vas3k / infomate.club

RSS feed aggregator with collections and NLP article summarization
https://infomate.club
Apache License 2.0
449 stars 88 forks source link

Need an endpoint for Telegram -> RSS parsing #15

Closed vas3k closed 4 years ago

vas3k commented 4 years ago

I used three different services to get RSS feeds from Telegram channels. One of them got me banned, the second one returns an error 500 from time to time, the third was Chinese and it inserts hieroglyphs into the feeds :D

I think it's time to use some github library and write our own parser. It would be nice to have an endpoint that takes the Telegram channel name and returns its RSS. Don't forget about pics and videos.

If anyone want to help me with that — feel free to comment below and open a PR.

vitalii-honchar commented 4 years ago

Hello! I'm interested in this. I will work on this in this weekends.

mishfish commented 4 years ago

Maybe this can help? https://github.com/RSS-Bridge/rss-bridge

uncletoxa commented 4 years ago

can you give an example of this three project? Probably it would be beneficial to understand what was wrong.

vas3k commented 4 years ago

@WeaxMe cool, thanks. You can create another django app for that. Like parsing or something. I suppose there will be more custom parsers in future.

@mishfish interesting project to use with other sites, but it doesn't seem to support Telegram (at least as I can see)

@ToxaZ https://tg.i-c-a.su/ — banned me http://tg2rss.xyz — 500 errors https://rsshub.app/ — Chinese :)

viacheslavleonov commented 4 years ago

@vas3k Looks like rss-bridge supports Telegram - https://github.com/RSS-Bridge/rss-bridge/blob/master/bridges/TelegramBridge.php

vas3k commented 4 years ago

@viacheslavleonov ah, but it's PHP :D

Nerten commented 4 years ago

rsshub too opensource https://github.com/DIYgod/RSSHub

vshlapakov commented 4 years ago

Wow, didn't know about RSSHub, thanks for sharing.

I also searched for anything working recently, found nothing, wrote the script this weekend based on telethone and rfeed https://gist.github.com/vshlapakov/5507afb30b01b8c3a666970ddcd01d1b.

I'm using a separate Telegram account for channels, the script walks through all of them and creates RSS feed file for each, then puts it right to the nginx-served directory to be read later by any RSS feed reader (also downloads images locally if possible). It's far from ideal but works well in most cases, though doesn't support videos yet (I haven't found a way to get a public link for a telegram media attachment, so went the most straightforward way).

The funniest thing here is that I came to the idea after hanging out at infomate.club.

vas3k commented 4 years ago

Done in #16 Closing