krasimir
|
4426120eff
|
fix typo in CheckGrammar
|
2009-09-14 12:34:46 +00:00 |
|
krasimir
|
70fea39e2b
|
added some tests in the testsuite for the compiler
|
2009-09-14 12:17:57 +00:00 |
|
krasimir
|
62ef772a2c
|
CheckGrammar is now using the printer in GF.Grammar.Printer. Fixed bug that was hiding the warnings
|
2009-09-14 12:16:02 +00:00 |
|
krasimir
|
cc151c4279
|
added needsTypeCheck parameter to CommandInfo. The argument to the command is typechecked only if needsTypeCheck=True
|
2009-09-13 15:39:11 +00:00 |
|
krasimir
|
4111438979
|
the testsuite\runtime\parser test now generates random trees to test the parser with
|
2009-09-13 15:17:15 +00:00 |
|
krasimir
|
509d0cf8d8
|
added wrapper functions for expression manipulations in PGF. The Haskell API now uses the wrappers
|
2009-09-11 14:42:45 +00:00 |
|
krasimir
|
1cdf171251
|
polish the PGF API and make Expr and Type abstract types. Tree is a type synonym of Expr
|
2009-09-11 13:45:34 +00:00 |
|
krasimir
|
28a7c4b5c7
|
now the datatype Tree is only internal. All API functions are working with Expr directly. Commands gt, gr, p and rf filter out the output via the typechecker
|
2009-09-08 08:40:28 +00:00 |
|
krasimir
|
9940c44259
|
export Hypo(..) from PGF
|
2009-09-07 19:16:27 +00:00 |
|
krasimir
|
76517518a3
|
make the PMCFG generation lazy again. it was made strict when the profiler was introduced
|
2009-09-07 14:45:17 +00:00 |
|
krasimir
|
b97d6abb81
|
hopefully complete and correct typechecker in PGF
|
2009-09-06 20:31:52 +00:00 |
|
krasimir
|
c99b64404d
|
bugfixes in the parser for PGF.Type.Type
|
2009-07-07 22:22:05 +00:00 |
|
krasimir
|
bb3040e2c4
|
bugfix in PGF.Expr.apply
|
2009-07-05 16:07:34 +00:00 |
|
krasimir
|
279ff9a6d2
|
PGF.Type.Hypo now can represent explicit and implicit arguments and argument without bound variable
|
2009-07-05 15:44:52 +00:00 |
|
krasimir
|
3394c171ed
|
PGF.Expr.eval now returns suspension when a meta variable is encountered
|
2009-07-05 15:34:16 +00:00 |
|
krasimir
|
95a577d269
|
simple profiler for PMCFG
|
2009-09-05 14:04:39 +00:00 |
|
aarne
|
785ef9224b
|
revised FoodsHeb
|
2009-09-03 14:26:31 +00:00 |
|
krasimir
|
c93d0a3a26
|
the -dump-X options should convert to UTF8
|
2009-08-27 17:27:20 +00:00 |
|
aarne
|
db59a4ac72
|
food grammars added, with copyright statements
|
2009-08-29 17:36:15 +00:00 |
|
aarne
|
f34dd228eb
|
added contrib and summerschool
|
2009-08-28 07:05:08 +00:00 |
|
krasimir
|
b6abe5f8f5
|
if there is oper patt = #(...) then it also should be decoded to Unicode
|
2009-08-27 15:57:38 +00:00 |
|
aarne
|
7e395c261c
|
some corrections in Bronzeage
|
2009-08-27 13:49:00 +00:00 |
|
aarne
|
8c2cee0f4b
|
Persian transliteration
|
2009-08-27 13:47:57 +00:00 |
|
krasimir
|
dbf87958f9
|
let the 'se utf8' command change the terminal codepage to 65001 on Windows
|
2009-08-25 18:35:33 +00:00 |
|
aarne
|
8b3b2946b2
|
ancient greek ut additions
|
2009-08-21 16:13:07 +00:00 |
|
krasimir
|
a73273895d
|
