To know if the gate should be bypassed, we check if the link contains
the pasameter instead of checking if the link ends with it. This is
impostant, for example if we were to implement searching for comments
within a post. If we wanted to search for comments within a post that we
have bypassed the gate to view: the link will look like
https://libreddit-instance/r/somesub/comments/post-id/post-title&bypass_nsfw_landing/?q=some-query&type=comment
* Redirect /:id to canonical URL for post.
This implements redirection of `/:id` (a short-form URL to a post) to
the post's canonical URL. Libreddit issues a `HEAD /:id` to Reddit to get
the canonical URL, and on success will send an HTTP 302 to a client with
the canonical URL set in as the value of the `Location:` header.
This also implements support for short IDs for non-ASCII posts, c/o
spikecodes.
Co-authored-by: spikecodes <19519553+spikecodes@users.noreply.github.com>
Implements HTTP compression, between both Reddit and Libreddit and Libreddit
and a web browser. Compression between Reddit and Libreddit is mandatory,
whereas compression between Libreddit and a client is opt-in (client must
specify a compressor in the Accept-Encoding header).
Supported compressors are gzip and brotli. gzip support is ubiquitous,
whereas brotli is supported by almost all modern browsers except Safari
(iOS, iPhone, macOS), although Safari may support brotli in the future.
Co-authored-by: Matthew E <matt@matthew.science>
Adds another on/off preference (default: on, keeps same
behaviour) for the fixed navbar.
When off the navbar will not remain at the top of the
page when scrolling.
This is useful for small displays such as phones where
otherwise the navbar takes up a sizeable portion of
the viewport.
* Parse video data from cross_post_parent_list as vanilla Reddit does.
introduce testdata directory for testing JSON parsing functions.
refactor Media::parse for slightly more readability.
Add various test cases.
* Trim down to just refactoring
Co-authored-by: Spike <19519553+spikecodes@users.noreply.github.com>
* Initial work on filtering subreddits and users
* Fix doubly-prefixed subreddit name in search alt text (e.g. r/r/pics)
* Don't set post title to "Comment" if empty - this could throw off actual posts with the title "Comment"
* Filter search results
* Fix filtering to differentiate between "this subject itself is filtered" vs "all posts on this current page have been filtered"
* Remove unnecessary check
* Clean up
* Cargo format
* Collapse comments from filtered users
Co-authored-by: spikecodes <19519553+spikecodes@users.noreply.github.com>
* Initial implementation of award parsing
* Posts: Implement awards as part of post
* Posts: remove parse_awards dead code
* Posts: initial implementation of displaying Awards at the post title
* Posts: Proxy static award images
* Client: i.redd.it should take path as argument not ID
* Posts: Just like Reddit make award size 16px
* Templates: limit the awards to 4 awards to increase performance
* Comments: Make awards a property of comments and display them
* Format and correct /img/:id
* Update comment.html
* [Optimization] Awards is not longer async
* [Revert] Posts can now display more than 4 awards again
* [Implementation] Awards not display on the frontpage
* [Implementation] Display count on awards
* Post: Start working on awards css
* Awards: Move the image size to css
* Awards: Start implementing tooltips
* Refactor awards code and tweak CSS indentation
* Unify Awards::new and Awards::parse
* Use native tooltips and brighten awards background
Co-authored-by: Spike <19519553+spikecodes@users.noreply.github.com>
