]>
Commit | Line | Data |
---|---|---|
1 | 2001-12-12 Akim Demaille <akim@epita.fr> | |
2 | ||
3 | * src/complain.h, src/complain.c: Remove error_one_per_line, rely | |
4 | on that of lib/error.c. | |
5 | ||
6 | 2001-12-12 Akim Demaille <akim@epita.fr> | |
7 | ||
8 | Some hosts don't like `/' in includes. | |
9 | ||
10 | * src/system.h: Include libgettext.h without qualifying the path. | |
11 | * src/Makefile.am (INCLUDES): Add $(top_srcdir)/intl, remove | |
12 | $(top_srcdir). | |
13 | ||
14 | 2001-12-11 Marc Autret <autret_m@epita.fr> | |
15 | ||
16 | * src/output.c (output_parser): Remove useless muscle. | |
17 | ||
18 | 2001-12-11 Marc Autret <autret_m@epita.fr> | |
19 | ||
20 | * src/bison.simple: Remove #line just before %%epilogue. It | |
21 | is now handled in ... | |
22 | * src/reader.c (read_additionnal_code): Add the output of a | |
23 | #line for the epilogue. | |
24 | ||
25 | 2001-12-10 Marc Autret <autret_m@epita.fr> | |
26 | ||
27 | * src/reader.c (copy_definition): Re-use CPP-outed code which | |
28 | replace precedent remove. | |
29 | * src/bison.simple: Remove #line before %%prologue because | |
30 | %%input-line is wrong at this time. | |
31 | ||
32 | 2001-12-10 Marc Autret <autret_m@epita.fr> | |
33 | ||
34 | * src/reader.c (symbols_output): Clean up. | |
35 | * src/output.c (output_gram, output): Clean up. | |
36 | ||
37 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
38 | ||
39 | * src/lalr.c (initialize_lookaheads): New. Extracted from... | |
40 | * src/LR0.c (set_state_table): here. | |
41 | * src/lalr.c (lalr): Call it. | |
42 | ||
43 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
44 | ||
45 | * src/state.h (shifts): Remove the `number' member: shifts are | |
46 | attached to state, hence no longer need to be labelled with a | |
47 | state number. | |
48 | ||
49 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
50 | ||
51 | Now that states have a complete set of members, the linked list of | |
52 | shifts is useless: just fill directly the state's shifts member. | |
53 | ||
54 | * src/state.h (shifts): Remove the `next' member. | |
55 | * src/LR0.c (first_state, last_state): Remove. | |
56 | Adjust the callers. | |
57 | (augment_automaton): Don't look for the shifts that must be added | |
58 | a shift on EOF: it is those of the state we looked for! But now, | |
59 | since shifts are attached, it is no longer needed to looking | |
60 | merely by its id: its number. | |
61 | ||
62 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
63 | ||
64 | * src/LR0.c (augment_automaton): Better variable locality. | |
65 | Remove an impossible branch: if there is a state corresponding to | |
66 | the start symbol being shifted, then there is shift for the start | |
67 | symbol from the initial state. | |
68 | ||
69 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
70 | ||
71 | * src/LR0.c (augment_automaton): Call `insert_eof_shifting_state' | |
72 | only when appropriate: when insert_start_shifting_state' is not | |
73 | invoked. | |
74 | * tests/regression.at (Rule Line Numbers): Adjust. | |
75 | ||
76 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
77 | ||
78 | * src/LR0.c (augment_automaton): Now that all states have shifts, | |
79 | merge the two cases addition shifts to the initial state. | |
80 | ||
81 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
82 | ||
83 | * src/lalr.c (set_state_table): Move to... | |
84 | * src/LR0.c: here. | |
85 | * src/lalr.c (lalr): Don't call it... | |
86 | * src/LR0.c (generate_states): do it. | |
87 | * src/LR0.h (first_state): Remove, only the table is used. | |
88 | ||
89 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
90 | ||
91 | * src/LR0.h (first_shift, first_reduction): Remove. | |
92 | * src/lalr.c: Don't use first_shift: find shifts through the | |
93 | states. | |
94 | ||
95 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
96 | ||
97 | * src/LR0.c: Attach shifts to states as soon as they are | |
98 | computed. | |
99 | * src/lalr.c (set_state_table): Instead of assigning shifts to | |
100 | state, just assert that the mapping was properly done. | |
101 | ||
102 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
103 | ||
104 | * src/LR0.c (insert_start_shift): Rename as... | |
105 | (insert_start_shifting_state): this. | |
106 | (insert_eof_shifting_state, insert_accepting_state): New. | |
107 | (augment_automaton): Adjust. | |
108 | Better locality of the variables. | |
109 | When looking if the start_symbol is shifted from the initial | |
110 | state, using `while (... symbol != start_symbol ...)' sounds | |
111 | better than `while (... symbol < start_symbol ...)': If fail | |
112 | to see how the order between symbols could be relevant! | |
113 | ||
114 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
115 | ||
116 | * src/getargs.h: Don't declare `spec_name_prefix' and | |
117 | `spec_file_prefix', declared by src/files.h. | |
118 | * src/files.c, src/files.h: Default for spec_name_prefix is "yy". | |
119 | * src/muscle_tab.c (muscle_init): Default prefix to NULL. | |
120 | * src/output.c (prepare): Adjust. | |
121 | * src/reader.c (symbols_output): Likewise. | |
122 | * src/vmsgetargs.c: Vaguely adjust, but who cares? | |
123 | ||
124 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
125 | ||
126 | * src/muscle_tab.c (muscle_init): NULL is a better default than | |
127 | `"0"'. | |
128 | ||
129 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
130 | ||
131 | * src/reader.c (reader): Calling symbols_output once is enough. | |
132 | ||
133 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
134 | ||
135 | Now that states have a complete set of members, the linked list of | |
136 | reductions is useless: just fill directly the state's reductions | |
137 | member. | |
138 | ||
139 | * src/state.h (struct reductions): Remove member `number' and | |
140 | `next'. | |
141 | * src/LR0.c (first_reduction, last_reduction): Remove. | |
142 | (save_reductions): Don't link the new reductions, store them in | |
143 | this_state. | |
144 | * src/lalr.c (set_state_table): No need to attach reductions to | |
145 | states, it's already done. | |
146 | * src/output.c (output_actions): No longer free the shifts, then | |
147 | the reductions, then the states: free all the states and their | |
148 | members. | |
149 | ||
150 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
151 | ||
152 | * src/options.c (OPTN, DRTV, BOTH): New. | |
153 | (option_table): Use them. | |
154 | ||
155 | * src/muscle_tab.c: Don't include xalloc.h and string.h: that's | |
156 | the job of system.h. | |
157 | * src/options.c: Don't include stdio.h and xalloc.h for the same | |
158 | reasons. | |
159 | ||
160 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
161 | ||
162 | * src/output.c (output, prepare): Make sure the values of the | |
163 | muscles `action' and `prologue' are 0-terminated. | |
164 | ||
165 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
166 | ||
167 | Clean up GCC warnings. | |
168 | ||
169 | * src/reader.c (copy_action): `buf' is not used. | |
170 | (parse_skel_decl): Be static. | |
171 | * src/muscle_tab.c (mhash1, mhash2, muscle_insert): Preserve `const'. | |
172 | * src/options.h (create_long_option_table): Have a real prototype. | |
173 | * lib/hash.c, lib/hash.h (hash_insert, hash_insert_at, hash_delete) | |
174 | (hash_delete_at): Return const void *. | |
175 | Adjust casts to preserve the const. | |
176 | ||
177 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
178 | ||
179 | * configure.in: Require 2.52g. | |
180 | M4 is not needed, but AUTOM4TE is. | |
181 | * m4/m4.m4: Remove. | |
182 | * tests/Makefile.am: Adjust. | |
183 | ||
184 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
185 | ||
186 | One structure for states is enough, even though theoretically | |
187 | there are LR(0) states and LALR(1) states. | |
188 | ||
189 | * src/lalr.h (state_t): Remove. | |
190 | (state_table): Be state_t **, not state_t *. | |
191 | * src/state.h (core, CORE_ALLOC): Rename as... | |
192 | (state_t, STATE_ALLOC): this. | |
193 | Add the LALR(1) members: shifts, reductions, errs. | |
194 | * src/LR0.c (state_table): Rename as... | |
195 | (state_hash): this, to avoid name clashes with the global | |
196 | `state_table'. | |
197 | * src/print_graph.c, src/LR0.c, src/LR0.h, src/conflicts.c | |
198 | * src/lalr.c, src/lalr.h, src/output.c, src/print.c: Adjust. | |
199 | ||
200 | 2001-12-10 Akim Demaille <akim@epita.fr> | |
201 | ||
202 | Bison dumps core on bash.y. | |
203 | Reported by Pascal Bart. | |
204 | ||
205 | * src/warshall.c (bitmatrix_print): New. | |
206 | (TC): Use it. | |
207 | When performing a transitive closure R(i, j) && R(j, k) => R(i, k), | |
208 | j must be the outer loop. | |
209 | * tests/regression.at (Broken Closure): New. | |
210 | ||
211 | 2001-12-05 Akim Demaille <akim@epita.fr> | |
212 | ||
213 | * tests/atlocal.in (CPPFLAGS): Do not leave a space between -I and | |
214 | its argument. | |
215 |