naiveai: if you want to switch the plugins you'll have to port the entire project :P
naiveai
Leo_Verto: yeah, I've been trying to find an alternative approach
Leo_Verto
but yeah, I definitely want to do that at some point in the future, probably after the current refactoring/throwing old stuff out spree though
naiveai
Leo_Verto: encoding is the main problem - I can't match 👍 in a regex
Leo_Verto
naiveai: again, scope creep, stick to what we've discussed for now
naiveai
Leo_Verto: i don't really think it's scope creep since you did say we should be able to handle unicode
I have a flexible enough design that I can add any arbritary character as a for/against, so I need that to work with unicode
Leo_Verto
unicode in vote name/options, yes, but you're adding another voting alias :P
I really just want to get 1.0 of this done, you can implement all the other stuff afterwards
naiveai
ok, lemme try just a little other things and then go on to tests, which shouldn't take that long
sampsyo has quit
antlarr2 is now known as antlarr
sampsyo joined the channel
KassOtsimine joined the channel
Gore|unemployed is now known as Gore|HasStoppedW
Gore|HasStoppedW is now known as Gore|unemployed
Hszzzzz has quit
UmkaDK joined the channel
UmkaDK has quit
UmkaDK joined the channel
Nyanko-sensei joined the channel
github joined the channel
github
[musicbrainz-server] samj1912 opened pull request #601: Revert "Fix SEARCH_X_ACCEL_REDIRECT to work with solr" (master...revert-600-solr) https://git.io/vNlGQ
github has left the channel
D4RK-PH0ENiX has quit
github joined the channel
[listenbrainz-server] mzfr opened pull request #336: Fixed bad URL in Docs [LB-280] (master...LB-280) https://git.io/vNlZs
github has left the channel
heyoni joined the channel
UmkaDK_ joined the channel
UmkaDK has quit
D4RK-PH0ENiX joined the channel
xps2 joined the channel
Nyanko-sensei has quit
naiveai
Leo_Verto: you remember the thing with the em-dash I talked about earlier? it works if I do it in the channel with `honcho start`, but not while running a test
presumably because the implementation of DummyApp is different in some way
Leo_Verto: for the moment I'm giving up on handling anything unicode at all
naiveai: ah yeah, botbot_shell has practically no unicode support but that's a limitation of a python 2 internal
naiveai: that test looks okay to me, although I kinda feel like our current way of testing plugins could be generally improved quite a bit
but yeah, not worse than the tests I've written :P
naiveai
yeah, to have it in some easy to write interactive mode
but for now, it looks ok
Leo_Verto
I'd like to somewhat automate testing in the future, e.g. provide the command structure and automagically test different combinations of unicode, whitespace, etc.
naiveai
Leo_Verto: also, where should I call flushall() on fakeredis? if one of my tests fails before the !endvote is called then it seems fakeredis still has the data stored
Leo_Verto
a new test function should start with a clean fakeredis
naiveai
huh
it doesn't seem to be to be the case when I'm running
yvanzo
Rotab: This is a small bug which is triggered when an edit is ninjaed in between 2am and the time subscriptions are processed.
naiveai
if I comment out the !endvote in one of my tests, the next test's !startvote complains that a vote is already running
Leo_Verto: is this a bug on the plugin side?
Freso: not sure what you mean by keeping the namespace contained in your comment on OTHER-321
naiveai: Instead of taking up "_{startvote,endvote,countvote,vote,…}" you only take up "_{vote}" in BB's command namespace.
naiveai
I don't see how that's particularly a bad thing, given that those names are extremely unlikely to correspond to anything else
Freso
"extremely unlike" ≠ impossible :)
naiveai
so should I rewrite it now?
Leo_Verto: what do you think?
Leo_Verto
I haven't talked about that because I have mixed feelings
with !vote <cmd> you wouldn't have to worry about whether it's !countvote or !countvotes, it'd obviously be !vote count
naiveai
aliases
I'll make countvote the same thing as countvotes
Leo_Verto
however from (albeit limited) experience, it seems less technical users don't "get" the !vote <cmd> syntax and would prefer a single command in some cases
although "less technical users" really isn't a problem here
how much of what you've done would you have to tear up to make the change?
essentially this would mean moving everything to a single "listens_to_command" except for the shorthands, wouldn't it?
naiveai
I'd have to change the vote command regex to something like "(count|start|end)(?P<args>.*)", then parse those args dyamically to correspond to the command, and then call the method corresponding to it.
(I think I've made it clear that I heavily, heavily dislike this idea btw)
Leo_Verto
Freso: are you okay with keeping it the way it's currently implemented? we can always change it later if it causes too much trouble
Freso
Yeah, sure.
naiveai
It's not like world-changinly hard, but it does require a fair amount of restructuring and rethinking
it'll push back this to like, today-end at best and tomorrow at worst
KassOtsimine has quit
KassOtsimine joined the channel
so, right now I'm not doing this, since I want to make some changes to the plugin anyway, we can discuss this in more depth later, when people have used it and I'll then act on the consensus
Leo_Verto
sure, focus on getting the tests and help for 1.0 done so I can accept a PR :P
phw joined the channel
phw has quit
phw joined the channel
phw is now known as outsidecontext
outsidecontext has quit
phw joined the channel
phw is now known as outsidecontext
outsidecontext has quit
outsidecontext joined the channel
naiveai: I'll be on the road for most of the evening, I'll try to check in inbetween
hey everyone, I'm submitting some listens to the listenbrainz API and somehow the additional_information.track_id and additional_information.release_id values are not taken into account (the track name and album name are not clickable in my profile)