[00:36:27] Feuerbach: for your second question, quick answer: no :P you can't match like that [00:36:36] they aren't data constructors [00:37:01] they're just functions that take arguments and construct an AP [00:37:27] you could define your lincat for AP such that you have a field isParticiple : Bool [00:38:00] and then when you build your ANP, you check that field of the AP and place it accordingly [01:17:40] *** Quits: Feuerbach (~feuerbach@unaffiliated/feuerbach) (Ping timeout: 252 seconds) [01:35:28] *** Quits: evariste (~evariste@2001:4b98:dc0:41:216:3eff:fef8:98d6) (Ping timeout: 245 seconds) [01:38:17] *** Joins: evariste (~evariste@2001:4b98:dc0:41:216:3eff:fef8:98d6) [01:43:43] *** Joins: inariksi1 (~inari@ksit.fixme.fi) [01:44:37] *** Quits: jonsterling (sid24013@gateway/web/irccloud.com/x-xiutcjstpxkdhzpc) (Ping timeout: 240 seconds) [01:44:41] *** Quits: inariksit (~inari@ksit.fixme.fi) (Ping timeout: 240 seconds) [01:50:28] *** Quits: esg_ (~emil@193.1.218.163) (Ping timeout: 240 seconds) [01:50:37] *** Joins: esg (~emil@193.1.218.163) [01:51:44] *** Joins: jonsterling (sid24013@gateway/web/irccloud.com/x-zkpmfbvfwmtemign) [03:27:35] Aargh. Don't want to import all tenses just to say, "was born in". [07:42:30] *** Joins: Feuerbach (~feuerbach@unaffiliated/feuerbach) [09:04:20] drbean: hmm, that could be added to Constructions.gf maybe [09:08:22] *** inariksi1 is now known as inariksit [09:27:09] *** Quits: Feuerbach (~feuerbach@unaffiliated/feuerbach) (Ping timeout: 252 seconds) [09:33:27] *** Joins: Feuerbach (~feuerbach@unaffiliated/feuerbach) [09:38:24] drbean: do you know of the module Construction? it's pretty recent addition, and quite small yet. I was just thinking if you need to express times, "I was born in May/Wednesday/1987", here you have constructions for habitual and punctual time expressions [09:38:28] http://www.grammaticalframework.org/lib/doc/browse/#english/ConstructionEng.gf [09:39:26] for chinese, english, finnish, french, german and swedish apparently [09:39:51] these get implemented based on what languages aarne/me/krasimir/whoever who has commit rights to the repo knows :P [10:02:33] *** Joins: myoriel (~Myoriel@aftr-88-217-180-51.dynamic.mnet-online.de) [10:50:02] *** Quits: Feuerbach (~feuerbach@unaffiliated/feuerbach) (Ping timeout: 245 seconds) [10:57:16] *** Joins: Feuerbach (~feuerbach@unaffiliated/feuerbach) [11:44:10] *** Joins: jmvanel (~jmvanel@78.193.21.40) [11:52:36] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 240 seconds) [12:08:50] *** Quits: Feuerbach (~feuerbach@unaffiliated/feuerbach) (Ping timeout: 252 seconds) [12:10:33] *** Joins: Feuerbach (~feuerbach@80.90.232.50) [12:10:33] *** Quits: Feuerbach (~feuerbach@80.90.232.50) (Changing host) [12:10:33] *** Joins: Feuerbach (~feuerbach@unaffiliated/feuerbach) [12:17:03] *** Quits: jmvanel (~jmvanel@78.193.21.40) (Ping timeout: 264 seconds) [12:31:56] *** Quits: Feuerbach (~feuerbach@unaffiliated/feuerbach) (Ping timeout: 252 seconds) [12:44:30] *** Joins: spectie (~fran@unaffiliated/spectie) [13:16:13] haha spectie, we're such apertium fans here :-D had lunch with GF group and apertium was mentioned quite often [13:16:49] actually aarne talks pretty often something like "I saw in the apertium mailing list that [...]", and I might say something "yeah I saw in the IRC channel" [13:17:55] and they are implementing some new chunking-based thing, which aarne refers to as "the apertium thing" or something like that [13:20:41] inariksit, hey! [13:20:45] \o/ [13:20:48] LOL [13:20:51] "the apertium thing" [13:20:51] :D [13:25:56] it's in the good sense :D aarne has been saying he's impressed by the quality of the translation of apertium [13:26:08] that's