Twarc

This release includes a bug fix for when the configuration file's directory does not exist. This functionality was accidentally removed in v2.1.5 and has been reinstated.

This release includes some new code to handle Operational Disconnect error messages from Twitter when streaming data. twarc will log the message, sleep for 5 seconds and then attempt to reconnect.

A bugfix release for logging CONSUMER_KEY correctly.

The log will now include a message about where the config file has been loaded from. In addition `--verbose` can be used to cause more information to be logged, such as the keys that are being used to access the API. Most of the time this is probably not a good idea since it makes your keys available in a log, but it can be useful in situations where debugging error responses from the API.

This release includes a small fix for `twarc timelines` to ignore empty lines. 476

This release adds two new twarc2 subcommands:

conversations

conversations will read a file of tweets and look for any conversations they are a part of and will download the full conversation thread for them.

twarc2 conversations tweets.jsonl > conversations.jsonl

You can also give it a file of conversation_ids to download instead:

twarc2 conversations ids.txt > conversations.jsonl

timelines

Similarly timelines will read in a file of tweet ids and will download the user timeline for any user who authored the tweets.

twarc2 timelines tweets.jsonl > timelines.jsonl

You can also give it a file of user ids or usernames:

twarc2 timelines users.txt > timelines.jsonl

This functionality was first developed in the twarc-timelines plugin which has been renamed to [twarc-timeline-archive](https://github.com/docnow/twarc-timeline-archive) because it does some extra things like writing timelines to separate directories and being able to be run on a schedule without redownloading previously downloaded data.