return "pick up where you left off" feature back to search about twarc

[edsu]

The problem is the "pick up" feature didn't really didn't work well once the --stream option was added. The code also got a bit complicated and tangled, and I needed to simplify it. I am sorry if this doesn't fit your use case. Please give the archive utility a try.

[skilfullycurled]

Ah I see. I didn't realize it was an architectural issue that was final. The archive.py file indicates otherwise.

"This functionality was initially part of twarc.py itself (not in a utility). If it proves useful perhaps it can go back in. But for now twarc.py writes to stdout to let you manage your data the way you want to."

I am happily using the archive.py function. That fits my use case perfectly, but my larger point was that I can't think of a use case where I would use --search given that archive.py is search plus. Again, I do not think of all things so my assertion is suspect.

Thanks, Ed!

[remagio]

Related to this issue, not to return past features, I would suggest to add a tricky feature. Probably useful for both --search --stream and archive.py: exit the process when twarc output "no new tweets…", catch python errors and do de same I found cases that it doesn't exit to the shell.
This way supervisor (or other linux system tools to running daemons) could be used. Still some trouble to catch error coming by twitter, you know.

[edsu]

I believe this problem of not exiting was fixed already. Update to the latest version, and please file a bug report if you notice a specific problem.