the parsing test now evaluates all languages
|
2009-08-19 22:35:57 +00:00 |
|
krasimir
|
8da1aab8ff
|
add functions and functionType to the PGF API
|
2009-08-19 20:59:43 +00:00 |
|
krasimir
|
f27c1785a7
|
let the GF.Grammar.Printer printer output the unicode symbols directly
|
2009-08-18 07:09:41 +00:00 |
|
krasimir
|
3ee04cc42d
|
fix the unicode encoding problem with the cc command
|
2009-08-18 07:08:44 +00:00 |
|
aarne
|
acb930a349
|
transliteration now needs addition only in one file; a code can be more than 2 chars; ancientgreek added
|
2009-08-07 13:20:45 +00:00 |
|
krasimir
|
b180ac61a5
|
merge some changes from the latest version of Data.Binary. Makes the binary decoding faster
|
2009-08-06 11:19:04 +00:00 |
|
krasimir
|
3473f0d274
|
added testcase for pattern matching on runtime variables
|
2009-07-22 09:53:39 +00:00 |
|
krasimir
|
4f21b57a86
|
updated the output in testsuite/compiler/check/lins
|
2009-07-22 09:51:57 +00:00 |
|
krasimir
|
5abc99b291
|
added testcases for evaluation with suspension
|
2009-07-07 11:40:14 +00:00 |
|
krasimir
|
fd03f9fa0f
|
added example for abstract syntax of RDF
|
2009-07-06 19:58:27 +00:00 |
|
aarne
|
b823a64795
|
updated bibliography
|
2009-06-24 14:55:51 +00:00 |
|
krasimir
|
d965cf4755
|
simplification: decodeFile doesn't have to check that the file is empty anymore
|
2009-06-24 12:21:09 +00:00 |
|
krasimir
|
f20dd0ba4a
|
make the Functor Get instance in Data.Binary strict to be sure that the decoder is strict
|
2009-06-24 12:08:32 +00:00 |
|
krasimir
|
f81a309eb1
|
bugfix: Use openBinaryFile in Data.Binary.decodeFile. It makes difference only on Windows
|
2009-06-24 12:07:38 +00:00 |
|
krasimir
|
cf4efe0acc
|
explicitly close the file handle in Data.Binary.decodeFile
|
2009-06-24 10:39:03 +00:00 |
|
krasimir
|
9e3131f169
|
make Data.Binary.decodeFile stricter. It seems like this fixes the file locking problem on Linux
|
2009-06-23 11:56:22 +00:00 |
|
krasimir
|
8e72ff3a1c
|
added preCopy and postCopy hooks in Setup.hs
|
2009-06-23 11:44:36 +00:00 |
|
aarne
|
032f8de862
|
made .cf grammars take their startcat from the first rule
|
2009-06-23 10:13:12 +00:00 |
|
aarne
|
e89fdae2fa
|
next-lib renamed to lib, lib to old-lib
|
2009-06-22 15:39:08 +00:00 |
|
aarne
|
3049b59b35
|
cleaning up overshadow failure to warning, after testing resource compilation
|
2009-06-22 14:51:58 +00:00 |
|
aarne
|
ff29836c9b
|
some more fixes for cleaner compilation of alltensenses
|
2009-06-22 14:45:02 +00:00 |
|
aarne
|
f8fdaa7acc
|
fixed warnings in present compilation of resource, esp. unreached patterns
|
2009-06-22 14:30:55 +00:00 |
|
aarne
|
b7f6393e9f
|
test unification in TypeCheck
|
2009-06-22 06:39:25 +00:00 |
|
aarne
|
beb8cad7d8
|
the construct lin C t now replaces lock fields (in source code; still tempor used internally); lock fields removed from english resource as an example
|
2009-06-20 13:50:34 +00:00 |
|
aarne
|
48c7555975
|
accepting + patterns in pre expressions
|
2009-06-20 08:39:30 +00:00 |
|