Can I get the error message from `pytest_runtest_protocol`?

General Discussions
1

I’m trying to capture test case details using the pytest_runtest_protocol hook in conftest.py. I can get the test name and outcome like this:

def pytest_runtest_protocol(item, log=True, nextitem=None):
    reports = runtestprotocol(item, nextitem=nextitem)
    for report in reports:
        if report.when == 'call':
            print(f'{item.name} --- {report.outcome}')
    return True

I read in the documentation that pytest_runtest_protocol can also hold exception details. Is it possible to get the actual error message or exception that caused the test to fail from this hook?

2

You can get the actual exception from the pytest_runtest_protocol hook by inspecting the longrepr attribute of the call report:

from _pytest.runner import runtestprotocol

def pytest_runtest_protocol(item, nextitem=None):
    reports = runtestprotocol(item, nextitem=nextitem)
    for report in reports:
        if report.when == "call" and report.failed:
            print(f"{item.name} failed: {report.longrepr}")
    return True

longrepr contains the full traceback and message, so you can log or process it. I used this in a plugin once to send failed test messages to Slack.

3

If you only want the exception text, you can convert longrepr to a string or, in newer pytest versions, use reprcrash.message:

from _pytest.runner import runtestprotocol

def pytest_runtest_protocol(item, nextitem=None):
    reports = runtestprotocol(item, nextitem=nextitem)
    for report in reports:
        if report.when == "call" and report.failed:
            msg = getattr(report.longrepr, "reprcrash", None)
            if msg:
                print(f"{item.name} failed: {msg.message}")
            else:
                print(f"{item.name} failed: {report.longrepr}")
    return True

This gives just the exception text without the full traceback. I found it handy for summarizing failures in a custom report.

4

I usually like logging both the outcome and the actual exception in a readable way:

from _pytest.runner import runtestprotocol

def pytest_runtest_protocol(item, nextitem=None):
    reports = runtestprotocol(item, nextitem=nextitem)
    for report in reports:
        if report.when == "call":
            if report.failed:
                print(f"Test {item.name} failed with: {report.longreprtext}")
            else:
                print(f"Test {item.name} passed")
    return True

longreprtext is basically str(report.longrepr) and is safe to print directly. I’ve used this approach in CI pipelines to capture both passes and failures cleanly.