 kr.angelov
|
3ad5493758
|
Use a separated tag for meta productions in the robust parser. This cleans up the code a lot
|
2012-06-13 05:49:30 +00:00 |
|
|
kr.angelov
|
c9c5675e1d
|
now there is a limit of 2000000 items in the chart of the robust parser. This prevents from explosion in the memory size but it will also prevent us from parsing some sentences.
|
2012-06-12 11:30:01 +00:00 |
|
|
kr.angelov
|
808c52744f
|
improved script for training from PennTreebank
|
2012-06-12 10:05:17 +00:00 |
|
|
kr.angelov
|
b27a440ef3
|
now the robust parser is purely top-down and the meta rules compete on a fair basis with the grammar rules
|
2012-06-12 09:29:51 +00:00 |
|
|
kr.angelov
|
06f9965d27
|
the viterbi probability for the epsilon categories is now updated properly
|
2012-05-25 07:30:35 +00:00 |
|
|
kr.angelov
|
ea5fe32d89
|
fixed an alignment bug in libgu which leads to crash on some platforms
|
2012-05-25 07:29:29 +00:00 |
|
|
kr.angelov
|
92488a6592
|
fixed memory leak in the pgf-translate tool
|
2012-05-25 07:28:18 +00:00 |
|
|
kr.angelov
|
5312ef23c6
|
added pgf-service.c in the robust parser'
|
2012-05-25 07:27:29 +00:00 |
|
|
kr.angelov
|
f4c17cb7aa
|
another attempt to port the robust parser to MacOS
|
2012-05-16 15:18:44 +00:00 |
|
|
kr.angelov
|
d2e528c7f4
|
small fix for compiling the robust parser on MacOS
|
2012-05-16 12:18:50 +00:00 |
|
|
kr.angelov
|
a6800fc0da
|
a new unbiased statistical parser. it is still far from perfect use it on your own risk.
|
2012-05-08 12:13:28 +00:00 |
|
|
kr.angelov
|
931066f6fc
|
yet another fix for parsing literals
|
2012-04-18 15:50:55 +00:00 |
|
|
kr.angelov
|
17bc8e5c89
|
some fixes in the robust parser and a new API for literals
|
2012-04-12 06:55:25 +00:00 |
|
|
kr.angelov
|
6644d93ec2
|
simple cleanup in the robust parser
|
2012-04-02 19:01:18 +00:00 |
|
|
kr.angelov
|
230f309317
|
libpgf: a new implementation for literals which also allows custom literals. the same mechanism is now used for the metavariables
|
2012-03-12 14:25:51 +00:00 |
|
|
kr.angelov
|
1726995921
|
libpgf: added simple lexer
|
2012-03-09 09:14:44 +00:00 |
|
|
kr.angelov
|
ed5de8335b
|
libpgf: implementation for built in literal categories
|
2012-03-07 16:39:29 +00:00 |
|
|
kr.angelov
|
96493c274b
|
libpgf: simple fix in the parser debugger
|
2012-03-07 12:23:07 +00:00 |
|
|
kr.angelov
|
a96da30489
|
libpgf: two APIs - one for finding all parse results and another for finding the best parse result
|
2012-03-07 11:00:17 +00:00 |
|
|
kr.angelov
|
0e90d1ba1f
|
libpgf: now all concrete functions and categories are explicitly linked to their abstract counter parts
|
2012-03-05 12:59:31 +00:00 |
|
|
kr.angelov
|
4d1b0859d0
|
libpgf: preliminary version for the statistical ranking. we use naive statistical model with random weight for the meta variables.
