Fix unicode support #51
Conversation
| @@ -87,11 +87,9 @@ def parse_body_part(pnum, md_text): | |||
|
|
|||
| # Store the content of the level 2 header | |||
| try: | |||
| content = codecs.unicode_escape_decode( | |||
There was a problem hiding this comment.
I think there was a reason why I had added this (it wasn't done this way initially)...let me try to figure out why this was done this way, I don't want to break other questions by making this change.
There was a problem hiding this comment.
diff mode on GitHub said it was to support LateX, but that doesn't make too much sense to me, since the tests questions that use LaTeX seem to be fine with the changes here
Edit: see this commit: 131ce85#diff-ec4c87e6a250f788996aabbdcca263c77e266b61454503016d38f2e678fdd945R161
There was a problem hiding this comment.
Thanks - so basically we are now allowing unicode characters in questions. Hopefully that fixes questions like this one: https://ca.prairielearn.com/pl/course/38/question/237974/preview
|
Fingers crossed ... hopefully this doesn't have downstream effects that will show up in Term 2 next year. |
Hopefully - I really don't see what point decoding unicode has, and it seems to have been added to target LaTeX support, if this causes an issue hopefully some of the questions added this summer will find it, since a decent variance of symbols should be used, and maybe there's some strange symbol that this will break on. |
|
So do I just 'git pull' on the branch where I was having this issue so the unicode characters are supported and appear properly now? |
|
I think I'll need to manually run the script to regenerate questions on opb000 and then it should fix. Locally, you should update |
|
Ok sounds good, updated those and now those questions are good to go, thanks |
Currently, any unicode characters get encoded into messy ascii escapes, this method of fixing it shouldn't break anything as the tests pass with no changes to them.
This was noticed in open-resources/instructor_datascience_bank#152, where some unicode box symbols were being converted to their ascii escape sequences, which do not look nice when shown to a user.
Resolves #8