]>
Commit | Line | Data |
---|---|---|
1 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
2 | ||
3 | * src/lalr.c (set_goto_map, initialize_F): Use SHIFT_SYMBOL. | |
4 | ||
5 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
6 | ||
7 | * src/conflicts.c, src/conflicts.h (print_reductions): Move to... | |
8 | * src/print.c: here. | |
9 | (lookaheadset, shiftset): New, used as additional storage by | |
10 | print_reductions. | |
11 | (print_results): Adjust. | |
12 | (print_shifts, print_gotos, print_errs): New, extracted from... | |
13 | (print_actions): here. | |
14 | * src/print_graph.c (print_actions): Remove dead code. | |
15 | ||
16 | ||
17 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
18 | ||
19 | * src/reader.c (copy_dollar, copy_at): Better checking of `n' in | |
20 | `$n' and `@n'. | |
21 | ||
22 | ||
23 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
24 | ||
25 | * src/lalr.c (add_lookback_edge): Use state_t instead of ints. | |
26 | (build_relations): Adjust. | |
27 | ||
28 | ||
29 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
30 | ||
31 | * src/lalr.c (set_goto_map): Remove a wrong but benign loop | |
32 | duplication. | |
33 | ||
34 | ||
35 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
36 | ||
37 | * src/reader.c (packgram): Catch nitems overflows. | |
38 | ||
39 | ||
40 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
41 | ||
42 | * src/files.c, src/files.h (guard_obstack): Remove. | |
43 | * src/output.c (output): Adjust. | |
44 | * src/reader.c (parse_braces): New, factoring... | |
45 | (copy_action, copy_guard): these two which are renamed as... | |
46 | (parse_action, parse_guard): these. | |
47 | As a voluntary consequence, using braces around guards is now | |
48 | mandatory. | |
49 | ||
50 | ||
51 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
52 | ||
53 | * src/gram.h (rule_t): `guard' and `guard_line' are new members. | |
54 | * src/reader.c (symbol_list): `guard' and `guard_line' are new | |
55 | members. | |
56 | (symbol_list_new): Adjust. | |
57 | (copy_action): action_line is the first line, not the last. | |
58 | (copy_guard): Just as for actions, store the `action' only, not | |
59 | the switch/case/break flesh. | |
60 | Don't parse the user action that might follow the guard, let... | |
61 | (readgram): do it, i.e., now, there can be an action after a | |
62 | guard. | |
63 | In other words the guard is just explicitly optional. | |
64 | (packgram): Adjust. | |
65 | * src/output.c (guards_output): New. | |
66 | (output_parser): Call it when needed. | |
67 | (output): Also free the guard and attrs obstacks. | |
68 | * src/files.c, src/files.h (obstack_save): Remove. | |
69 | (output_files): Remove. | |
70 | As a result, if one needs the former `.act' file, using an | |
71 | appropriate skeleton which requires actions and guards is now | |
72 | required. | |
73 | * src/main.c (main): Adjust. | |
74 | * tests/semantic.at: New. | |
75 | * tests/regression.at: Use `input.y' as input file name. | |
76 | Avoid 8+3 problems by requiring input.c when the test needs the | |
77 | parser. | |
78 | ||
79 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
80 | ||
81 | * src/reader.c (symbol_list_new): Be sure to initialize all the | |
82 | fields. | |
83 | ||
84 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
85 | ||
86 | All the hacks using a final pseudo state are now useless. | |
87 | ||
88 | * src/LR0.c (set_state_table): state_table holds exactly nstates. | |
89 | * src/lalr.c (nLA): New. | |
90 | (initialize_LA, compute_lookaheads, initialize_lookaheads): Use it | |
91 | instead of lookaheadsp from the pseudo state (nstate + 1). | |
92 | ||
93 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
