[Apertium-stuff] GSoC19 - Unsupervised weighting of automata progress update

2019-07-21 Thread Amr Mohamed Hosny Anwar
Dear Francis, Nick, Tommi, Hope this mail finds you well. I would like to share with the blog posts that I have used to document the project's progress. Firstly, The scores for the implemented methods that are computed using a custom script

Re: [Apertium-stuff] Ôdp: [Pd: Implementing apertium-html-tools]

2019-07-21 Thread Sushain Cherivirala
Yeah, I think it would work by just adding "szl": "ślōnski" to languages but I haven't checked. That variable is intended for autoglottonyms. That being said, I noted that apertium-pol-szl is not on beta.apertium.org. I checked the server and it looks like while apertium-pol-szl exists on the

Re: [Apertium-stuff] Ôdp: Ôdp: [Pd: Implementing apertium-html-tools]

2019-07-21 Thread Jonathan Washington
We support plenty of other languages that don't have two-letter codes (chv, kaa, ...), and we've never had to do it that way to my knowledge. Maybe Sushain can clarify? -- Jonathan On Sun, Jul 21, 2019, 11:35 Grzegorz Kulik wrote: > No, you define a language's name with the two-letter code: >

[Apertium-stuff] Ôdp: Ôdp: [Pd: Implementing apertium-html-tools]

2019-07-21 Thread Grzegorz Kulik
No, you define a language's name with the two-letter code: var languages = {'af': 'Afrikaans',... and then you link it to the three-letter one var iso639Codes = {'afr': 'af',... I don't know about the others but I've got a pol-szl pair, so I have to do it like that. :) Greg We niydziela, 21 lip

Re: [Apertium-stuff] Ôdp: [Pd: Implementing apertium-html-tools]

2019-07-21 Thread Jonathan Washington
Shouldn't it work without needing a two-letter code? -- Jonathan On Sun, Jul 21, 2019, 09:58 Grzegorz Kulik wrote: > Hi Sushain, > > thank you for your answer, it pointed me in the right direction. Turns out > Silesian wasn't defined at all in localization.js or in min.js. In 'var > languages'

[Apertium-stuff] Ôdp: [Pd: Implementing apertium-html-tools]

2019-07-21 Thread Grzegorz Kulik
Hi Sushain, thank you for your answer, it pointed me in the right direction. Turns out Silesian wasn't defined at all in localization.js or in min.js. In 'var languages' I created a fake two-letter code for Silesian and then in 'var iso639Codes' I linked it to the three-letter real one. Now

Re: [Apertium-stuff] CG question on unification

2019-07-21 Thread Hèctor Alòs i Font
Thank you very much, Tino! Changing $$ to && seems to solve the problem. Hèctor Missatge de Tino Didriksen del dia dg., 21 de jul. 2019 a les 12:09: > There's two kinds of unification: > https://visl.sdu.dk/cg3/chunked/sets.html#set-unification > > $$-unification must match the exact same tag.

Re: [Apertium-stuff] CG question on unification

2019-07-21 Thread Tino Didriksen
There's two kinds of unification: https://visl.sdu.dk/cg3/chunked/sets.html#set-unification $$-unification must match the exact same tag. In the first cohort, $$Gender resolves to 'm'. Then it looks for 'm' in the second cohort, and fails to find. &&-unification would work. It matches the same