projects / bison.git / blame
| Commit | Line | Data |
|---|---|---|
| 80e25d4d AD |
1 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 2 | ||
| 3 | * src/LR0.c: Attach shifts to states as soon as they are | |
| 4 | computed. | |
| 5 | * src/lalr.c (set_state_table): Instead of assigning shifts to | |
| 6 | state, just assert that the mapping was properly done. | |
| 7 | ||
| 8 | ||
| 0ab3728b AD |
9 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 10 | ||
| 11 | * src/LR0.c (insert_start_shift): Rename as... | |
| 12 | (insert_start_shifting_state): this. | |
| 13 | (insert_eof_shifting_state, insert_accepting_state): New. | |
| 14 | (augment_automaton): Adjust. | |
| 15 | Better locality of the variables. | |
| 16 | When looking if the start_symbol is shifted from the initial | |
| 17 | state, using `while (... symbol != start_symbol ...)' sounds | |
| 18 | better than `while (... symbol < start_symbol ...)': If fail | |
| 19 | to see how the order between symbols could be relevant! | |
| 20 | ||
| 78af9bbc AD |
21 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 22 | ||
| 23 | * src/getargs.h: Don't declare `spec_name_prefix' and | |
| 24 | `spec_file_prefix', declared by src/files.h. | |
| 25 | * src/files.c, src/files.h: Default for spec_name_prefix is "yy". | |
| 26 | * src/muscle_tab.c (muscle_init): Default prefix to NULL. | |
| 27 | * src/output.c (prepare): Adjust. | |
| 28 | * src/reader.c (symbols_output): Likewise. | |
| 29 | * src/vmsgetargs.c: Vaguely adjust, but who cares? | |
| 30 | ||
| 31 | ||
| bdef2a41 AD |
32 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 33 | ||
| 34 | * src/muscle_tab.c (muscle_init): NULL is a better default than | |
| 35 | `"0"'. | |
| 36 | ||
| 37 | ||
| 3735969c AD |
38 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 39 | ||
| 40 | * src/reader.c (reader): Calling symbols_output once is enough. | |
| 41 | ||
| 42 | ||
| 49701457 AD |
43 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 44 | ||
| 45 | Now that states have a complete set of members, the linked list of | |
| 46 | reductions is useless: just fill directly the state's reductions | |
| 47 | member. | |
| 48 | ||
| 49 | * src/state.h (struct reductions): Remove member `number' and | |
| 50 | `next'. | |
| 51 | * src/LR0.c (first_reduction, last_reduction): Remove. | |
| 52 | (save_reductions): Don't link the new reductions, store them in | |
| 53 | this_state. | |
| 54 | * src/lalr.c (set_state_table): No need to attach reductions to | |
| 55 | states, it's already done. | |
| 56 | * src/output.c (output_actions): No longer free the shifts, then | |
| 57 | the reductions, then the states: free all the states and their | |
| 58 | members. | |
| 59 | ||
| 60 | ||
| 0edad749 AD |
61 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 62 | ||
| 63 | * src/options.c (OPTN, DRTV, BOTH): New. | |
| 64 | (option_table): Use them. | |
| 65 | ||
| 66 | ||
| 67 | * src/muscle_tab.c: Don't include xalloc.h and string.h: that's | |
| 68 | the job of system.h. | |
| 69 | * src/options.c: Don't include stdio.h and xalloc.h for the same | |
| 70 | reasons. | |
| 71 | ||
| 72 | ||
| 5449dd0f AD |
73 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 74 | ||
| 75 | * src/output.c (output, prepare): Make sure the values of the | |
| 76 | muscles `action' and `prologue' are 0-terminated. | |
| 77 | ||
| a870c567 AD |
78 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 79 | ||
| 80 | Clean up GCC warnings. | |
| 81 | ||
| 82 | * src/reader.c (copy_action): `buf' is not used. | |
| 83 | (parse_skel_decl): Be static. | |
| 84 | * src/muscle_tab.c (mhash1, mhash2, muscle_insert): Preserve `const'. | |
| 85 | * src/options.h (create_long_option_table): Have a real prototype. | |
| 86 | * lib/hash.c, lib/hash.h (hash_insert, hash_insert_at, hash_delete) | |
| 87 | (hash_delete_at): Return const void *. | |
| 88 | Adjust casts to preserve the const. | |
| 89 | ||
| 80df8768 AD |
90 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 91 | ||
| 92 | * configure.in: Require 2.52g. | |
| 93 | M4 is not needed, but AUTOM4TE is. | |
| 94 | * m4/m4.m4: Remove. | |
| 95 | * tests/Makefile.am: Adjust. | |
| 96 | ||
| 97 | ||
| f693ad14 AD |
98 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 99 | ||
| 100 | One structure for states is enough, even though theoretically | |
| 101 | there are LR(0) states and LALR(1) states. | |
| 102 | ||
| 103 | * src/lalr.h (state_t): Remove. | |
| 104 | (state_table): Be state_t **, not state_t *. | |
| 105 | * src/state.h (core, CORE_ALLOC): Rename as... | |
| 106 | (state_t, STATE_ALLOC): this. | |
| 107 | Add the LALR(1) members: shifts, reductions, errs. | |
| 108 | * src/LR0.c (state_table): Rename as... | |
| 109 | (state_hash): this, to avoid name clashes with the global | |
| 110 | `state_table'. | |
| 111 | * src/print_graph.c, src/LR0.c, src/LR0.h, src/conflicts.c | |
| 112 | * src/lalr.c, src/lalr.h, src/output.c, src/print.c: Adjust. | |
| 113 | ||
| 114 | ||
| 74ffbcb6 AD |
115 | 2001-12-10 Akim Demaille <akim@epita.fr> |
| 116 | ||
| 117 | Bison dumps core on bash.y. | |
| 118 | Reported by Pascal Bart. | |
| 119 | ||
| 120 | * src/warshall.c (bitmatrix_print): New. | |
| 121 | (TC): Use it. | |
| 122 | When performing a transitive closure R(i, j) && R(j, k) => R(i, k), | |
| 123 | j must be the outer loop. | |
| 124 | * tests/regression.at (Broken Closure): New. | |
| 125 | ||
| 07708e19 AD |
126 | 2001-12-05 Akim Demaille <akim@epita.fr> |
| 127 | ||
| 128 | * tests/atlocal.in (CPPFLAGS): Do not leave a space between -I and | |
| 129 | its argument. | |
| 130 |