d1cddd04aa
rewrote App so that it just inherits Translate and excludes some functions. This minimizes the duplication of code between Translate and App. Also added some functions, which make App 15% slower but increase the coverage and quality. This trade-off has to be studied further.
2014-04-09 13:04:19 +00:00
6fc7271950
adjusted the contents of Translate.gf: (1) inherit everything in Idiom since it is useful and cheap ; (2) inherit ComplVV and SlashV2V from Verb rather than Extensions, since it is more efficient and already available for all languages. Actually the previous version didn't have these functions at all, which affected the quality quite a bit.
2014-04-09 09:40:43 +00:00
edcb328b70
revised TopDict Ger,Ita,Spa with the new Dictionary improvements
2014-04-08 21:18:39 +00:00
5d4b849b14
Fix build problem in PGFService.hs when configured without c-runtime
2014-04-08 20:45:54 +00:00
2ca8024ba4
smart paradigm for refl verbs ending "se" in Spa
2014-04-08 20:28:27 +00:00
70dddec7e1
further clean-up in DictionaryIta and Spa
2014-04-08 20:21:37 +00:00
d7e25e7bdd
DictionarySpa and Ita: retrieved a few thousand unpredictable noun genders from Wiktionary
2014-04-08 19:44:18 +00:00
0341fd035f
PGF web service: add lexer support
...
The PGF web API commands parse, translate, translategroup, c-parse and
c-translate now support a lexer parameter. Possible values are "text",
"code" and "mixed".
This is used in the Wide Coverage Demo App to save one server roundtrip.
2014-04-08 16:10:33 +00:00
8c3d91f405
small fixes in App grammar - don't affect performance
2014-04-08 16:14:05 +00:00
8fc7cc5541
some words in DictionaryGer
2014-04-08 16:13:46 +00:00
0d564c15ff
smartened Paradigms in Ita and Spa to recognize some more nouns as feminine
2014-04-08 15:37:43 +00:00
72b6d8da9c
checked 600 entries in DictionaryFre ; CheckDict.hs, functions for dictionary checking
2014-04-08 15:12:34 +00:00
cf34ee8890
Move basic lexing functions from GF.Text.Lexing to the new module PGF.Lexing
...
They are thus part of the PGF Run-Time Library, making it possible to add
lexing functionality in PGF service in a natural way.
2014-04-08 14:07:49 +00:00
4f2bd4aa5f
Wide Coverage Demo App: remove quality markers ("+" or "*") before unlexing
...
Otherwise the first word won't be capitalized.
2014-04-08 13:46:57 +00:00
be62b047f4
added preliminary API for custom literals from Java. Not functional yet
2014-04-08 13:54:49 +00:00
2198d18d3a
TopDictFre checked to 500
2014-04-08 12:53:03 +00:00
52377b46cf
bugfix in the C runtime that shows up with only some grammars
2014-04-08 12:41:41 +00:00
581a0336a5
Wide Coverage Demo App: restore swap button
2014-04-08 11:32:54 +00:00
feb4ff5dd1
plural of stad in Dut
2014-04-08 09:19:50 +00:00
04f9a50da0
some new ParadigmsFre and DictionarySwe
2014-04-07 20:23:11 +00:00
02a0372b41
checked 300+ words in TopDictFre ; not yet merged in DictionaryFre
2014-04-07 20:22:42 +00:00
9f3e9d8626
demos/translation.t2t: more space for the demo app
2014-04-07 13:01:00 +00:00
620f30d187
Wide Coverage Demo Web App: add segmentation
...
Click on sentences to choose among alternative translations.
2014-04-07 12:51:13 +00:00
baef6eee90
Expr and ExprProb are now serializable objects
2014-04-07 12:52:51 +00:00
e4555d2dce
fix in the keyboard
2014-04-07 12:42:02 +00:00
dc476cce50
added alternative_item.xml
2014-04-07 12:41:23 +00:00
2825f9e420
VPI chunks linearized ; nouns with ión in Spanish and zione in Italian marked as feminine
2014-04-07 12:02:52 +00:00
1671383e1c
another fix in DictionaryBul
2014-04-07 09:53:31 +00:00
024321b520
fixes in DictionaryBul
2014-04-07 09:52:12 +00:00
78e7f0e87d
a bit more text in the help
2014-04-07 09:43:20 +00:00
e8303dc761
show alternative translations on Android
2014-04-07 09:14:14 +00:00
0df4d4bef6
restored passives in Translate, generalized IdRP in Eng
2014-04-07 07:57:55 +00:00
72c4527767
Wide coverage demo app: show an error message even if the error string is empty
2014-04-06 23:44:31 +00:00
ceb6de1561
Wide coverage demo app: show an error message if linearization fails
2014-04-06 23:13:18 +00:00
b7973f2f5d
restored the initial segment of TopDictSwe
2014-04-06 20:19:46 +00:00
0721769935
gftranslate.js: use Translate10.pgf
2014-04-06 19:59:04 +00:00
79fe3f2b49
swede in DictionaryEng ; CompoundCN probability in translate.probs tweaked to avoid too aggressive compounding
2014-04-06 19:45:35 +00:00
37c3afa9b4
added "todo" dictionaries
2014-04-06 19:19:51 +00:00
82a333c602
normalized Dictionary Fin,Chin,Hin,Fre to a format easier to process automatically; other Dictionary files were already in this format: each rule prefixed by "lin", sorted, checked parts uncommented, unchecked or problematic parts commented, one rule per line
2014-04-06 16:26:16 +00:00
5d36c4734d
Fixes for compiling Translate10.pgf
2014-04-04 19:04:16 +00:00
3fa7b3e04b
149 new words in DictinarySwe.gf (mostly geographical names)
2014-04-04 19:02:42 +00:00
0aff5f4aa4
type error in finnish revealed
2014-04-04 17:01:06 +00:00
0577ec19a4
fixed type errors in finnish revealed by improved type checker
2014-04-04 16:38:36 +00:00
0ae79efdbf
TypeCheck.RConcrete: contravariance in table types
2014-04-04 16:50:12 +00:00
c4a45f687f
translate10 do the right thing
2014-04-04 14:45:50 +00:00
3f24bd302a
path in AppFin
2014-04-04 14:32:36 +00:00
81f76ba658
Make for Translate10
2014-04-04 14:32:05 +00:00
316e473a1e
added Spa and Ita to translator/ ; omitted some Extensions functions to double the parsing speed
2014-04-04 14:13:11 +00:00
ab3244fbe5
polarities restored in Hin translation
2014-04-04 12:05:20 +00:00
15f0edae32
Removed imports of TypeCheck.Concrete, fixed problems in TypeCheck.RConcrete
...
+ The current type checker for concrete syntax is in
GF.Compile.TypeCheck.RConcrete, but GF.Compile.TypeCheck.Concrete was
still imported in GFI.
+ Fixed a bug that allowed Ints n as a subtype of Ints m, regardless of
m and n. It now requires n<=m. Note: the type checker still allows Int
as a subtype of Ints m, regardless of m.
+ Fixed a potential efficiency problem with large record types, by reducing
the number of recursive calls from |R|*|S| to |R| when checking if R<=S.
+ Fixed a misleading comment: "alpha g t u" checks that u is a subtype of t,
the other way around. Similarly, "checkIfEqLType gr g t u trm" checks that
u is a subtype of t, not the other way around, and not that t is equal to u.
2014-04-04 13:51:07 +00:00
aarne