Data

Tools

Indexes

Applications

Experiments

Awesome

An open API service indexing awesome lists of open source software.

https://github.com/daniguardiola/egolatron

❤️ Self-like detector for tweets!
https://github.com/daniguardiola/egolatron

chrome-extension javascript reverse-engineering twitter web-extension

Last synced: about 2 months ago
JSON representation

❤️ Self-like detector for tweets!

Awesome Lists containing this project

README 

          
# Egolatron: self-like detector for tweets!



This is a Chrome extension that detects tweets that have been liked by their own authors, and highlights them with shiny self-love medals!



[![Available on the Chrome Webstore](https://daniguardio.la/img/webstore-promo.png)](https://chrome.google.com/webstore/detail/egolatron/gncgbgdmioamlfdcliheeepnmgknlekm)



Built for fun by [Dani Guardiola](https://daniguardiola.me) ([@daniguardiolame](https://twitter.com/DaniGuardiola)).



### It stops working randomly, fix it you lazy f%#k!



It's not my fault, it's Twitter and its rate-limiting policy.



If you want to understand why this happens, read on.



### It doesn't work if I'm not logged in!



Yep. Can't do anything about it. Sorry. Again, read below to understand why this is the behavior.



## How it works



The extension will check for new tweets in the current DOM in very short intervals. When it finds new tweets, it will analyze them by retrieving the list of users who liked them and looking for the author. The API call implements pagination, but this approach works because (as far as I have observed) if the author liked the tweet, it shows up on the first page.



The API is used directly from the client context, reusing the user's (bearer) token that the official Twitter web client has obtained for its own session. This token is retrieved with a hacky but effective method that I came up with after some reverse engineering (feel free to check the code). This approach has a downside: the extension won't work if no user is logged in.



The requested resource (users who liked a tweet) seems to be part of Twitter's v2 API, but I couldn't find any public documentation, so I'm guessing it is an internal resource that only the official clients have access to.



After a tweet has been analyzed, the result is stored in memory (the tweet ID and whether it's been 'self-liked' or not). This cache-like behavior reduces the volume of calls and allows re-rendering the styles if a previously analyzed tweet comes up on screen again. The memory is reset every time the Twitter page is reloaded, but not when navigating inside the app, so it persists through a full session. Of course, this also means that it doesn't take into account any changes that happen once a tweet has been analyzed.



The styles are quite simple, just a red border on top, and a (hand-crafted) medal SVG, animated with CSS for extra-fanciness. The styles won't be applied until the tweet is on-screen (fully visible), even if it has been analyzed previously or while off-screen. The animation will start over if the tweet is re-rendered by the Twitter client, which happens by navigating, or just by scrolling, as the Twitter client dynamically loads and unloads the tweets DOM when they are scrolled into and out of view (for performance reasons).



## The Twitter rate limit



Twitter rate-limits the requests you can send in a certain period of time. If this limit is exceeded, no more requests can be made until the period resets. In my tests, the limits have been 180 requests for every 15-minute block.



To get detailed real-time information about the rate limit state (total requests, remaining requests, when the period resets and whether the limit has been reached), enable the stats with the instructions below. Any new tweets found during the rate-limited state will be ignored.



## How to look under the hood



Other than by reading the code itself, you can enable three baked-in debugging options. To do this, you need to open the devtools console, and switch to the extension context ("Egolatron" instead of "top" in the dropdown). Then, execute the following:



- For detailed logs: `_egolatron.setLog(true)`

- For detailed stats: `_egolatron.setStats(true)`

- For debug styles: `_egolatron.setDebug(true)`



To disable, pass `false` instead.



Enabling debug styles will also enable logs and stats, overriding their individual settings. The debug styles will show a blue border for tweets that are being analyzed, green for tweets confirmed to not be 'self-liked' and red for 'self-liked' tweets (this will replace the fancy styles with the animated medal).



## How to install directly (in Chromium)



- Download (and extract) or clone this repo.

- Open the extension management page by navigating to `chrome://extensions`.

  - The extension management page can also be opened by clicking on the Chrome menu, hovering over 'More tools' then selecting 'Extensions'.

- Enable developer mode by clicking the toggle switch next to 'Developer mode'.

- Click the 'Load unpacked' button and select the extension directory (`../egolatron/extension/`).



## Where to find the code



Just check the [`extension/content-script.js`](extension/content-script.js) file.



It's all there, well-organized and with many many comments. Maybe too many. There are no dependencies or build scripts.



## How to thank me for this incredible and necessary contribution to humanity



Star this repo, follow me on Twitter [@daniguardiolame](https://twitter.com/daniguardiolame) and check out my website: [daniguardiola.me](https://daniguardiola.me)



## Changelog



- 1.3: Fixed false positives and rate-limit bugs, improved medal design.

- 1.2: Disabled debug mode, which I left enabled by mistake in 1.1.

- 1.1: Fixed bearerToken grabber and improved tweet ID and username detection.

- 1.0: Initial version.