fix: wait_for_logs can now fail early when the container stops

[ArthurBook]

Addresses my suggestion made in issue 681.

This PR adds a flag that checks is the status is not running and raises a RuntimeError to avoid waiting for logs after the container already has exited. The idea is to save wait time when there is a long startup time in case the container fails early.

from testcontainers.core import container, waiting_utils

if __name__ == "__main__":
    waiting_utils.wait_for_logs(
        container.DockerContainer("flyway/flyway").start(),
        r"Successfully applied \d+ migrations to schema",
        timeout=10,
        raise_on_exit=True,
    )

# > RuntimeError(f"Container exited before emitting logs satisfying predicate")
# ( Raised almost immediately ) 

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.

Please upload report for BASE (main@c7d9b81). Learn more about missing BASE report.

Files	Patch %	Lines
core/testcontainers/core/waiting_utils.py	0.00%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #682   +/-   ##
=======================================
  Coverage        ?   78.07%           
=======================================
  Files           ?       12           
  Lines           ?      602           
  Branches        ?       89           
=======================================
  Hits            ?      470           
  Misses          ?      106           
  Partials        ?       26           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[alexanderankin]

why is true is not the default, any background on this?

[ArthurBook]

why is true is not the default, any background on this?

I think it would be a good default.
Kept it False just to not change existing functionality without explicit opt-in.

[alexanderankin]

i would like to roll it out everywhere else, wonder what would be a good strategy for that, unless i am overthinking it, in which case we should just do it.

[kiview]

TBH, I think the false behavior is more a bug than actual expected behavior. If I were to make the call, I would not only make true the default, but also don't make it user configurable at all (can any of you come up with a use case were the false behavior is useful?).

[alexanderankin]

pr title = "fix: wait_for_logs can now fail early when the container stops"

next pr title = ${pr title/can/WILL/}

[ArthurBook]

Hi!
Agree with above, for all our use cases, the expected behavior would be to raise RuntimeError if the container exited. I think its a good idea to drop the raise_on_exit param and fix behavior to True.

Another option is to allow user to define the set of Container.status that (don't) cause a fail. This would allow for a bit more configurability. WDYT?