I've added a lazy path: type /l/languagename and you'll be redirected to /languages/lang1234. But this could use some work:
What happens if more than one language is returned? I don't think there are any cases of this yet, but it's possible. So there should be disambiguation pages. (And since the whole point is that going to the language index and clicking through is slow, the disambiguation page should play well with the tab key.)
What to do about Glottolog's country-name disambiguation parentheticals? (e.g. Mari the Uralic family vs. Mari (Papua New Guinea) the Sepik language. Once disambiguation pages exist, it would be nice to be able to ignore the parentheticals. (The point is to be lazy, after all.)
If it's not a language name, maybe it's a doculect name. Should be able to type ngadha and get Ngad'a.
We're ignoring everything that isn't an ASCII letter - is there instead a way to normalize non-ASCII characters to ASCII ones? So that you can type xaracuu and get Xârâcùù. Right now, you get Xârâcùù by typing xrc. (Is there a good way to do this in Postgres so we don't have to add a stub column?)
I've added a lazy path: type
/l/languagename
and you'll be redirected to/languages/lang1234
. But this could use some work:Mari
the Uralic family vs.Mari (Papua New Guinea)
the Sepik language. Once disambiguation pages exist, it would be nice to be able to ignore the parentheticals. (The point is to be lazy, after all.)ngadha
and get Ngad'a.xaracuu
and get Xârâcùù. Right now, you get Xârâcùù by typingxrc
. (Is there a good way to do this in Postgres so we don't have to add a stub column?)