Basic flask server that serves fixed twitter video embeds to desktop discord by using either the Twitter API or Youtube-DL to grab tweet video information. This also automatically embeds the first link in the text of non video tweets (API Only)
fxtwitter exposed all recently processed tweets publicly via a "latest" and "top" page.
Even though Tweets are public, it was a personal concern for me that a tweet with potentially sensitive information in it could suddenly be shown to however many people were browsing the latest tweets page, and could be used as a tool for harassment. This was removed in [The following commit](https://github.com/dylanpdx/BetterTwitFix/commit/87ba86ba502e73ddb370bd4e5b964548d3272400#diff-a11c36d9b2d53672d6b3d781dca5bef9129159947de66bc3ffaec5fab389d80cL115)
You can also simply type out 'vx' directly before 'twitter.com' in any valid twitter video url, and that will convert it into a working vxTwitter url, for example:
I do not monitor any tweets processed by this server. Additionally, if you plan on hosting the code yourself and are concerned about this, be sure to check how to disable logging on the web server you are using (i.e Nginx)
this script uses the youtube-dl python module, along with flask, twitter and pymongo, so install those with pip (you can use `pip install -r requirements.txt`) and start the server with `python twitfix.py`
I have included some files to give you a head start on setting this server up with uWSGI, though if you decide to use uWSGI I suggest you set up mongoDB link caching
**API** - This will be where you put the credentials for your twitter API if you use this method
**database** - This is where you put the URL to your mongoDB database if you are using one
**link_cache** - (Options: **db**, **json**)
- **db**: Caches all links to a mongoDB database. This should be used it you are using uWSGI and are not just running the script on its own as one worker
- **json**: This saves cached links to a local **links.json** file
- **youtube-dl**: the original method for grabbing twitter video links, this uses a guest token provided via youtube-dl and should work well for individual instances, but may not scale up to a very large amount of usage
- **api**: this directly uses the twitter API to grab tweet info, limited to 900 calls per 15m
- **hybrid**: This will start off by using the twitter API to grab tweet info, but if the rate limit is reached or the api fails for any other reason it will switch over to youtube-dl to avoid downtime
**combination_method** - using c.vxtwitter as the url causes vxTwitter to combine all images in the post into one. This is CPU intensive, so you might not want it running on the same machine that's serving requests. When `combination_method` is set to `local`, it will use the local machine to combine the images. This requires pillow to be installed. If you want to use another server, replace `local` with the URL to the endpoint which combines images. Both methods use the code in the `combineImg` module. Inside, there's also a `Dockerfile` intended to be deployed as a combination endpoint on an (AWS Lambda function)[https://docs.aws.amazon.com/lambda/latest/dg/images-create.html].
We check for t.co links in non video tweets, and if one is found, we direct the discord useragent to embed that link directly, this means that twitter links containing youtube / vimeo links will automatically embed those as if you had just directly linked to that content