94 | ||
95 | * src/output.c (action_row, token_actions): Use a state_t instead | |
96 | of a integer, and nlookaheads instead of the following state's | |
97 | lookaheadsp. | |
98 | ||
99 | ||
100 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
101 | ||
102 | * src/conflicts.c (log_resolution, flush_shift) | |
103 | (resolve_sr_conflict, set_conflicts, solve_conflicts) | |
104 | (count_sr_conflicts, count_rr_conflicts, conflicts_output) | |
105 | (conflicts_print, print_reductions): Use a state_t instead of an | |
106 | integer when referring to a state. | |
107 | As much as possible, depend upon nlookaheads, instead of the | |
108 | `lookaheadsp' member of the following state (since lookaheads of | |
109 | successive states are successive, the difference between state n + 1 | |
110 | and n served as the number of lookaheads for state n). | |
111 | * src/lalr.c (add_lookback_edge): Likewise. | |
112 | * src/print.c (print_core, print_actions, print_state) | |
113 | (print_results): Likewise. | |
114 | * src/print_graph.c (print_core, print_actions, print_state) | |
115 | (print_graph): Likewise. | |
116 | * src/conflicts.h: Adjust. | |
117 | ||
118 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
119 | ||
120 | * src/bison.hairy: Formatting/comment changes. | |
121 | ANSIfy. | |
122 | Remove `register' indications. | |
123 | Add plenty of `static'. | |
124 | ||
125 | 2001-12-27 Akim Demaille <akim@epita.fr> | |
126 | ||
127 | * src/output.c (prepare): Drop the muscle `ntbase' which | |
128 | duplicates ntokens. | |
129 | * src/bison.simple: Formatting/comment changes. | |
130 | Use YYNTOKENS only, which is documented, but not YYNTBASE, which | |
131 | is an undocumented synonym. | |
132 | ||
133 | ||
134 | 2001-12-22 Akim Demaille <akim@epita.fr> | |
135 | ||
136 | * src/output.c (output_table_data): Change the prototype to use | |
137 | `int' for array ranges: some invocations do pass an int, not a | |
138 | short. | |
139 | Reported by Wayne Green. | |
140 | ||
141 | 2001-12-22 Akim Demaille <akim@epita.fr> | |
142 | ||
143 | Some actions of web2c.y are improperly triggered. | |
144 | Reported by Mike Castle. | |
145 | ||
146 | * src/lalr.c (traverse): s/F (i)[k] = F (j)[k]/F (j)[k] = F (i)[k]/. | |
147 | * tests/regression.at (Web2c): Rename as... | |
148 | (Web2c Report): this. | |
149 | (Web2c Actions): New. | |
150 | ||
151 | 2001-12-22 Akim Demaille <akim@epita.fr> | |
152 | ||
153 | Reductions in web2c.y are improperly reported. | |
154 | Reported by Mike Castle. | |
155 | ||
156 | * src/conflicts.c (print_reductions): Fix. | |
157 | * tests/regression.at (Web2c): New. | |
158 | ||
159 | 2001-12-18 Akim Demaille <akim@epita.fr> | |
160 | ||
161 | Some host fail on `assert (!"foo")', which expands to | |
162 | ((!"foo") ? (void)0 : __assert("!"foo."", __FILE__, __LINE__)) | |
163 | Reported by Nelson Beebee. | |
164 | ||
165 | * src/output.c, src/vcg.c: Replace `assert (!"it succeeded")' with | |
166 | `#define it_succeeded 0' and `assert (it_succeeded)'. | |
167 | ||
168 | 2001-12-17 Marc Autret <autret_m@epita.fr> | |
169 | ||
170 | * src/bison.simple: Don't hard code the skeleton line and filename. | |
171 | * src/output.c (output_parser): Rename 'line' as 'output_line'. | |
172 | New line counter 'skeleton_line' (skeleton-line muscle). | |
173 | ||
174 | 2001-12-17 Paul Eggert <eggert@twinsun.com> | |
175 | ||
176 | * NEWS, doc/bison.texinfo, doc/bison.1, doc/bison.rnh: Document that | |
177 | YYDEBUG must be defined to a nonzero value. | |
178 | ||
179 | * src/bison.simple (yytname): Do not assume that the user defines | |
180 | YYDEBUG to a properly parenthesized expression. | |
181 | ||
182 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
