"Go to Pouet, but not to the BBS"
category: offtopic [glöplog]
Ok, sent a pull request, I'm new to this so... sorry if I fucked up :(
Whoops, didn't see your pull request Tomoya - but thanks! :)
Narrowed it a bit down, so it's not running everywhere - only where needed. Not sure if these needs to be fixed for the UserScript as well, but ppl will complain if that's the case :3
tl;dr [X] fixed
Narrowed it a bit down, so it's not running everywhere - only where needed. Not sure if these needs to be fixed for the UserScript as well, but ppl will complain if that's the case :3
tl;dr [X] fixed
Thanks for the pull request.
A good idea with some good ideas in its implementation ( fixing the split htmlentities ). I'll check another approach than the jungle hunt and all the hoops in the DOM to find the HTML entities. :p
A good idea with some good ideas in its implementation ( fixing the split htmlentities ). I'll check another approach than the jungle hunt and all the hoops in the DOM to find the HTML entities. :p
Well, so insane it sounds, but it was the most sane approach to make sure it only targets the oneliner on both pages. That parentNode thing still makes me cringe though :P
unicode.user.js uses XPATH to find all the elements whose text contain a & character, then it replaces the & by a single & and look for the Թ sequences ( withstanding linebreaks ) and replace them by the corresponding character. This works everywhere: onliner, comments, prod title, group names, etc...
How do you like this approach ?
How do you like this approach ?
I don't see how this is more elegant than to tell gargaj to just pass the onliner to html_entity_decode() ;)
dfox:
he won't be able to do that until pouet v2, so guys... just wait for it.
Code:
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">he won't be able to do that until pouet v2, so guys... just wait for it.
Saga: Your point being?
Just saved the page as HTML, replaced the & to & and voila, it displays correctly, even with the "wrong" encoding in the header.
It might be a different story for the complete platform - but this seems to be working just fine.
well i guess i just misunderstood your idea - i thought you want to double-decode the double-encoded entities (which would result in plain unicode characters in an ISO-encoded file) - however I guess it would be the easiest (TM) to fix the DB storage rather than its output, which is something that gargaj is obviously going to do.
p01: As it seems to work everywhere else, even in the BBS I didn't find it necessary, apart from some odd encoded names. Other than that, I didn't really wanted to fuck with the site, and introduce loopholes - or odd stuff he later has to fix. Kudos though :)