Meredith L. Patterson
|
f37a13ef41
|
Fixed the 'unknown HCF_CHOICE' problem with indirect; test_leftrec is not an LL(k) test
|
2013-05-23 03:38:22 -07:00 |
|
Sven M. Hallberg
|
0b98b19ef0
|
Merge remote-tracking branch 'mlp/master' into LL
|
2013-05-23 11:58:05 +02:00 |
|
Sven M. Hallberg
|
7a681faeeb
|
remove a dead line
|
2013-05-23 11:57:46 +02:00 |
|
Meredith L. Patterson
|
3978266651
|
further merging of pesco's and aegis' changes
|
2013-05-23 02:16:52 -07:00 |
|
Meredith L. Patterson
|
1c7e9947a4
|
merging in aegis' JNI bindings
|
2013-05-23 00:58:27 -07:00 |
|
Meredith L. Patterson
|
570c798be7
|
Regex tests too. Only int_range() and nothing() work.
|
2013-05-23 00:16:08 -07:00 |
|
Meredith L. Patterson
|
22b0471050
|
Only LL(k)'s many() and many1(), and benchmark 1, are failing.
|
2013-05-23 00:12:47 -07:00 |
|
Meredith L. Patterson
|
10584ba22a
|
Refactor for separate tests for packrat and LL(k).
|
2013-05-22 23:50:48 -07:00 |
|
Sven M. Hallberg
|
2c8d76d7ea
|
refactor and bughunt unsuccessfully trying to find out why
h_first(2, ...) won't give me length-2 strings
|
2013-05-22 22:45:25 +02:00 |
|
Sven M. Hallberg
|
289c57aef5
|
h_stringmap_update wasn't merging results correctly
|
2013-05-22 21:58:34 +02:00 |
|
Sven M. Hallberg
|
e4d0b636dd
|
add h_hashtable_merge function
|
2013-05-22 21:57:46 +02:00 |
|
Sven M. Hallberg
|
194ddee90b
|
generalize grammar data structure allocation to k>1
|
2013-05-22 21:10:47 +02:00 |
|
Sven M. Hallberg
|
46525ede2d
|
remove some left-over dead code that I think is bullshit
|
2013-05-22 20:51:23 +02:00 |
|
Sven M. Hallberg
|
475d0ea9f3
|
comma-separate properly in h_pprint_stringset
|
2013-05-22 20:49:05 +02:00 |
|
Sven M. Hallberg
|
f5d5c36756
|
generalize grammar analysis to k>1
|
2013-05-22 20:38:36 +02:00 |
|
Sven M. Hallberg
|
337cbc2695
|
ugh, forgot to commit actions.c!
|
2013-05-22 12:20:44 +02:00 |
|
Sven M. Hallberg
|
91c04314a4
|
refactor h_llk_lookup to take the input stream as argument
|
2013-05-20 21:28:16 +02:00 |
|
Sven M. Hallberg
|
fb3aae7aec
|
fix free of previous backend data
|
2013-05-20 17:10:38 +02:00 |
|
Sven M. Hallberg
|
1afef6c3d2
|
reshape for h_attr_bool
|
2013-05-20 16:59:38 +02:00 |
|
Sven M. Hallberg
|
f48903d85e
|
reshape for h_ignore
|
2013-05-20 16:31:37 +02:00 |
|
Sven M. Hallberg
|
a42d97f053
|
fix desugar_optional
|
2013-05-20 16:25:42 +02:00 |
|
Sven M. Hallberg
|
3be766b741
|
reshape for h_epsilon
|
2013-05-20 16:25:02 +02:00 |
|
Sven M. Hallberg
|
e03513b59a
|
reshape for h_sequence
|
2013-05-20 15:32:15 +02:00 |
|
Sven M. Hallberg
|
e5300e2836
|
roll the positive cases for int16, int32 back into the existing test
|
2013-05-20 15:26:14 +02:00 |
|
Sven M. Hallberg
|
482851b3e2
|
reshape for h_action and h_choice
|