183 | ||
184 | * src/state.h (state_t): Rename lookaheads as lookaheadsp. | |
185 | nlookaheads is a new member. | |
186 | Adjust all users. | |
187 | * src/lalr.h (nlookaheads): Remove this orphan declaration. | |
188 | * src/lalr.c (initialize_lookaheads): Set nlookaheads for each | |
189 | state. | |
190 | ||
191 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
192 | ||
193 | * src/files.h, src/files.c (open_files, close_files): Remove. | |
194 | * src/main.c (main): Don't open/close files, nor invoke lex_free, | |
195 | let... | |
196 | * src/reader.c (reader): Do it. | |
197 | ||
198 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
199 | ||
200 | * src/conflicts.c (print_reductions): Formatting changes. | |
201 | ||
202 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
203 | ||
204 | * src/conflicts.c (flush_shift): Also adjust lookaheadset. | |
205 | (flush_reduce): New. | |
206 | (resolve_sr_conflict): Adjust. | |
207 | ||
208 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
209 | ||
210 | * src/output.c (output_obstack): Be static and rename as... | |
211 | (format_obstack): this, to avoid any confusion with files.c's | |
212 | output_obstack. | |
213 | * src/reader.h (muscle_obstack): Move to... | |
214 | * src/output.h: here, since it's defined in output.c. | |
215 | ||
216 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
217 | ||
218 | * src/output.c (action_row, save_column, default_goto) | |
219 | (sort_actions, matching_state, pack_vector): Better variable | |
220 | locality. | |
221 | ||
222 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
223 | ||
224 | * src/output.c: Various formatting changes. | |
225 | ||
226 | 2001-12-17 Akim Demaille <akim@epita.fr> | |
227 | ||
228 | * src/files.c (output_files): Free the output_obstack. | |
229 | * src/main.c (main): Call print and print_graph conditionally. | |
230 | * src/print.c (print): Work unconditionally. | |
231 | * src/print_graph.c (print_graph): Work unconditionally. | |
232 | * src/conflicts.c (log_resolution): Output only if verbose_flag. | |
233 | ||
234 | 2001-12-16 Marc Autret <autret_m@epita.fr> | |
235 | ||
236 | * src/output.c (actions_output): Fix. When we use %no-lines, | |
237 | there is one less line per action. | |
238 | ||
239 | 2001-12-16 Marc Autret <autret_m@epita.fr> | |
240 | ||
241 | * src/bison.simple: Remove a useless #line directive. | |
242 | s/#line %%line %%skeleton/#line %%line "%%parser-file-name"/'. | |
243 | * src/output.c (get_lines_number): New. | |
244 | (output_parser): Adjust, now takes care about the lines of a | |
245 | output muscles. | |
246 | Fix line numbering. | |
247 | (actions_output): Computes the number of lines taken by actions. | |
248 | (output_master_parser): Insert new skeleton which is the name of | |
249 | the output parser file name. | |
250 | ||
251 | 2001-12-15 Marc Autret <autret_m@epita.fr> | |
252 | ||
253 | * src/bison.simple [YYERROR_VERBOSE]: Restore backward compatibility. | |
254 | ||
255 | 2001-12-15 Marc Autret <autret_m@epita.fr> | |
256 | ||
257 | * src/output.c (output_gram): Keep track of the hairy one. | |
258 | ||
259 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
260 | ||
261 | Make `make distcheck' work. | |
262 | ||
263 | * lib/Makefile.am (INCLUDES): Add top_srcdir/intl, since hash uses | |
264 | system.h which uses libgettext.h. | |
265 | ||
266 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
267 | ||
268 | * src/nullable.c (set_nullable): Useless rules must be skipped, | |
269 | otherwise, since we range over their symbols, we might look at a | |
270 | nonterminal which no longer ``exists'', i.e., it is not counted in | |
271 | `nvars', hence we overflow our arrays. | |
272 | ||