|
2012-03-02 19:25:01 +00:00 |
|
|
kr.angelov
|
58b6bbd242
|
libpgf: simple optimization in the implementation for heaps
|
2012-03-02 17:43:46 +00:00 |
|
|
kr.angelov
|
2775faad7a
|
libpgf: an API for building heaps from buffers in gu/seq.c
|
2012-02-29 18:51:21 +00:00 |
|
|
kr.angelov
|
e31c883075
|
libpgf: the first prototype for the robust parser
|
2012-02-29 14:43:08 +00:00 |
|
|
kr.angelov
|
b054549082
|
libpgf: the choice object should remember integers instead of single bytes
|
2012-02-29 13:41:59 +00:00 |
|
|
kr.angelov
|
5fa1418194
|
libpgf: another fix in the parser debugger
|
2012-02-28 16:37:12 +00:00 |
|
|
kr.angelov
|
4ddf24429b
|
libpgf: the time measurement in pgf-translate should be made after pgf_parse_result for more accurate timing
|
2012-02-28 13:14:24 +00:00 |
|
|
kr.angelov
|
dcbeb63849
|
libpgf: fix in the parser debugger
|
2012-02-28 13:12:38 +00:00 |
|
|
kr.angelov
|
eb4bfb008e
|
buf fix in the compiler for HOAS
|
2012-02-28 08:10:47 +00:00 |
|
|
kr.angelov
|
686fcca9a4
|
libpgf: pretty printing for expressions with metavaraibles
|
2012-02-27 13:50:35 +00:00 |
|
|
kr.angelov
|
667ca8e5f7
|
libpgf: fix in pgf_read_into_map
|
2012-02-24 15:15:07 +00:00 |
|
|
kr.angelov
|
1d239ab567
|
libpgf: simple fix in the grammar printer and the reader
|
2012-02-24 13:52:21 +00:00 |
|
|
kr.angelov
|
b99fa6aa9a
|
libpgf: now we have both complete bottom up index for robust parsing and fast lexical lookup from the same index
|
2012-02-22 21:27:54 +00:00 |
|
|
kr.angelov
|
42410f80d2
|
libpgf: two small fixes in the parser debugger
|
2012-02-22 14:06:49 +00:00 |
|
|
kr.angelov
|
786bf883b7
|
libpgf: the map curr_lindefs must be allocated from a temporary pool
|
2012-02-22 08:49:08 +00:00 |
|
|
kr.angelov
|
7ddd0d5f3e
|
libpgf: added index for fast lexicon lookup. Still not perfect
|
2012-02-21 21:17:50 +00:00 |
|
|
kr.angelov
|
a178608f37
|
utils/pgf-translate now reports the parsing time
|
2012-02-21 20:17:18 +00:00 |
|
|
kr.angelov
|
e725b10d66
|
fixes in the script for the conversion of PennTreebank
|
2012-02-21 17:26:04 +00:00 |
|
|
kr.angelov
|
a55a224dce
|
libpgf: now the debugging mode for the parser is available only with compilation option.
|
2012-02-18 19:30:16 +00:00 |
|
|
kr.angelov
|
5259be92e3
|
libpgf: remove the now redundant field extra_ccats in PgfConcr
|
2012-02-18 16:25:53 +00:00 |
|
|
kr.angelov
|
47e5e8c966
|
libpgf: now the linearization index is created during the grammar loading which also makes the types PgfLzr and PgfParser redundant.
|
2012-02-18 16:22:40 +00:00 |
|
|
kr.angelov
|
aed7cc429a
|
libpgf: simplify the loading of PgfCncCat
|
2012-02-17 14:26:08 +00:00 |
|
|
kr.angelov
|
4f8ad8c23f
|
libpgf: added printer.c
|
2012-02-17 14:11:29 +00:00 |
|
|
kr.angelov
|
6f0795d8a3
|
libpgf: switch to using callbacks and lazy prediction in the parser. this reduce the parsing time from 11 sec down to 3 sec.