2013-05-20 15:24:19 +02:00 |
|
Sven M. Hallberg
|
29cee318f9
|
move the h_act_* functions to src/actions.c
|
2013-05-20 14:58:20 +02:00 |
|
Sven M. Hallberg
|
31256ba867
|
reshape for h_left and friends
|
2013-05-20 14:44:18 +02:00 |
|
Sven M. Hallberg
|
745aaebe02
|
fix desugaring of h_whitespace and add reshape action
|
2013-05-14 18:29:18 +02:00 |
|
Sven M. Hallberg
|
d081cf4284
|
reshape bits parsers
|
2013-05-14 17:20:05 +02:00 |
|
Sven M. Hallberg
|
c794be5b6f
|
consider HCF_CHARSET terminal symbols (as the LL driver already treats them)
|
2013-05-14 16:15:58 +02:00 |
|
Sven M. Hallberg
|
efae603c6b
|
fix an off-by-one alloc error in desugar_bits
|
2013-05-14 16:12:09 +02:00 |
|
Sven M. Hallberg
|
4138506273
|
unwrap (reshape) singleton terminal symbols correctly
|
2013-05-14 14:14:31 +02:00 |
|
Sven M. Hallberg
|
010d957fe8
|
add token reshaper
|
2013-05-14 12:14:45 +02:00 |
|
Sven M. Hallberg
|
1fac551429
|
add reshapers to 'many' parsers (makes the benchmark run with llk!)
|
2013-05-14 11:55:30 +02:00 |
|
Sven M. Hallberg
|
623fd6b030
|
add "reshape" action to HCFChoice
|
2013-05-14 11:54:43 +02:00 |
|
Sven M. Hallberg
|
0431528065
|
detect parse error in llk dummy test
|
2013-05-14 11:51:54 +02:00 |
|
Sven M. Hallberg
|
b488e5485f
|
always put end_token in the start symbol's follow set
|
2013-05-12 16:51:35 +02:00 |
|
Sven M. Hallberg
|
f33390a510
|
fix memory issues in benchmark.c
|
2013-05-12 02:54:28 +02:00 |
|
Sven M. Hallberg
|
fa1e59fe68
|
Revert "don't touch parser->backend in benchmark.c"
This reverts commit d81d820dca.
|
2013-05-12 01:18:46 +02:00 |
|
Sven M. Hallberg
|
75326263df
|
silence a warning when compiling with NDEBUG
|
2013-05-12 00:20:42 +02:00 |
|
Sven M. Hallberg
|
1c8764e1da
|
minor cosmetics
|
2013-05-12 00:20:27 +02:00 |
|
Sven M. Hallberg
|
fdcfc243b8
|
formatting
|
2013-05-11 23:41:47 +02:00 |
|
Sven M. Hallberg
|
d81d820dca
|
don't touch parser->backend in benchmark.c
|
2013-05-11 22:24:49 +02:00 |
|
Sven M. Hallberg
|
f1810d8be6
|
make h_llktable_free a no-op on NULL, just to be safe
|
2013-05-11 22:22:02 +02:00 |
|
Sven M. Hallberg
|
31c5189974
|
call previous backend's free function in h_compile
|
2013-05-11 22:21:30 +02:00 |
|
Sven M. Hallberg
|
1fe6b2ade9
|
add h_packrat_free
|
2013-05-11 22:19:36 +02:00 |
|
Sven M. Hallberg
|
286458f58a
|
parser->data got renamed to backend_data, also h_llk_free
|
2013-05-11 22:02:59 +02:00 |
|
Sven M. Hallberg
|
a1105d413a
|
make h_parse_result_free a no-op on NULL
|
2013-05-11 21:45:02 +02:00 |
|
Sven M. Hallberg
|
207fc6adef
|
fix (I think?) the test case failure detection login in benchmark.c
|
2013-05-11 21:44:01 +02:00 |
|
Sven M. Hallberg
|
62d48d2f5a
|
derp, forgot that table lookup can fail
|
2013-05-11 21:41:53 +02:00 |
|