273 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
274 | ||
275 | The header can also be produced directly, without any obstack! | |
276 | Yahoo! | |
277 | ||
278 | * src/files.c, src/files.h (defines_obstack): Remove. | |
279 | (compute_header_macro): Global. | |
280 | (defines_obstack_save): Remove. | |
281 | * src/reader.c (parse_union_decl): No longer output to | |
282 | defines_obstack: its content can be found in the `stype' muscle | |
283 | anyway. | |
284 | (output_token_translations): Merge into... | |
285 | (symbols_output): this. | |
286 | Rename as... | |
287 | (symbols_save): this. | |
288 | (reader): Adjust. | |
289 | * src/output.c (header_output): New. | |
290 | (output): Call it. | |
291 | ||
292 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
293 | ||
294 | * src/reader.c (parse_union_decl): Instead of handling two obstack | |
295 | simultaneously, use one to define the `stype' muscle, and use the | |
296 | value of the latter to fill defines_obstack. | |
297 | (copy_comment): Remove. | |
298 | (copy_comment2): Work for a single obstack. | |
299 | Rename as... | |
300 | (copy_comment): this. | |
301 | ||
302 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
303 | ||
304 | * src/lex.c, src/lex.h (xgetc): No longer static. | |
305 | * src/reader.c (parse_union_decl): Revamp. | |
306 | ||
307 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
308 | ||
309 | Still making progress in separating Bison into (i) input, (ii) | |
310 | process, (iii) output: now we can directly output the parser file | |
311 | without using table_obstack at all. | |
312 | ||
313 | * src/files.c, src/files.h (table_obstack): Bye bye. | |
314 | (parser_file_name): New. | |
315 | * src/files.c (compute_output_file_names): Compute it. | |
316 | * src/output.c (actions_output, output_parser) | |
317 | (output_master_parser): To a file instead of an obstack. | |
318 | ||
319 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
320 | ||
321 | Attach actions to rules, instead of pre-outputting them to | |
322 | actions_obstack. | |
323 | ||
324 | * src/gram.h (rule_t): action and action_line are new members. | |
325 | * src/reader.c (symbol_list): Likewise. | |
326 | (copy_action): Save the actions within the rule. | |
327 | (packgram): Save them in rule_table. | |
328 | * src/output.c (actions_output): New. | |
329 | (output_parser): Use it on `%%actions'. | |
330 | (output_rule_data): Don't free rule_table. | |
331 | (output): Do it. | |
332 | (prepare): Don't save the `action' muscle. | |
333 | * src/bison.simple: s/%%action/%%actions/. | |
334 | ||
335 | 2001-12-15 Akim Demaille <akim@epita.fr> | |
336 | ||
337 | * src/reader.c (copy_action): When --yacc, don't append a `;' | |
338 | to the user action: let it fail if lacking. | |
339 | Suggested by Arnold Robbins and Tom Tromey. | |
340 | ||
341 | 2001-12-14 Akim Demaille <akim@epita.fr> | |
342 | ||
343 | * src/lex.c (literalchar): Simply return the char you decoded, non | |
344 | longer mess around with obstacks and int pointers. | |
345 | Adjust all callers. | |
346 | ||
347 | 2001-12-14 Akim Demaille <akim@epita.fr> | |
348 | ||
349 | * src/lex.c (literalchar): Don't escape the special characters, | |
350 | just decode them, and keep them as char (before, eol was output as | |
351 | the 2 char string `\n' etc.). | |
352 | * src/output.c (output_rule_data): Use quotearg to output the | |
353 | token strings. | |
354 | ||
355 | 2001-12-13 Paul Eggert <eggert@twinsun.com> | |
356 | ||
357 | * src/bison.simple (YYSIZE_T, YYSTACK_ALLOC, YYSTACK_FREE): | |
358 | Do not infringe on the global user namespace when using C++. | |
359 | (YYFPRINTF, YYSTDERR): New macros, needed for the above. | |
360 | All uses of `fprintf' and `stderr' changed. | |
