a |
|
|
1907 // continuation. If so, this eats it, up to and including the new |
|
|
1908 // line character. |
|
|
1909 |
|
2857
|
1910 static bool |
|
3096
|
1911 have_ellipsis_continuation (bool trailing_comments_ok) |
|
973
|
1912 { |
|
|
1913 char c1 = yyinput (); |
|
|
1914 if (c1 == '.') |
|
|
1915 { |
|
|
1916 char c2 = yyinput (); |
|
1091
|
1917 if (c2 == '.' && have_continuation (trailing_comments_ok)) |
|
2857
|
1918 return true; |
|
973
|
1919 else |
|
|
1920 { |
|
3246
|
1921 unput (c2); |
|
|
1922 unput (c1); |
|
973
|
1923 } |
|
|
1924 } |
|
|
1925 else |
|
3246
|
1926 unput (c1); |
|
973
|
1927 |
|
2857
|
1928 return false; |
|
973
|
1929 } |
|
|
1930 |
|
1001
|
1931 // See if we have a continuation line. If so, eat it and the leading |
|
|
1932 // whitespace on the next line. |
|
1088
|
1933 // |
|
|
1934 // Return value is the same as described for eat_whitespace(). |
|
1001
|
1935 |
|
1826
|
1936 static yum_yum |
|
1001
|
1937 eat_continuation (void) |
|
|
1938 { |
|
1826
|
1939 int retval = ATE_NOTHING; |
|
3665
|
1940 |
|
1001
|
1941 int c = yyinput (); |
|
3665
|
1942 |
|
1001
|
1943 if ((c == '.' && have_ellipsis_continuation ()) |
|
3105
|
1944 || (c == '\\' && have_continuation ())) |
|
1001
|
1945 retval = eat_whitespace (); |
|
|
1946 else |
|
3246
|
1947 unput (c); |
|
1001
|
1948 |
|
|
1949 return retval; |
|
|
1950 } |
|
|
1951 |
|
973
|
1952 static int |
|
975
|
1953 handle_string (char delim, int text_style) |
|
973
|
1954 { |
|
4051
|
1955 OSSTREAM buf; |
|
973
|
1956 |
|
3805
|
1957 int bos_line = input_line_number; |
|
|
1958 int bos_col = current_input_column; |
|
|
1959 |
|
973
|
1960 int c; |
|
1031
|
1961 int escape_pending = 0; |
|
973
|
1962 |
|
|
1963 while ((c = yyinput ()) != EOF) |
|
|
1964 { |
|
|
1965 current_input_column++; |
|
|
1966 |
|
3105
|
1967 if (c == '\\') |
|
973
|
1968 { |
|
1053
|
1969 if (escape_pending) |
|
|
1970 { |
|
3802
|
1971 buf << static_cast<char> (c); |
|
1053
|
1972 escape_pending = 0; |
|
|
1973 } |
|
|
1974 else |
|
|
1975 { |
|
3096
|
1976 if (have_continuation (false)) |
|
1053
|
1977 escape_pending = 0; |
|
|
1978 else |
|
|
1979 { |
|
3802
|
1980 buf << static_cast<char> (c); |
|
1053
|
1981 escape_pending = 1; |
|
|
1982 } |
|
|
1983 } |
|
1031
|
1984 continue; |
|
973
|
1985 } |
|
|
1986 else if (c == '.') |
|
|
1987 { |
|
3096
|
1988 if (! have_ellipsis_continuation (false)) |
|
3802
|
1989 buf << static_cast<char> (c); |
|
973
|
1990 } |
|
|
1991 else if (c == '\n') |
|
|
1992 { |
|
1053
|
1993 error ("unterminated string constant"); |
|
973
|
1994 break; |
|
|
1995 } |
|
|
1996 else if (c == delim) |
|
|
1997 { |
|
1031
|
1998 if (escape_pending) |
|
3802
|
1999 buf << static_cast<char> (c); |
|
973
|
2000 else |
|
|
2001 { |
|
|
2002 c = yyinput (); |
|
|
2003 if (c == delim) |
|
3802
|
2004 buf << static_cast<char> (c); |
|
973
|
2005 else |
|
|
2006 { |
|
3246
|
2007 unput (c); |
|
4051
|
2008 buf << OSSTREAM_ENDS; |
|
|
2009 std::string s = do_string_escapes (OSSTREAM_STR (buf)); |
|
|
2010 OSSTREAM_FREEZE (buf); |
|
975
|
2011 |
|
1826
|
