Commit Graph

6976 Commits

Author SHA1 Message Date
Inari Listenmaa
2c3372a5fd Merge pull request #167 from odanoburu/por-genrp
(Por) add GenRP to ExtendPor
2019-02-12 15:02:31 +01:00
Inari Listenmaa
2ac36ba8e4 Merge branch 'master' of https://github.com/GrammaticalFramework/gf-rgl into persian 2019-02-11 14:31:46 +01:00
Inari Listenmaa
a532389bbd (Pes) Fix orthography + add a few words in LexiconPes 2019-02-11 11:30:05 +01:00
Inari Listenmaa
9ea34d871c (Pes) More comments in ParadigmsPes 2019-02-11 11:29:45 +01:00
odanoburu
db671b2e31 (Por) add GenRP to ExtendPor 2019-02-10 10:22:12 -02:00
Inari Listenmaa
57b59ff386 Merge pull request #166 from odanoburu/iberic-clause-inversion
(Por,Spa) don't invert word order in questions most of the time
2019-02-10 08:47:34 +01:00
Inari Listenmaa
56328d56f1 Merge pull request #164 from odanoburu/usecomp_estar
(Extend,Por,Spa) add UseComp_estar
2019-02-10 08:41:49 +01:00
odanoburu
de301c45cd (Por,Spa) don't invert word order in questions most of the time
- don't invert in QuestIAdv either
(Romance) allow for finer-grained choice of when to invert word order
in questions
2019-02-09 12:54:54 -02:00
Inari Listenmaa
f4daeb4f54 Merge pull request #165 from inariksit/persian
Persian
2019-02-09 15:36:33 +01:00
Inari Listenmaa
5fa83b90e4 (Pes) Add possibility to use vowel diacritic (kasre) in mkEzafa
It's still not using kasre, but if you want to use it, just change one 
line.
2019-02-09 15:28:20 +01:00
Inari Listenmaa
b0ea6680fb (Pes) comment out conjThat from insertVV; TODO where it should go? 2019-02-09 15:27:06 +01:00
Inari Listenmaa
812602e6f8 (Pes) Cleanup; add comments in ParadigmsPes
Hide some confusing constructors from the public API, add mkA2 and a 
pattern matching mkV
2019-02-09 15:26:12 +01:00
odanoburu
95eb384223 (Extend,Por,Spa) add UseComp_estar
- with default UseComp lin
2019-02-09 09:37:01 -02:00
Inari Listenmaa
ceb876eb37 Merge pull request #163 from inariksit/persian-fixes
Persian fixes
2019-02-07 20:22:14 +01:00
Inari Listenmaa
55decd9317 (Pes) Noun is indefinite if predicative and modified by adjective
According to this source: 
https://sites.la.utexas.edu/persian_online_resources/nouns/noun-in-a-predicative-position/
2019-02-07 19:49:06 +01:00
Inari Listenmaa
a45cc462e3 (Pes) Fixes in VV constructions
* Add the conjunction که 'that' in VV
* Fix word order in imperatives formed of VV+complement
2019-02-07 19:37:17 +01:00
Inari Listenmaa
61838b818e (Pes) Fix orthography and add a vowel diacritic in mkEzafa and mkEnclic 2019-02-07 19:34:21 +01:00
Inari Listenmaa
04c5f433ec Merge pull request #161 from inariksit/persian
Persian
2019-02-07 11:39:51 +01:00
Inari Listenmaa
79a375a03e (Pes) add beVerb and use in Lexicon for some compound verbs 2019-02-07 11:10:47 +01:00
Inari Listenmaa
d0d760d68b (Pes) Remove hardcoded have_V2 for one constructed in MorphoPes 2019-02-07 11:10:04 +01:00
Inari Listenmaa
8a6b0f8f01 (Pes) Added ZWNJ to verb forms + big cleanup and refactoring 2019-02-07 11:00:26 +01:00
Inari Listenmaa
75cf3643ec (Pes) Remove unused MakeStructuralPes 2019-02-06 20:12:42 +01:00
Inari Listenmaa
b0c9a3527b Merge pull request #160 from odanoburu/iberic-neg-imp
(Por) also bind clitic with hyphen in imperative
2019-02-06 18:40:21 +01:00
odanoburu
b976158388 (Por) also bind clitic with hyphen in imperative 2019-02-06 13:23:23 -02:00
Inari Listenmaa
1da12df1db Merge pull request #159 from odanoburu/iberic-neg-imp
(Cat,Por,Spa) Iberic imperative
2019-02-06 15:23:34 +01:00
odanoburu
32bde9dd1b (Cat) improve Catalan imperative 2019-02-06 11:19:55 -02:00
Inari Listenmaa
28917900ba Merge pull request #158 from odanoburu/ud-treebank
(UD treebank) fix two trees using it_Pron instead of ImpersCl
2019-02-05 19:21:06 +01:00
odanoburu
8939dee214 (Por) update lexicon given past changes in morphological paradigms 2019-02-05 13:59:04 -02:00
odanoburu
2679717852 (Por,Spa) fix negative imperative
- reflexive pronoun is before the verb in the negative, not after
- verb form is not the same as positive imperative, but the same as
  present subjunctive

thanks to @inariksit for catching this bug!