361 | ||
362 | * doc/bison.texinfo: Document YYFPRINTF, YYSTDERR. | |
363 | ||
364 | 2001-12-13 Akim Demaille <akim@epita.fr> | |
365 | ||
366 | The computation of nullable is broken: it doesn't handle empty | |
367 | RHS's properly. | |
368 | ||
369 | * tests/torture.at (GNU AWK Grammar): New. | |
370 | * tests/sets.at (Nullable): New. | |
371 | * src/nullable.c (set_nullable): Instead of blindly looping over | |
372 | `ritems', loop over the rules, and then over their rhs's. | |
373 | ||
374 | Work around Autotest bugs. | |
375 | ||
376 | * src/warshall.c (bitmatrix_print): Don't use `+--+' as table | |
377 | frame, because Autotest understand lines starting with a `+' as | |
378 | traces from the shell. Then, they are not processed properly. | |
379 | Admittedly an Autotest bug, but we don't have time to wait for | |
380 | Autotest to catch up. | |
381 | * tests/regression.at (Broken Closure): Adjust to the new table | |
382 | frames. | |
383 | Move to... | |
384 | * tests/sets.at: here. | |
385 | ||
386 | 2001-12-13 Akim Demaille <akim@epita.fr> | |
387 | ||
388 | * src/closure.c (closure): Use nrules instead of playing tricks | |
389 | with BITS_PER_WORD. | |
390 | ||
391 | 2001-12-13 Akim Demaille <akim@epita.fr> | |
392 | ||
393 | * src/print.c (print_actions): Output the handling of `$' as the | |
394 | traces do: shifting the token EOF. Before EOF was treated as a | |
395 | nonterminal. | |
396 | * tests/regression.at: Adjust some tests. | |
397 | * src/print_graph.c (print_core): Complete the set of items via | |
398 | closure. The next-to-final and final states are still unsatisfying, | |
399 | but that's to be addressed elsewhere. | |
400 | No longer output the rule numbers, but do output the state number. | |
401 | A single loop for the shifts + gotos is enough, but picked a | |
402 | distinct color for each. | |
403 | (print_graph): Initialize and finalize closure. | |
404 | ||
405 | 2001-12-13 Akim Demaille <akim@epita.fr> | |
406 | ||
407 | * src/reader.c (readgram): Remove dead code, an strip useless | |
408 | braces. | |
409 | (get_type): Remove, unused. | |
410 | ||
411 | 2001-12-12 Akim Demaille <akim@epita.fr> | |
412 | ||
413 | * src/complain.h, src/complain.c: Remove error_one_per_line, rely | |
414 | on that of lib/error.c. | |
415 | ||
416 | 2001-12-12 Akim Demaille <akim@epita.fr> | |
417 | ||
418 | Some hosts don't like `/' in includes. | |
419 | ||
420 | * src/system.h: Include libgettext.h without qualifying the path. | |
421 | * src/Makefile.am (INCLUDES): Add $(top_srcdir)/intl, remove | |
422 | $(top_srcdir). | |
423 | ||
424 | 2001-12-11 Marc Autret <autret_m@epita.fr> | |
425 | ||
426 | * src/output.c (output_parser): Remove useless muscle. | |
427 | ||
428 | 2001-12-11 Marc Autret <autret_m@epita.fr> | |
429 | ||
430 | * src/bison.simple: Remove #line just before %%epilogue. It | |
431 | is now handled in ... | |
432 | * src/reader.c (read_additionnal_code): Add the output of a | |
433 | #line for the epilogue. | |
434 | ||
435 | 2001-12-10 Marc Autret <autret_m@epita.fr> | |
436 | ||
437 | * src/reader.c (copy_definition): Re-use CPP-outed code which | |
438 | replace precedent remove. | |
439 | * src/bison.simple: Remove #line before %%prologue because | |
440 | %%input-line is wrong at this time. | |
441 | ||
442 | 2001-12-10 Marc Autret <autret_m@epita.fr> | |
443 | ||
444 | * src/reader.c (symbols_output): Clean up. | |
445 | * src/output.c (output_gram, output): Clean up. | |
446 | ||
447 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
448 | ||
449 | * src/lalr.c (initialize_lookaheads): New. Extracted from... | |
450 | * src/LR0.c (set_state_table): here. | |
451 | * src/lalr.c (lalr): Call it. | |
452 | ||