2012 if (text_style && lexer_flags.doing_set) |
|
975
|
2013 { |
|
3523
|
2014 s = std::string (1, delim) + s + std::string (1, delim); |
|
975
|
2015 } |
|
|
2016 else |
|
|
2017 { |
|
2857
|
2018 lexer_flags.quote_is_transpose = true; |
|
|
2019 lexer_flags.cant_be_identifier = true; |
|
|
2020 lexer_flags.convert_spaces_to_comma = true; |
|
975
|
2021 } |
|
|
2022 |
|
3805
|
2023 yylval.tok_val = new token (s, bos_line, bos_col); |
|
973
|
2024 token_stack.push (yylval.tok_val); |
|
3400
|
2025 |
|
4037
|
2026 if (delim == '"') |
|
|
2027 gripe_matlab_incompatible ("\" used as string delimiter"); |
|
|
2028 else if (delim == '\'') |
|
3400
|
2029 gripe_single_quote_string (); |
|
|
2030 |
|
4064
|
2031 return STRING; |
|
973
|
2032 } |
|
|
2033 } |
|
|
2034 } |
|
|
2035 else |
|
|
2036 { |
|
3802
|
2037 buf << static_cast<char> (c); |
|
973
|
2038 } |
|
|
2039 |
|
1031
|
2040 escape_pending = 0; |
|
973
|
2041 } |
|
|
2042 |
|
|
2043 return LEXICAL_ERROR; |
|
|
2044 } |
|
|
2045 |
|
3208
|
2046 static bool |
|
|
2047 next_token_is_assign_op (void) |
|
|
2048 { |
|
|
2049 bool retval = false; |
|
|
2050 |
|
|
2051 int c0 = yyinput (); |
|
|
2052 |
|
|
2053 switch (c0) |
|
|
2054 { |
|
|
2055 case '=': |
|
|
2056 { |
|
|
2057 int c1 = yyinput (); |
|
|
2058 unput (c1); |
|
|
2059 if (c1 != '=') |
|
|
2060 retval = true; |
|
|
2061 } |
|
|
2062 break; |
|
|
2063 |
|
|
2064 case '+': |
|
|
2065 case '-': |
|
|
2066 case '*': |
|
|
2067 case '/': |
|
|
2068 case '\\': |
|
|
2069 case '&': |
|
|
2070 case '|': |
|
|
2071 { |
|
|
2072 int c1 = yyinput (); |
|
|
2073 unput (c1); |
|
|
2074 if (c1 == '=') |
|
|
2075 retval = true; |
|
|
2076 } |
|
|
2077 break; |
|
|
2078 |
|
|
2079 case '.': |
|
|
2080 { |
|
|
2081 int c1 = yyinput (); |
|
|
2082 if (match_any (c1, "+-*/\\")) |
|
|
2083 { |
|
|
2084 int c2 = yyinput (); |
|
|
2085 unput (c2); |
|
|
2086 if (c2 == '=') |
|
|
2087 retval = true; |
|
|
2088 } |
|
|
2089 unput (c1); |
|
|
2090 } |
|
|
2091 break; |
|
|
2092 |
|
|
2093 case '>': |
|
|
2094 { |
|
|
2095 int c1 = yyinput (); |
|
|
2096 if (c1 == '>') |
|
|
2097 { |
|
|
2098 int c2 = yyinput (); |
|
|
2099 unput (c2); |
|
|
2100 if (c2 == '=') |
|
|
2101 retval = true; |
|
|
2102 } |
|
|
2103 unput (c1); |
|
|
2104 } |
|
|
2105 break; |
|
|
2106 |
|
|
2107 case '<': |
|
|
2108 { |
|
|
2109 int c1 = yyinput (); |
|
|
2110 if (c1 == '<') |
|
|
2111 { |
|
|
2112 int c2 = yyinput (); |
|
|
2113 unput (c2); |
|
|
2114 if (c2 == '=') |
|
|
2115 retval = true; |
|
|
2116 } |
|
|
2117 unput (c1); |
|
|
2118 } |
|
|
2119 break; |
|
|
2120 |
|
|
2121 default: |
|
|
2122 break; |
|
|
2123 } |
|
|
2124 |
|
|
2125 unput (c0); |
|
|
2126 |
|
|
2127 return retval; |
|
|
2128 } |
|
|
2129 |
|
971
|
2130 static int |
|
3351
|
2131 handle_close_bracket (int spc_gobbled) |
|
971
|
2132 { |
|
3208
|
2133 int retval = ']'; |
|
|
2134 |
|
1826
|
2135 if (! nesting_level.none ()) |
|
971
|
2136 { |
|
1826
|
2137 nesting_level.remove (); |
|
3351
|
2138 lexer_flags.bracketflag--; |
|
971
|
2139 } |
|
|
2140 |
|
3351
|
2141 if (lexer_flags.bracketflag == 0) |