(Por) handle politeness better
formal -> 2nd person singular (tu)
informal sg -> 3rd person singular (você, conjugated correctly)
informal pl -> 3rd person plural (vocês, conjugated correctly)

unlike Spa, Por uses 2nd person as polite form; the PorRG thus would
have no direct access to vós (second person plural), which is rather
rare (in Brazilian Por, at least). colloquial speech in Brazil
actually mangles both, using tu with você conjugation, etc; but this
out of scope for the RGL.

(Por) fix imperative
- should use hyphen to bind the reflexive pronoun
2019-02-05 13:58:15 -02:00
odanoburu
28a9453750 (UD treebank) fix two trees using it_Pron instead of ImpersCl 2019-02-05 09:47:10 -02:00
Inari Listenmaa
f2cba3fb4d Merge pull request #157 from inariksit/persian
Persian
2019-02-04 16:59:09 +01:00
Inari Listenmaa
7cb2298b55 (Pes) Add zero-width non-joiner + mkN in ParadigmsPes 2019-02-04 10:44:44 +01:00
Inari Listenmaa
19d4934671 (Pes) Add isNum to SymbOrd 2019-02-04 10:07:48 +01:00
Inari Listenmaa
271feb0295 Merge pull request #156 from odanoburu/por-dict-update-2
(Por) update Dict
2019-02-01 15:07:07 +01:00
odanoburu
a3a0a3958b (Por) update Dict
-- opening new PR to save space (dictionary diffs are big!)

- don't add gender information to noun fun names if uneeded
- update README with this info
- update gfdoc sections in MorphoPor
2019-02-01 10:54:00 -02:00
Inari Listenmaa
494acc4153 (Pes) Add ExtendPes, so far only GenNP implemented 2019-01-31 17:05:16 +01:00
Inari Listenmaa
8a69cd40c7 (Pes) Cleanup + automatic whitespace removal 2019-01-31 17:04:47 +01:00
Inari Listenmaa
781f049fdd (Pes) Add isNum to Num + Ord; fix bug in DetCN, DetQuant and DetQuantOrd
See https://sites.la.utexas.edu/persian_online_resources/nouns/

"A noun modified by a number is never pluralized; that is to say, it is 
invariably singular"

In the old grammar, DetQuant would always make the Det into numeral, and 
so Sg was always chosen in DetCN.
2019-01-31 16:38:46 +01:00
Inari Listenmaa
7a48a09931 (Pes) Use enclitic form of NP in predicative
According to this source: 
https://sites.la.utexas.edu/persian_online_resources/nouns/noun-in-a-predicative-position/
2019-01-31 16:29:40 +01:00
Inari Listenmaa
edc41a3a5c Merge pull request #154 from inariksit/arabic
(Ara) Use PerGenNum in the table for RCl and RS; compiles 15sec faster
2019-01-30 17:31:03 +01:00
Inari Listenmaa
80d7e293ee (Ara) Fix bug in PossPron 2019-01-30 17:18:53 +01:00
Inari Listenmaa
bf0fd56960 (Ara) Use PerGenNum in the table for RCl and RS; compiles 15sec faster 2019-01-30 16:23:11 +01:00
Inari Listenmaa
269191cf9a Merge pull request #153 from inariksit/arabic
Arabic
2019-01-29 20:30:25 +01:00
Inari Listenmaa
7cdcc0931f (Ara) Small typofix + add comment 2019-01-29 17:57:38 +01:00
Inari Listenmaa
e468b48ded (Ara) Fix the corner case where ReflVP is called after VPSlashPrep.
Also fix noPrep to accPrep (the default object case).

Tested with gftest -o, only changes were in combinations of ReflVP and 
VPSlashPrep.
2019-01-29 17:57:21 +01:00
Inari Listenmaa
ac5352f852 (Ara) Replace two 2-valued params in VP with one 3-valued param.
* is copula / is have_V2 / is neither

There is no difference; all three options are is mutually exclusive. 
There 2 params in the first place only by accident. First one was added 
by AED 10 years agon, second by me only some weeks ago, without thinking 
that the params are mutually exclusive.

Also tested with gftest, no difference found after 2 hours of running.
2019-01-29 17:54:44 +01:00
Inari Listenmaa
08c0b3fe54 (Ara) Fixes in DetNP + some Dets and Quants 2019-01-29 15:46:59 +01:00
Inari Listenmaa
267ef1b178 (Ara) Move mkQuant functions from Paradigms to Res 2019-01-29 14:11:05 +01:00
Aarne Ranta
5a50706345 changed suffix in SymbOrd to :e which is more general, but not perfect of course 2019-01-29 11:31:11 +01:00
Herbert Lange
c30a33b2cb recreating missing 2019-01-28 18:25:09 +01:00