453 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
454 | ||
455 | * src/state.h (shifts): Remove the `number' member: shifts are | |
456 | attached to state, hence no longer need to be labelled with a | |
457 | state number. | |
458 | ||
459 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
460 | ||
461 | Now that states have a complete set of members, the linked list of | |
462 | shifts is useless: just fill directly the state's shifts member. | |
463 | ||
464 | * src/state.h (shifts): Remove the `next' member. | |
465 | * src/LR0.c (first_state, last_state): Remove. | |
466 | Adjust the callers. | |
467 | (augment_automaton): Don't look for the shifts that must be added | |
468 | a shift on EOF: it is those of the state we looked for! But now, | |
469 | since shifts are attached, it is no longer needed to looking | |
470 | merely by its id: its number. | |
471 | ||
472 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
473 | ||
474 | * src/LR0.c (augment_automaton): Better variable locality. | |
475 | Remove an impossible branch: if there is a state corresponding to | |
476 | the start symbol being shifted, then there is shift for the start | |
477 | symbol from the initial state. | |
478 | ||
479 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
480 | ||
481 | * src/LR0.c (augment_automaton): Call `insert_eof_shifting_state' | |
482 | only when appropriate: when insert_start_shifting_state' is not | |
483 | invoked. | |
484 | * tests/regression.at (Rule Line Numbers): Adjust. | |
485 | ||
486 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
487 | ||
488 | * src/LR0.c (augment_automaton): Now that all states have shifts, | |
489 | merge the two cases addition shifts to the initial state. | |
490 | ||
491 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
492 | ||
493 | * src/lalr.c (set_state_table): Move to... | |
494 | * src/LR0.c: here. | |
495 | * src/lalr.c (lalr): Don't call it... | |
496 | * src/LR0.c (generate_states): do it. | |
497 | * src/LR0.h (first_state): Remove, only the table is used. | |
498 | ||
499 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
500 | ||
501 | * src/LR0.h (first_shift, first_reduction): Remove. | |
502 | * src/lalr.c: Don't use first_shift: find shifts through the | |
503 | states. | |
504 | ||
505 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
506 | ||
507 | * src/LR0.c: Attach shifts to states as soon as they are | |
508 | computed. | |
509 | * src/lalr.c (set_state_table): Instead of assigning shifts to | |
510 | state, just assert that the mapping was properly done. | |
511 | ||
512 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
513 | ||
514 | * src/LR0.c (insert_start_shift): Rename as... | |
515 | (insert_start_shifting_state): this. | |
516 | (insert_eof_shifting_state, insert_accepting_state): New. | |
517 | (augment_automaton): Adjust. | |
518 | Better locality of the variables. | |
519 | When looking if the start_symbol is shifted from the initial | |
520 | state, using `while (... symbol != start_symbol ...)' sounds | |
521 | better than `while (... symbol < start_symbol ...)': If fail | |
522 | to see how the order between symbols could be relevant! | |
523 | ||
524 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
525 | ||
526 | * src/getargs.h: Don't declare `spec_name_prefix' and | |
527 | `spec_file_prefix', declared by src/files.h. | |
528 | * src/files.c, src/files.h: Default for spec_name_prefix is "yy". | |
529 | * src/muscle_tab.c (muscle_init): Default prefix to NULL. | |
530 | * src/output.c (prepare): Adjust. | |
531 | * src/reader.c (symbols_output): Likewise. | |
532 | * src/vmsgetargs.c: Vaguely adjust, but who cares? | |
533 | ||
534 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