|
1001
|
2142 BEGIN 0; |
|
|
2143 |
|
3208
|
2144 if (next_token_is_assign_op () && ! lexer_flags.looking_at_return_list) |
|
971
|
2145 { |
|
3208
|
2146 retval = CLOSE_BRACE; |
|
971
|
2147 } |
|
|
2148 else |
|
|
2149 { |
|
3208
|
2150 int c1 = yyinput (); |
|
971
|
2151 unput (c1); |
|
|
2152 |
|
3388
|
2153 if (lexer_flags.bracketflag) |
|
971
|
2154 { |
|
3246
|
2155 int bin_op = next_token_is_bin_op (spc_gobbled); |
|
3263
|
2156 |
|
3246
|
2157 int postfix_un_op = next_token_is_postfix_unary_op (spc_gobbled); |
|
971
|
2158 |
|
3263
|
2159 int sep_op = next_token_is_sep_op (); |
|
971
|
2160 |
|
3263
|
2161 if (! (postfix_un_op || bin_op || sep_op) |
|
3351
|
2162 && nesting_level.is_bracket () |
|
1826
|
2163 && lexer_flags.convert_spaces_to_comma) |
|
971
|
2164 { |
|
3388
|
2165 maybe_warn_separator_insert (','); |
|
|
2166 |
|
|
2167 if (Vwhitespace_in_literal_matrix != 2) |
|
|
2168 { |
|
|
2169 unput (','); |
|
|
2170 return ']'; |
|
|
2171 } |
|
971
|
2172 } |
|
|
2173 } |
|
|
2174 } |
|
|
2175 |
|
2857
|
2176 lexer_flags.quote_is_transpose = true; |
|
|
2177 lexer_flags.cant_be_identifier = false; |
|
|
2178 lexer_flags.convert_spaces_to_comma = true; |
|
3208
|
2179 |
|
|
2180 return retval; |
|
971
|
2181 } |
|
|
2182 |
|
1072
|
2183 static void |
|
|
2184 maybe_unput_comma (int spc_gobbled) |
|
|
2185 { |
|
3388
|
2186 if (nesting_level.is_bracket ()) |
|
1072
|
2187 { |
|
3246
|
2188 int bin_op = next_token_is_bin_op (spc_gobbled); |
|
1072
|
2189 |
|
3246
|
2190 int postfix_un_op = next_token_is_postfix_unary_op (spc_gobbled); |
|
1072
|
2191 |
|
|
2192 int c1 = yyinput (); |
|
|
2193 int c2 = yyinput (); |
|
2970
|
2194 |
|
1072
|
2195 unput (c2); |
|
|
2196 unput (c1); |
|
2970
|
2197 |
|
3263
|
2198 int sep_op = next_token_is_sep_op (); |
|
2970
|
2199 |
|
1072
|
2200 int dot_op = (c1 == '.' |
|
|
2201 && (isalpha (c2) || isspace (c2) || c2 == '_')); |
|
2970
|
2202 |
|
3388
|
2203 if (postfix_un_op || bin_op || sep_op || dot_op) |
|
|
2204 return; |
|
|
2205 |
|
3985
|
2206 int index_op = (c1 == '(' || c1 == '{'); |
|
3388
|
2207 |
|
|
2208 if (index_op) |
|
|
2209 { |
|
3985
|
2210 // If there is no space before the indexing op, we don't |
|
|
2211 // insert a comma. |
|
3388
|
2212 if (! spc_gobbled) |
|
|
2213 return; |
|
|
2214 |
|
|
2215 maybe_warn_separator_insert (','); |
|
1072
|
2216 |
|
3388
|
2217 // If there is a space, we only insert a comma if we are |
|
|
2218 // trying to be Matlab-like. |
|
|
2219 if (Vwhitespace_in_literal_matrix == 1) |
|
|
2220 unput (','); |
|
|
2221 } |
|
|
2222 else |
|
|
2223 { |
|
|
2224 maybe_warn_separator_insert (','); |
|
|
2225 |
|
|
2226 if (Vwhitespace_in_literal_matrix != 2) |
|
|
2227 unput (','); |
|
|
2228 } |
|
1072
|
2229 } |
|
|
2230 } |
|
|
2231 |
|
767
|
2232 // Figure out exactly what kind of token to return when we have seen |
|
|
2233 // an identifier. Handles keywords. |
|
|
2234 |
|
146
|
2235 static int |
|
3974
|
2236 handle_identifier (void) |
|
146
|
2237 { |
|
3974
|
2238 std::string tok = strip_trailing_whitespace (yytext); |
|
|
2239 |
|
