Skip to content

Handle LS experiments more reliably #126

Description

@integraledelebesgue

At the moment, when Maat launches LS, it only checks the exit code:

def _ls_has_errors(ls: StepReport) -> bool:
# Check for total errors count greater than 0.
match = re.search(r"total: (\d+) errors", ls.log_str, re.M)
return match and int(match.group(1)) > 0

This way, we only care about the process success, not the actual test. We can't verify if LS performed any actions or simply exited with status 0. We also don't know when the experiment was skipped, especially when filtering in #124.

Beyond checking the exit code, we can e.g. assert some logs in the output.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    Projects

    Status
    Backlog LS

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions