Twitter @asf_releases. An automated Twitted bot aggregator. Retrieves the last releases and tweets them for you.
Uses a simple logger in the server, Markmail hacky Python API, and Tweepy. Data is stored in a local SQLite database.
The bot is configured through the aggregator.cfg file. It is a simple INI file, with field groups. The following code listing contains an example aggregator.cfg configuration file.
[markmail]
max_pages=2
[twitter]
tweet_url_length=22
Fields details.
- max_pages Maximum of pages to retrieve from Markmail
- tweet_url_length URL length to discount from tweet message length
- database SQLite database connection
The Twitter credentials are stored in a .env dotEnv file.
TWITTER_CONSUMER_KEY=
TWITTER_CONSUMER_SECRET=
TWITTER_ACCESS_KEY=
TWITTER_ACCESS_TOKEN=
This file is also in .gitignore, so that it does not get committed by accident.
The entry point script markmail_consumer.py accepts one parameter, --dry-run to allow you running the script and inspect the would-be output. This way you do not need the Twitter credentials, and can check if there is anything wrong with the data or with the script.
- reads configuration from an INI file with group fields. The file contains settings for aggregator, Markmail and Twitter
- reads the last execution time and subject used by the bot
- retrieves Markmail messages, up to a maximum of pages (step #1)
- finds messages which title matches to a REGEX, and that are more recent then the last execution time (step #2)
- posts a tweet for each message found
- updates last execution time and subject (used by step #2)
Every step includes logging that can be found in the server.
For a complete list, see requirements.txt. You can install them all with pip by running
pip install -r requirements.txt, or use conda to manually install them with
Anaconda.
There is one dependency that is not a PIP package, but a Git submodule. You can initialise it by running the following set of commands.
git submodule init
git submodule update
You can change values in the sqlite database. As in the following example, where we set the last execution date manually.
import sqlite3
conn = sqlite3.connect('database.sqlite', detect_types=sqlite3.PARSE_DECLTYPES|sqlite3.PARSE_COLNAMES)
import datetime
date = datetime.datetime(2016, 1, 1, 0, 0, 0)
c = conn.cursor()
c.execute('INSERT INTO executions(last_execution, subject, count) VALUES(?, "", 0)', (date,))
conn.commit()
conn.close()
The code is licensed under the Apache License v2. The Markmail API used by this code, it licensed under the GPL license. See LICENSE.txt for more.
The bot is hosted at TupiLabs VPS servers, and the results can be see at @asf_releases.