|
2240 int c = yytext[yyleng-1]; |
|
|
2241 |
|
|
2242 int cont_is_spc = eat_continuation (); |
|
|
2243 |
|
|
2244 int spc_gobbled = (cont_is_spc || c == ' ' || c == '\t'); |
|
|
2245 |
|
1826
|
2246 // It is almost always an error for an identifier to be followed |
|
|
2247 // directly by another identifier. Special cases are handled |
|
|
2248 // below. |
|
883
|
2249 |
|
2857
|
2250 lexer_flags.cant_be_identifier = true; |
|
883
|
2251 |
|
2970
|
2252 // If we are expecting a structure element, avoid recognizing |
|
|
2253 // keywords and other special names and return STRUCT_ELT, which is |
|
|
2254 // a string that is also a valid identifier. But first, we have to |
|
|
2255 // decide whether to insert a comma. |
|
747
|
2256 |
|
1826
|
2257 if (lexer_flags.looking_at_indirect_ref) |
|
1072
|
2258 { |
|
2970
|
2259 do_comma_insert_check (); |
|
|
2260 |
|
1072
|
2261 maybe_unput_comma (spc_gobbled); |
|
2819
|
2262 |
|
|
2263 yylval.tok_val = new token (tok, input_line_number, |
|
|
2264 current_input_column); |
|
|
2265 |
|
|
2266 token_stack.push (yylval.tok_val); |
|
|
2267 |
|
2857
|
2268 lexer_flags.cant_be_identifier = false; |
|
|
2269 lexer_flags.quote_is_transpose = true; |
|
|
2270 lexer_flags.convert_spaces_to_comma = true; |
|
2819
|
2271 |
|
|
2272 current_input_column += yyleng; |
|
|
2273 |
|
2970
|
2274 return STRUCT_ELT; |
|
1072
|
2275 } |
|
747
|
2276 |
|
1826
|
2277 // If we have a regular keyword, or a plot STYLE, return it. |
|
|
2278 // Keywords can be followed by identifiers (TOK_RETURN handles |
|
|
2279 // that). |
|
146
|
2280 |
|
|
2281 int kw_token = is_keyword (tok); |
|
2970
|
2282 |
|
146
|
2283 if (kw_token) |
|
|
2284 { |
|
|
2285 if (kw_token == STYLE) |
|
|
2286 { |
|
1060
|
2287 current_input_column += yyleng; |
|
2857
|
2288 lexer_flags.quote_is_transpose = false; |
|
|
2289 lexer_flags.convert_spaces_to_comma = true; |
|
146
|
2290 return kw_token; |
|
|
2291 } |
|
|
2292 else |
|
|
2293 TOK_RETURN (kw_token); |
|
|
2294 } |
|
|
2295 |
|
1826
|
2296 // See if we have a plot keyword (title, using, with, or clear). |
|
146
|
2297 |
|
1826
|
2298 if (lexer_flags.plotting) |
|
941
|
2299 { |
|
1826
|
2300 // Yes, we really do need both of these plot_range variables. |
|
|
2301 // One is used to mark when we are past all possiblity of a plot |
|
|
2302 // range, the other is used to mark when we are actually between |
|
|
2303 // the square brackets that surround the range. |
|
146
|
2304 |
|
1826
|
2305 if (! lexer_flags.in_plot_range) |
|
2857
|
2306 lexer_flags.past_plot_range = true; |
|
941
|
2307 |
|
3351
|
2308 // Option keywords can't appear in brackets, braces, or parentheses. |
|
1273
|
2309 |
|
|
2310 int plot_option_kw = 0; |
|
2970
|
2311 |
|
1826
|
2312 if (nesting_level.none ()) |
|
1273
|
2313 plot_option_kw = is_plot_keyword (tok); |
|
3575
|
2314 |
|
1826
|
2315 if (lexer_flags.cant_be_identifier && plot_option_kw) |
|
941
|
2316 TOK_RETURN (plot_option_kw); |
|
|
2317 } |
|
146
|
2318 |
|
3480
|
2319 int c1 = yyinput (); |
|
|
2320 |
|
|
2321 bool next_tok_is_paren = (c1 == '('); |