|
2012-01-26 12:32:26 +00:00 |
|
|
kr.angelov
|
a2414bc625
|
libpgf: use a temporal pool for allocating the arrays in the continuation map of the parser
|
2012-01-26 09:03:08 +00:00 |
|
|
kr.angelov
|
8256750599
|
libpgf: fix a warning in reader.c
|
2012-01-26 08:58:23 +00:00 |
|
|
kr.angelov
|
ecb5e9f23a
|
libpgf: fix in string.h
|
2012-01-26 08:58:03 +00:00 |
|
|
kr.angelov
|
58de345f99
|
libpgf: few fixes to make the loading of grammars with def rules possible
|
2012-01-24 14:47:11 +00:00 |
|
|
kr.angelov
|
acc4f352af
|
libpgf: extra_ccat is now redundant and was removed
|
2012-01-23 19:47:08 +00:00 |
|
|
kr.angelov
|
4f289698f0
|
libpgf: forgot to add utils/pgf-print.c
|
2012-01-23 16:34:02 +00:00 |
|
|
kr.angelov
|
1af48d34ec
|
pgf-translate: add \n at the end of two error messages
|
2012-01-23 15:51:37 +00:00 |
|
|
kr.angelov
|
5ccd75c8b9
|
libpgf: debugging framework for the parser
|
2012-01-23 15:49:29 +00:00 |
|
|
kr.angelov
|
fcd2c2905e
|
libpgf: the concrete categories were allocated from the temporary pool
|
2012-01-23 13:43:17 +00:00 |
|
|
kr.angelov
|
4e4881dba5
|
libpgf: printing of literals and flags
|
2012-01-23 10:17:20 +00:00 |
|
|
kr.angelov
|
7fcaa81858
|
libpgf: move the lindefs field from PgfCncCat to PgfCCat. display the list in the grammar printout
|
2012-01-23 09:46:45 +00:00 |
|
|
kr.angelov
|
28bce1cdc0
|
added an API for printing the PGF to human readable format
|
2012-01-21 10:27:55 +00:00 |
|
|
kr.angelov
|
d164ddfbe1
|
fix the crash in the reader when there aren't any productions per category.
|
2012-01-21 10:25:43 +00:00 |
|
|
kr.angelov
|
2eee382a62
|
initial import of the C runtime
|
2012-01-20 13:41:10 +00:00 |
|
|
kr.angelov
|
796dd530ee
|
the translation script from the Penn format to GF RGL is now in examples/PennTreebank
|
2012-01-10 19:36:28 +00:00 |
|
|
kr.angelov
|
bb6905e36f
|
the parser now use nub instead of nubsort which means that the abstract syntax trees will be returned lazily
|
2011-12-19 13:10:33 +00:00 |
|
|
kr.angelov
|
7c9bbd844b
|
Now graphvizAbstractTree suppress the visualization of implicit arguments.
|
2011-12-08 09:18:38 +00:00 |
|
|
kr.angelov
|
1160244271
|
Phrasebook is updated for Bulgarian
|
2011-12-06 13:54:25 +00:00 |
|
|
kr.angelov
|
c52f05ed46
|
The typechecker is still unfinished but at least it can typecheck the English resource grammar
|
2011-12-02 12:33:26 +00:00 |
|
|
kr.angelov
|
e6135d963e
|
missing cases in GF.Grammar.Binary
|
2011-12-02 09:58:18 +00:00 |
|
|
kr.angelov
|
b859edee38
|
Hopefully complete Value type and a little bit more on computations.
|
2011-11-30 20:56:31 +00:00 |
|
|
kr.angelov
|
7863b21c1a
|
more stuff in the new type checker
|
2011-11-30 14:55:52 +00:00 |
|
|
kr.angelov
|
f9af731c9c
|
fix the compilation after the patch for the new type checker
|
2011-11-29 18:45:51 +00:00 |
|
|
kr.angelov
|
fffc5f2507
|
Sketch of the new type checker for the concrete syntax. Enabled only with -new-comp
|
2011-11-29 12:12:51 +00:00 |
|
|
kr.angelov
|
c05be64838
|
now if some module is compiled with -no-pmcfg then the PMCFG code is generated at the end during the linking phase. Now the default compilation of the libraries with cabal is with -no-pmcfg.
|
2011-11-24 14:23:04 +00:00 |
|
|
kr.angelov
|
26d61354dd
|
make addSequencesB(V) strict. Otherwise we get stack overflow when compiling LangFre
|
2011-11-22 20:31:58 +00:00 |
|
|
kr.angelov
|
3d973adb33
|
rebuildModule should do location globalization too
|
2011-11-17 22:11:56 +00:00 |
|
|
kr.angelov
|
4f95b5656b
|
bugfix in the overload resolution. It was accidentally introduced as part of the per module PMCFG generation
|
2011-11-17 21:36:45 +00:00 |
|
|
kr.angelov
|
04fdb41927
|
fix in the versioning of the .gfo files
|
2011-11-15 19:46:02 +00:00 |
|
|
kr.angelov
|
a2626e24dd
|
now we store version number in every .gfo file. If the file is compiled with different compiler then we simply recompile it.
