Don't depend of enabled pdf/epub to show downloads #5502
Conversation
| @@ -258,32 +256,25 @@ def get_subdomain_url(self): | |||
| def get_downloads(self, pretty=False): | |||
| project = self.project | |||
| data = {} | |||
| if pretty: | |||
There was a problem hiding this comment.
This code was duplicates, just refactored it to something simpler
|
Apart from the |
| self.slug, | ||
| ) | ||
| if project.has_htmlzip(self.slug): | ||
| data[prettify('HTML')] = project.get_production_media_url( |
There was a problem hiding this comment.
| data[prettify('HTML')] = project.get_production_media_url( | |
| data[prettify('HTMLZIP')] = project.get_production_media_url( |
I believe this should be HTMLZIP instead of HTML. The template that displays these links checks for dict.htmlzip.
There was a problem hiding this comment.
Yeah, I just realized that when prettify was set, we named this to html instead of htmlzip. I need to see if we use this in the footer api, or maybe it's safe to rename the template.
There was a problem hiding this comment.
We currently use the name HTML in the footer api and on the theme https://github.com/rtfd/sphinx_rtd_theme/blob/a49a812c8821123091166fae1897d702cdc2d627/sphinx_rtd_theme/versions.html#L17-L21
So, it's safe to rename this in the template
In readthedocs#5502 we refactored the function but we used to depend on `pretty` to change the name of the key from `html` to `htmlzip`, we use `html` in the footer and in the theme. So, it's safe to change the value in the template. See https://github.com/rtfd/readthedocs.org/pull/5502/files#r272787188
Fixes #5497
We just need to rely if the epub/pdf exists, we already do this with the localmedia download.