|
|
2322 |
|
|
2323 bool next_tok_is_eq = false; |
|
|
2324 if (c1 == '=') |
|
|
2325 { |
|
|
2326 int c2 = yyinput (); |
|
|
2327 unput (c2); |
|
|
2328 |
|
|
2329 if (c2 != '=') |
|
|
2330 next_tok_is_eq = true; |
|
|
2331 } |
|
|
2332 |
|
|
2333 unput (c1); |
|
1001
|
2334 |
|
1826
|
2335 // Make sure we put the return values of a function in the symbol |
|
|
2336 // table that is local to the function. |
|
146
|
2337 |
|
2877
|
2338 // If we are defining a function and we have not seen the function |
|
|
2339 // name yet and the next token is `=', then this identifier must be |
|
|
2340 // the only return value for the function and it belongs in the |
|
|
2341 // local symbol table. |
|
|
2342 |
|
1826
|
2343 if (next_tok_is_eq |
|
2877
|
2344 && lexer_flags.defining_func |
|
|
2345 && ! lexer_flags.parsed_function_name) |
|
146
|
2346 curr_sym_tab = tmp_local_sym_tab; |
|
|
2347 |
|
2702
|
2348 // Kluge alert. |
|
|
2349 // |
|
|
2350 // If we are looking at a text style function, set up to gobble its |
|
2745
|
2351 // arguments. |
|
|
2352 // |
|
|
2353 // If the following token is `=', or if we are parsing a function |
|
3189
|
2354 // return list or function parameter list, or if we are looking at |
|
|
2355 // something like [ab,cd] = foo (), force the symbol to be inserted |
|
|
2356 // as a variable in the current symbol table. |
|
2702
|
2357 |
|
|
2358 if (is_text_function_name (tok) && ! is_variable (tok)) |
|
|
2359 { |
|
2745
|
2360 if (next_tok_is_eq |
|
|
2361 || lexer_flags.looking_at_return_list |
|
3189
|
2362 || lexer_flags.looking_at_parameter_list |
|
4101
|
2363 || lexer_flags.looking_at_matrix_or_assign_lhs) |
|
2745
|
2364 { |
|
|
2365 force_local_variable (tok); |
|
|
2366 } |
|
2702
|
2367 else if (! next_tok_is_paren) |
|
|
2368 { |
|
|
2369 if (tok == "gset") |
|
2857
|
2370 lexer_flags.doing_set = true; |
|
2702
|
2371 |
|
|
2372 BEGIN TEXT_FCN; |
|
|
2373 } |
|
|
2374 } |
|
|
2375 |
|
1826
|
2376 // Find the token in the symbol table. |
|
146
|
2377 |
|
|
2378 yylval.tok_val = new token (lookup_identifier (tok), |
|
|
2379 input_line_number, |
|
|
2380 current_input_column); |
|
|
2381 |
|
|
2382 token_stack.push (yylval.tok_val); |
|
|
2383 |
|
1826
|
2384 // After seeing an identifer, it is ok to convert spaces to a comma |
|
|
2385 // (if needed). |
|
146
|
2386 |
|
2857
|
2387 lexer_flags.convert_spaces_to_comma = true; |
|
146
|
2388 |
|
2877
|
2389 if (! next_tok_is_eq) |
|
|
2390 { |
|
|
2391 lexer_flags.quote_is_transpose = true; |
|
146
|
2392 |
|
2877
|
2393 do_comma_insert_check (); |
|
|
2394 |
|
|
2395 maybe_unput_comma (spc_gobbled); |
|
146
|
2396 } |
|
|
2397 |
|
2877
|
2398 current_input_column += yyleng; |
|
146
|
2399 |
|
|
2400 return NAME; |
|
|
2401 } |
|
|
2402 |
|
767
|
2403 // Print a warning if a function file that defines a function has |
|
|
2404 // anything other than comments and whitespace following the END token |
|
|
2405 // that matches the FUNCTION statement. |
|
|
2406 |
|
1
|
2407 void |
|
|
2408 check_for_garbage_after_fcn_def (void) |
|
|
2409 { |
|