very generous ;) [13:26:16] i am always complaining about it :) [13:26:32] hehe, maybe it's just a case of grass being greener on the other side [13:30:40] aye [13:36:03] this is probably super bad publicity for GF, but I ran the apertium story (english) through our wide coverage translator, the results are not so impressive :D http://www.grammaticalframework.org/~inari/apertium-story/ [13:37:58] well, everything gets parsed (didn't try parsing other language versions, just english), but the lexical coverage is quite bad... :P [13:38:01] "quand [mary_PN] [GerundNP] [GerundNP] finit [GerundNP] elle [look_V] [around_Adv]" [13:38:47] :/ [13:38:57] yeah, the vocab coverage is quite bad :/ [13:39:03] how about if you add the vocab ? [13:39:09] it's only about 200 words [13:39:09] *** Joins: Feuerbach (~feuerbach@80.90.232.50) [13:39:09] *** Quits: Feuerbach (~feuerbach@80.90.232.50) (Changing host) [13:39:09] *** Joins: Feuerbach (~feuerbach@unaffiliated/feuerbach) [13:39:45] yeah, that would be a good addition, this story has very frequent vocabulary [13:40:01] I did some changes on finnish already based on that translation [13:41:57] aarne is more focused on getting wide coverage translation to work in GF, I sent him these too and he seemed interested, maybe he's doing something on his own already [13:42:06] et que est [james_PN] [do_V] proche le [13:42:14] this is transfer [13:42:35] qu'est ce que ... [13:42:44] do you have that structure in the french RG? [13:43:35] qu'est ce que? yes we do [13:43:51] originally RG wasn't really meant for translation, just a resource [13:44:24] now the trend has been to take RG as a base and extend it with domain-specific rules, and give higher probabilities to those rules [13:45:17] but yeah, in this case the english is just parsed to something, possibly in chunks, and it is linearized to french [13:45:45] ah [13:45:55] the rule for qu'est-ce que etc. is probably in ExtraFre or something, and it isn't going to be a translation for this english thing [13:45:59] or Iäd need to check [13:46:07] :) [13:46:08] (whoops now I need to see my supervisor actually :P) [13:46:18] :D [13:56:23] why, when I try to parse a simple Russian word, it gives me a huge number of parses? and how to limit them? [13:56:46] e.g. i present/LangRus.gfo; p "мальчик" [13:57:10] prints many lines like PhrUtt NoPConj (UttCN (SentCN (UseN boy_N) (EmbedQS (UseQCl ?4 ?5 (QuestCl (ImpersCl (ComplSlash (VPSlashPrep (UseComp (CompNP (AdvNP (DetNP (DetQuant DefArt NumSg)) (PrepNP possess_Prep (DetNP (DetQuant DefArt NumSg)))))) part_Prep) (AdvNP (DetNP (DetQuant DefArt NumPl)) (PrepNP part_Prep (DetNP (DetQuant IndefArt NumPl))))))))))) NoVoc [13:57:28] I tried -depth=1, but it doesn't help [14:01:18] whoa that's weird, my guess is that in the russian grammar there are rules that are linearized as empty strings, and the parser happily recognizes them in nothingness [14:36:41] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 245 seconds) [14:38:35] *** Joins: spectie (~fran@unaffiliated/spectie) [14:53:17] *** Quits: Feuerbach (~feuerbach@unaffiliated/feuerbach) (Ping timeout: 252 seconds) [14:53:30] *** Joins: jmvanel (~jmvanel@78.193.21.40) [14:54:13] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 246 seconds) [14:56:37] *** Joins: spectie (~fran@unaffiliated/spectie) [15:01:34] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 246 seconds) [15:02:01] *** Joins: spectie (~fran@unaffiliated/spectie) [15:11:24] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 264 seconds) [15:13:50] *** Joins: Feuerbach (~feuerbach@80.90.232.50) [15:13:50] *** Quits: Feuerbach (~feuerbach@80.90.232.50) (Changing host) [15:13:50] *** Joins: Feuerbach (~feuerbach@unaffiliated/feuerbach) [15:48:34] *** Joins: spectie (~fran@hum-tf4-ans166.hum.uit.no) [15:48:34] *** Quits: spectie (~fran@hum-tf4-ans166.hum.uit.no) (Changing host) [15:48:34] *** Joins: spectie (~fran@unaffiliated/spectie) [15:51:37] Feuerbach: yep, it's that reason [15:51:45] so, parsing this as a NP gives the first results [15:51:45] Lang> p -cat=NP "мальчик" [15:51:46] DetCN (DetQuant DefArt NumSg) (UseN boy_N) [15:51:46] DetCN (DetQuant IndefArt NumSg) (UseN boy_N) [15:51:46] MassNP (UseN boy_N) [15:52:10] then it goes on to all kinds of silly parses like AdvNP (DetNP (DetQuant DefArt NumSg)) (PrepNP part_Prep (AdvNP (DetCN (DetQuant DefArt NumSg) (UseN boy_N)) (PrepNP part_Prep (DetNP (DetQuant DefArt NumSg))))) [15:52:35] looking what part_Prep does, it's linearized as an empty string [15:52:36] StructuralRus.gf: part_Prep = { s = "" ; c = Nom}; [15:55:37] inariksit: I see. thanks! [15:57:00] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 264 seconds) [16:00:49] *** Joins: spectie (~fran@unaffiliated/spectie) [16:33:54] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 240 seconds) [16:34:51] *** Quits: jmvanel (~jmvanel@78.193.21.40) (Quit: Quitte) [16:35:20] *** Joins: spectie (~fran@unaffiliated/spectie) [16:57:17] *** Joins: jmvanel (~jmvanel@78.193.21.40) [16:57:54] *** Quits: Feuerbach (~feuerbach@unaffiliated/feuerbach) (Ping timeout: 240 seconds) [16:58:07] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 252 seconds) [17:06:33] *** Joins: spectie (~fran@hum-tf4-ans189.hum.uit.no) [17:06:33] *** Quits: spectie (~fran@hum-tf4-ans189.hum.uit.no) (Changing host) [17:06:33] *** Joins: spectie (~fran@unaffiliated/spectie) [17:09:13] *** Quits: spectie (~fran@unaffiliated/spectie) (Client Quit) [17:09:26] *** Joins: spectie (~fran@unaffiliated/spectie) [17:11:53] *** Joins: spectre (~fran@hum-tf4-ans166.hum.uit.no) [17:15:43] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 252 seconds) [18:17:54] *** Quits: spectre (~fran@hum-tf4-ans166.hum.uit.no) (Ping timeout: 240 seconds) [18:31:13] *** Quits: jmvanel (~jmvanel@78.193.21.40) (Ping timeout: 246 seconds) [18:45:26] http://www.grammaticalframework.org/~inari/pgf-library.html I promised to write some kind of tutorial on using GF from haskell/other language [18:45:47] does anyone have good ideas what to include? :-D this is obviously a super beta version [19:11:00] *** Joins: spectie (~fran@unaffiliated/spectie) [20:46:15] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 240 seconds) [20:46:41] *** Joins: spectie (~fran@unaffiliated/spectie) [20:54:37] *** Quits: myoriel (~Myoriel@aftr-88-217-180-51.dynamic.mnet-online.de) (Ping timeout: 252 seconds) [21:00:52] *** Joins: jmvanel (~jmvanel@bdv75-13-78-237-141-165.fbx.proxad.net) [21:51:15] *** Quits: jmvanel (~jmvanel@bdv75-13-78-237-141-165.fbx.proxad.net) (Ping timeout: 241 seconds) [22:51:06] *** Joins: spectre (~fran@c78CF00C3.dhcp.as2116.net) [23:07:30] *** Quits: spectre (~fran@c78CF00C3.dhcp.as2116.net) (Remote host closed the connection) [23:35:53] *** Quits: spectie (~fran@unaffiliated/spectie) (Ping timeout: 240 seconds) [23:38:05] *** Joins: spectie (~fran@c10C800C3.dhcp.as2116.net) [23:38:05] *** Quits: spectie (~fran@c10C800C3.dhcp.as2116.net) (Changing host) [23:38:05] *** Joins: spectie (~fran@unaffiliated/spectie) [23:40:16] *** Quits: spectie (~fran@unaffiliated/spectie) (Excess Flood) [23:40:37] *** Joins: spectie (~fran@unaffiliated/spectie) [23:44:14] *** Joins: spectei (~fran@c78CF00C3.dhcp.as2116.net) [23:44:14] *** Quits: spectei (~fran@c78CF00C3.dhcp.as2116.net) (Changing host) [23:44:14] *** Joins: spectei (~fran@unaffiliated/spectie) [23:48:37] *** Joins: Feuerbach (~feuerbach@unaffiliated/feuerbach)