NOTE: THIS BOUNTY IS TAKEN SINCE BLOCK 19640323
The other day, I found this bot:
https://github.com/E-D-A/steem-curatebot
by @danielsaori
Yesterday, I forked the code and converted it to do @deutschbot's job:
https://github.com/fxnormaluser/steem-deutschbot
This is what it does:
Steem deutschbot is a fork of:
https://github.com/E-D-A/steem-curatebot
Steemit curation bot with Discord integration.
The bot will login to Discord and starts to listen to the blockchain in the background.
It holds a list of 'trusted curators' for a certain hashtag on Steem. (#deutsch)
It streams all votes of the 'trusted curators' and their vote_weight and calculates a score for each post.
If a certain threshold is reached, the bot will vote for a post and broadcast its vote to Discord.
Commands:
$blacklist STEEMUSER
$curator STEEMUSER
$set_threshold 10000 (= 100% vote_weigh)
( leave argument blank to return a list )
The script works and does that job;
It's not nice code, at all:
- it crashes unexpectedly ( 25 Steem )
- it makes too many API calls ( @gtg will not like it ) ( 25 Steem )
Perhaps, parsing the stream to a local db would be a solution ... - ... the db system is awful ( 25 Steem )
- I have to initialize the
Steem()
Object at different points in the code for it to work and I don't exactly understand why. ( 25 Steem ) - Optional:
I have to useset_shared_steemd_instance
to manually override which nodes Steem-Python connects to.
This would need to be fixed in the official repository. ( 25 Steem )
If you can fix or upgrade my code, feel free to do so.
If you can fix one or more of the issues mentioned above, I will reward you in Steem;
125 Steem in total*
This bounty ends, when the payout-period of this post ends.
- Make a pull request on github
- Comment and link your PR under this post
Of course, your code has to work.
In case of multiple entries, I will try to pick the smartest solution.
However, if your solution works, you will probably win a price, anyways.
It took me a whole day to come up with the new deutschbot.py.
I figured, it would take a skilled programmer another day to clean up my mess.
Therefore, I hope 75 - 125 STEEM is enough.
*Disclaimer:
I will simultaneously work on the code, too.
If I find the best solution myself, I reserve the right to keep some of the bounty.
Furthermore, if no valid entries are made, I will keep the bounty for a later try.
Feel free to contact me here or in Discord, if you have any questions.
hi @felixxx,
The script looks like too messy to refactor. I can do a complete rewrite in 24 hours for a total 125 STEEM. What do you think about that? I am experienced w/ discord and steem api's.
Don't know enough Python but totally agree sometimes to clean up the mess takes a longer time than to redesign and rewrite the thing :-)
I agree, that it really is messy.
If you did a complete redesign, I'd like to plan ahead for some extra features.
I'll just wait for some more replies, to give others a chance - Hope that's ok.
No problem.
I have a couple of ideas:
Don't parse every block. Instead use get_account_history for each curator and filter only votes. Much more cleaner code and less overhead. This will remove almost half of the code.
Use mongodb to log already sent votes and check from there.
Use supervisord to restart the bot if bot goes down. Fixing the main reason is the real solution however you can always use Supervisord to auto-restart the process.
Custom node settings are not valid on some internal classes of steem-python. There are a couple of pull requests waiting for approval about this issue but they didn't do that yet. Using get_account_history with a custom list of nodes working well though. So, this won't be a o problem.
Let me know, if you decide for a rewrite.
my Discord id: emre#9263.
Another feature of the bot ( which now runs as a separate script ) is to drop a comment under posts, when the author has used #deutsch for the first time.
To integrate that I'd need to get all blocks.
Yes, you will need to listen then. However, I strongly suggest keeping them separate processes. Totally different things. For the sake of "seperation of concerns". They may exist in the same repository but I don't recommend same process.
For the welcome functionality:
$deutchebot --welcome
It will listen all blocks and welcome users if they use the tag first time.
For the discord/curate functionality:
$deutchebot --curate
It will not listen all blocks, will work with get_account_history.
I understand, but ...
... I wanted both functions to broadcast to discord with the same bot ID :/
ah, if the "welcome bot" does something with the discord, it's okay. :)
Hmm This sounds a pretty interesting, but the code seems to be missing something. I'm not sure what, just a feeling.
Cant help but like this idea! :) go for it !
a very good post
anyone who reads it will like it
I really like your posts
@muhammadkariman
nice
danke schön!
you are one of those people who are always grateful and patient. you deserve success in all areas.
Beautiful cute post so sweet