1826
|
2410 // By making a newline be the next character to be read, we will |
|
|
2411 // force the parser to return after reading the function. Calling |
|
3246
|
2412 // unput with EOF does not work. |
|
1
|
2413 |
|
3665
|
2414 std::string comment_buf; |
|
|
2415 |
|
2857
|
2416 bool in_comment = false; |
|
3665
|
2417 bool beginning_of_comment = true; |
|
|
2418 |
|
1
|
2419 int lineno = input_line_number; |
|
3665
|
2420 |
|
|
2421 int c = 0; |
|
|
2422 |
|
1
|
2423 while ((c = yyinput ()) != EOF) |
|
|
2424 { |
|
|
2425 switch (c) |
|
|
2426 { |
|
|
2427 case ' ': |
|
|
2428 case '\t': |
|
|
2429 case ';': |
|
|
2430 case ',': |
|
3665
|
2431 if (in_comment) |
|
|
2432 { |
|
3802
|
2433 comment_buf += static_cast<char> (c); |
|
3665
|
2434 beginning_of_comment = false; |
|
|
2435 } |
|
1
|
2436 break; |
|
777
|
2437 |
|
1
|
2438 case '%': |
|
|
2439 case '#': |
|
3665
|
2440 if (in_comment) |
|
|
2441 { |
|
|
2442 if (! beginning_of_comment) |
|
3802
|
2443 comment_buf += static_cast<char> (c); |
|
3665
|
2444 } |
|
|
2445 else |
|
|
2446 { |
|
4037
|
2447 maybe_gripe_matlab_incompatible_comment (c); |
|
3665
|
2448 in_comment = true; |
|
|
2449 beginning_of_comment = true; |
|
|
2450 } |
|
1
|
2451 break; |
|
777
|
2452 |
|
3802
|
2453 case '\n': |
|
|
2454 if (in_comment) |
|
|
2455 { |
|
|
2456 comment_buf += static_cast<char> (c); |
|
|
2457 octave_comment_buffer::append (comment_buf); |
|
|
2458 comment_buf.resize (0); |
|
|
2459 in_comment = false; |
|
|
2460 beginning_of_comment = false; |
|
|
2461 } |
|
|
2462 break; |
|
|
2463 |
|
|
2464 case '\r': |
|
|
2465 if (in_comment) |
|
|
2466 comment_buf += static_cast<char> (c); |
|
|
2467 c = yyinput (); |
|
|
2468 if (c == EOF) |
|
|
2469 break; |
|
|
2470 else if (c == '\n') |
|
|
2471 { |
|
|
2472 if (in_comment) |
|
|
2473 { |
|
|
2474 comment_buf += static_cast<char> (c); |
|
|
2475 octave_comment_buffer::append (comment_buf); |
|
|
2476 comment_buf.resize (0); |
|
|
2477 in_comment = false; |
|
|
2478 beginning_of_comment = false; |
|
|
2479 } |
|
|
2480 break; |
|
|
2481 } |
|
|
2482 |
|
|
2483 // Fall through... |
|
|
2484 |
|
1
|
2485 default: |
|
|
2486 if (in_comment) |
|
3665
|
2487 { |
|
3802
|
2488 comment_buf += static_cast<char> (c); |
|
3665
|
2489 beginning_of_comment = false; |
|
|
2490 break; |
|
|
2491 } |
|
1
|
2492 else |
|
|
2493 { |
|
|
2494 warning ("ignoring trailing garbage after end of function\n\ |
|
1755
|
2495 near line %d of file `%s.m'", lineno, curr_fcn_file_name.c_str ()); |
|
1
|
2496 |
|
3246
|
2497 unput ('\n'); |
|
1
|
2498 return; |
|
|
2499 } |
|
|
2500 } |
|
|
2501 } |
|
3665
|
2502 |
|
|
2503 if (! comment_buf.empty ()) |
|
|
2504 octave_comment_buffer::append (comment_buf); |
|
|
2505 |
|
3246
|
2506 unput ('\n'); |
|
1
|
2507 } |
|
|
2508 |
|
1826
|
2509 void |
|
|
2510 lexical_feedback::init (void) |
|
|
2511 { |
|
|
2512 // Not initially defining a matrix list. |
|
3351
|
2513 bracketflag = 0; |
|
1826
|
2514 |
|
|
2515 // Not initially inside a loop or if statement. |
|
|
2516 looping = 0; |
|
|
2517 |
|
2857
|