535 | ||
536 | * src/muscle_tab.c (muscle_init): NULL is a better default than | |
537 | `"0"'. | |
538 | ||
539 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
540 | ||
541 | * src/reader.c (reader): Calling symbols_output once is enough. | |
542 | ||
543 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
544 | ||
545 | Now that states have a complete set of members, the linked list of | |
546 | reductions is useless: just fill directly the state's reductions | |
547 | member. | |
548 | ||
549 | * src/state.h (struct reductions): Remove member `number' and | |
550 | `next'. | |
551 | * src/LR0.c (first_reduction, last_reduction): Remove. | |
552 | (save_reductions): Don't link the new reductions, store them in | |
553 | this_state. | |
554 | * src/lalr.c (set_state_table): No need to attach reductions to | |
555 | states, it's already done. | |
556 | * src/output.c (output_actions): No longer free the shifts, then | |
557 | the reductions, then the states: free all the states and their | |
558 | members. | |
559 | ||
560 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
561 | ||
562 | * src/options.c (OPTN, DRTV, BOTH): New. | |
563 | (option_table): Use them. | |
564 | ||
565 | * src/muscle_tab.c: Don't include xalloc.h and string.h: that's | |
566 | the job of system.h. | |
567 | * src/options.c: Don't include stdio.h and xalloc.h for the same | |
568 | reasons. | |
569 | ||
570 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
571 | ||
572 | * src/output.c (output, prepare): Make sure the values of the | |
573 | muscles `action' and `prologue' are 0-terminated. | |
574 | ||
575 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
576 | ||
577 | Clean up GCC warnings. | |
578 | ||
579 | * src/reader.c (copy_action): `buf' is not used. | |
580 | (parse_skel_decl): Be static. | |
581 | * src/muscle_tab.c (mhash1, mhash2, muscle_insert): Preserve `const'. | |
582 | * src/options.h (create_long_option_table): Have a real prototype. | |
583 | * lib/hash.c, lib/hash.h (hash_insert, hash_insert_at, hash_delete) | |
584 | (hash_delete_at): Return const void *. | |
585 | Adjust casts to preserve the const. | |
586 | ||
587 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
588 | ||
589 | * configure.in: Require 2.52g. | |
590 | M4 is not needed, but AUTOM4TE is. | |
591 | * m4/m4.m4: Remove. | |
592 | * tests/Makefile.am: Adjust. | |
593 | ||
594 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
595 | ||
596 | One structure for states is enough, even though theoretically | |
597 | there are LR(0) states and LALR(1) states. | |
598 | ||
599 | * src/lalr.h (state_t): Remove. | |
600 | (state_table): Be state_t **, not state_t *. | |
601 | * src/state.h (core, CORE_ALLOC): Rename as... | |
602 | (state_t, STATE_ALLOC): this. | |
603 | Add the LALR(1) members: shifts, reductions, errs. | |
604 | * src/LR0.c (state_table): Rename as... | |
605 | (state_hash): this, to avoid name clashes with the global | |
606 | `state_table'. | |
607 | * src/print_graph.c, src/LR0.c, src/LR0.h, src/conflicts.c | |
608 | * src/lalr.c, src/lalr.h, src/output.c, src/print.c: Adjust. | |
609 | ||
610 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
611 | ||
612 | Bison dumps core on bash.y. | |
613 | Reported by Pascal Bart. | |
614 | ||
615 | * src/warshall.c (bitmatrix_print): New. | |
616 | (TC): Use it. | |
617 | When performing a transitive closure R(i, j) && R(j, k) => R(i, k), | |
618 | j must be the outer loop. | |
619 | * tests/regression.at (Broken Closure): New. | |
620 | ||
621 | 2001-12-05 Akim Demaille <akim@epita.fr> | |
622 | ||
623 | * tests/atlocal.in (CPPFLAGS): Do not leave a space between -I and | |
624 | its argument. | |
625 |