|
2011-11-15 19:12:22 +00:00 |
|
|
kr.angelov
|
8a10aa5cf9
|
now the pretty printer in GF has a new mode called Internal. This is the most detailed mode and it can print even things that are not in the GF syntax. For example PMCFG snippets and indirections.
|
2011-11-15 15:55:45 +00:00 |
|
|
kr.angelov
|
59e098a440
|
Now the errors messages from GF.Compile.Update also follow the new format
|
2011-11-15 15:33:52 +00:00 |
|
|
kr.angelov
|
b1a82e8e8a
|
the indirections in the .gf-tags files now point directly to the origin of the corresponding indentifiers
|
2011-11-15 14:30:31 +00:00 |
|
|
kr.angelov
|
61c16f2eb2
|
more structured format for errors and warnings from the compiler
|
2011-11-15 13:33:44 +00:00 |
|
|
kr.angelov
|
8a7aa15c57
|
bugfix in the new PGF generation
|
2011-11-14 16:27:59 +00:00 |
|
|
kr.angelov
|
43fc9f462a
|
remove the -output-file option
|
2011-11-14 16:10:13 +00:00 |
|
|
kr.angelov
|
7be8566f35
|
the new design for -tags
|
2011-11-14 16:08:56 +00:00 |
|
|
kr.angelov
|
416d231c5e
|
Now PMCFG is compiled per module and at the end we only link it. The new compilation schema is few times faster.
|
2011-11-10 14:09:41 +00:00 |
|
|
kr.angelov
|
b25ec6ae89
|
remove few obsolete options from GF.Infra.Options
|
2011-11-10 09:02:46 +00:00 |
|
|
kr.angelov
|
0d04d56241
|
qualified names in the tags file
|
2011-11-04 16:20:41 +00:00 |
|
|
kr.angelov
|
734c66710e
|
merge GF.Infra.Modules and GF.Grammar.Grammar. This is a preparation for the separate PGF building
|
2011-11-02 13:57:11 +00:00 |
|
|
kr.angelov
|
5fe49ed9f7
|
Now the compiler maintains more precise information for the source locations of the different definitions. There is a --tags option which generates a list of all identifiers with their source locations.
|
2011-11-02 11:44:59 +00:00 |
|
|
kr.angelov
|
8a263781d0
|
add Windows binary
|
2011-10-26 13:30:42 +00:00 |
|
|
kr.angelov
|
bb599029c9
|
change the precedence for the left argument of ->
|
2011-09-22 16:24:02 +00:00 |
|
|
kr.angelov
|
89fb9a7fdf
|
fix in the proof search. when we start the generation from a template, we must start the refinement from the expression that the typecheker generated
|
2011-09-21 13:30:09 +00:00 |
|
|
kr.angelov
|
2001788b02
|
few more fixes in the NLG grammar
|
2011-08-30 17:29:13 +00:00 |
|
|
kr.angelov
|
a29a4408a8
|
added quantifier ambiguity in the NLG example
|
2011-08-30 15:23:59 +00:00 |
|
|
kr.angelov
|
c70df9200c
|
added example for NLG from logical formula. See examples/nlg
|
2011-08-30 14:58:16 +00:00 |
|
|
kr.angelov
|
4215fc31bf
|
bugfixes in the typechecker and the tree generator
|
2011-08-30 11:24:59 +00:00 |
|
|
kr.angelov
|
41a60772c7
|
bugfix: allow higher-order variables in the linearization
|
2011-08-30 09:22:44 +00:00 |
|
|
kr.angelov
|
5339aa8074
|
fix SUMO after the update in DictEng
|
2011-08-23 15:05:04 +00:00 |
|
|
kr.angelov
|
873a160537
|
FCat is just a plain integer now
|
2007-10-12 16:00:37 +00:00 |
|
|
kr.angelov
|
acc32ec199
|
handle (F ..) references in the lintypes
|
2007-10-04 18:23:59 +00:00 |
|
|
kr.angelov
|
890b4ca7e0
|
don't try to search for argument that has been already found
|
2007-09-25 14:57:14 +00:00 |
|
|
kr.angelov
|
bb8e7d2d58
|
whitespace only
|
2007-09-25 14:56:46 +00:00 |
|
|
kr.angelov
|
6613d5f40a
|
simple optimization of the parser that makes it about two times faster
|
2007-09-25 12:59:25 +00:00 |
|