2518 // Not initially defining a function. |
|
|
2519 beginning_of_function = false; |
|
|
2520 defining_func = false; |
|
2877
|
2521 parsed_function_name = false; |
|
2857
|
2522 |
|
|
2523 // Not parsing a function return or parameter list. |
|
|
2524 looking_at_return_list = false; |
|
|
2525 looking_at_parameter_list = false; |
|
|
2526 |
|
3796
|
2527 // Not parsing a matrix or the left hand side of multi-value |
|
|
2528 // assignment statement. |
|
|
2529 looking_at_matrix_or_assign_lhs = false; |
|
|
2530 |
|
2857
|
2531 // Next token can be identifier. |
|
|
2532 cant_be_identifier = false; |
|
|
2533 |
|
|
2534 // No need to do comma insert or convert spaces to comma at |
|
|
2535 // beginning of input. |
|
|
2536 convert_spaces_to_comma = true; |
|
|
2537 do_comma_insert = false; |
|
|
2538 |
|
|
2539 // Not initially doing any plotting or setting of plot attributes. |
|
|
2540 doing_set = false; |
|
|
2541 in_plot_range = false; |
|
|
2542 in_plot_style = false; |
|
3165
|
2543 in_plot_axes = false; |
|
2857
|
2544 in_plot_using = false; |
|
|
2545 past_plot_range = false; |
|
|
2546 plotting = false; |
|
|
2547 |
|
1826
|
2548 // Not initially looking at indirect references. |
|
2857
|
2549 looking_at_indirect_ref = false; |
|
1826
|
2550 |
|
|
2551 // Quote marks strings intially. |
|
2857
|
2552 quote_is_transpose = false; |
|
1826
|
2553 } |
|
|
2554 |
|
3388
|
2555 static void |
|
|
2556 maybe_warn_separator_insert (char sep) |
|
|
2557 { |
|
3523
|
2558 std::string nm = curr_fcn_file_full_name; |
|
3388
|
2559 |
|
|
2560 if (Vwarn_separator_insert) |
|
|
2561 { |
|
|
2562 if (nm.empty ()) |
|
|
2563 warning ("potential auto-insertion of `%c' near line %d", |
|
|
2564 sep, input_line_number); |
|
|
2565 else |
|
|
2566 warning ("potential auto-insertion of `%c' near line %d of file %s", |
|
|
2567 sep, input_line_number, nm.c_str ()); |
|
|
2568 } |
|
|
2569 } |
|
|
2570 |
|
3400
|
2571 static void |
|
|
2572 gripe_single_quote_string (void) |
|
|
2573 { |
|
3523
|
2574 std::string nm = curr_fcn_file_full_name; |
|
3400
|
2575 |
|
|
2576 if (Vwarn_single_quote_string) |
|
|
2577 { |
|
|
2578 if (nm.empty ()) |
|
|
2579 warning ("single quote delimited string near line %d", |
|
|
2580 input_line_number); |
|
|
2581 else |
|
|
2582 warning ("single quote delimited string near line %d of file %s", |
|
|
2583 input_line_number, nm.c_str ()); |
|
|
2584 } |
|
|
2585 } |
|
|
2586 |
|
4037
|
2587 static void |
|
|
2588 gripe_matlab_incompatible (const std::string& msg) |
|
|
2589 { |
|
|
2590 if (Vwarn_matlab_incompatible) |
|
|
2591 warning ("potential Matlab compatibility problem: %s", msg.c_str ()); |
|
|
2592 } |
|
|
2593 |
|
|
2594 static void |
|
|
2595 maybe_gripe_matlab_incompatible_comment (char c) |
|
|
2596 { |
|
|
2597 if (c == '#') |
|
|
2598 gripe_matlab_incompatible ("# used as comment character"); |
|
|
2599 } |
|
|
2600 |
|
|
2601 static void |
|
|
2602 gripe_matlab_incompatible_continuation (void) |
|
|
2603 { |
|
|
2604 gripe_matlab_incompatible ("\\ used as line continuation marker"); |
|
|
2605 } |
|
|
2606 |
|
|
2607 static void |
|
|
2608 gripe_matlab_incompatible_operator (const std::string& op) |
|
|
2609 { |
|
|
2610 std::string t = op; |
|
|
2611 int n = t.length (); |
|
|
2612 if (t[n-1] == '\n') |
|
|
2613 t.resize (n-1); |
|
|
2614 gripe_matlab_incompatible (t + " used as operator"); |
|
|
2615 } |
|
|
2616 |
|
|
2617 static int |
|
|
2618 warn_matlab_incompatible (void) |
|
|
2619 { |
|
|
2620 Vwarn_matlab_incompatible = check_preference ("warn_matlab_incompatible"); |
|
|
2621 |
|
|
2622 return 0; |
|
|
2623 } |
|
|
2624 |
|
3388
|
2625 static int |
|
|
2626 warn_separator_insert (void) |
|
|
2627 { |
|
|
2628 Vwarn_separator_insert = check_preference ("warn_separator_insert"); |
|
|
2629 |
|
|
2630 return 0; |
|
|
2631 } |
|
|
2632 |
|
3246
|
2633 static int |
|
3400
|
2634 warn_single_quote_string (void) |
|
|
2635 { |
|
|
2636 Vwarn_single_quote_string = check_preference ("warn_single_quote_string"); |
|
|
2637 |
|
|
2638 return 0; |
|
|
2639 } |
|
|
2640 |
|
|
2641 static int |
|
2167
|
2642 whitespace_in_literal_matrix (void) |
|
|
2643 { |
|
|
2644 int pref = 0; |
|
3009
|
2645 |
|
3523
|
2646 std::string val = builtin_string_variable ("whitespace_in_literal_matrix"); |
|
3009
|
2647 |
|
2167
|
2648 if (! val.empty ()) |
|
|
2649 { |
|
3565
|
2650 if (val == "ignore") |
|
2167
|
2651 pref = 2; |
|
3565
|
2652 else if (val == "traditional") |
|
2167
|
2653 pref = 1; |
|
|
2654 } |
|
3009
|
2655 |
|
2167
|
2656 Vwhitespace_in_literal_matrix = pref; |
|
3009
|
2657 |
|
2167
|
2658 return 0; |
|
|
2659 } |
|
|
2660 |
|
|
2661 void |
|
|
2662 symbols_of_lex (void) |
|
|
2663 { |
|
4037
|
2664 DEFVAR (warn_matlab_incompatible, 0.0, warn_matlab_incompatible, |
|
|
2665 "-*- texinfo -*-\n\ |
|
|
2666 @defvr {Built-in Variable} warn_matlab_incompatible\n\ |
|
|
2667 Print warnings for Octave language features that may cause\n\ |
|
|
2668 compatibility problems with Matlab.\n\ |
|
|
2669 @end defvr"); |
|
|
2670 |
|
3388
|
2671 DEFVAR (warn_separator_insert, 0.0, warn_separator_insert, |
|
3428
|
2672 "-*- texinfo -*-\n\ |
|
|
2673 @defvr {Built-in Variable} warn_separator_insert\n\ |
|
3567
|
2674 Print warning if commas or semicolons might be inserted\n\ |
|
3428
|
2675 automatically in literal matrices.\n\ |
|
|
2676 @end defvr"); |
|
3388
|
2677 |
|
3400
|
2678 DEFVAR (warn_single_quote_string, 0.0, warn_single_quote_string, |
|
3428
|
2679 "-*- texinfo -*-\n\ |
|
|
2680 @defvr {Built-in Variable} warn_single_quote_string\n\ |
|
|
2681 Print warning if a signle quote character is used to introduce a\n\ |
|
|
2682 string constant.\n\ |
|
|
2683 @end defvr"); |
|
3400
|
2684 |
|
3258
|
2685 DEFVAR (whitespace_in_literal_matrix, "", whitespace_in_literal_matrix, |
|
3428
|
2686 "-*- texinfo -*-\n\ |
|
|
2687 @defvr {Built-in Variable} whitespace_in_literal_matrix\n\ |
|
|
2688 Control auto-insertion of commas and semicolons in literal matrices.\n\ |
|
|
2689 @end defvr"); |
|
|
2690 |
|
2167
|
2691 } |
|
|
2692 |
|
1994
|
2693 /* |
|
|
2694 ;;; Local Variables: *** |
|
|
2695 ;;; mode: C++ *** |
|
|
2696 ;;; End: *** |
|
|
2697 */ |
