CI: Use azure-pipelines reporting for pytest data

[datapythonista]

Azure Pipelines can show errors and warnings clearer, by adding some specific syntax to the log.

We are already using this for the linting, where we output flake8 errors like ##vso[task.logissue type=error;sourcepath=%(path)s;linenumber=%(row)s;columnnumber=%(col)s;code=%(code)s;]%(text)s (see ci/code_checks.sh).

This makes it easier to find the errors in the logs, as they appear in the summary. See for example the Checks_and_docs build in https://dev.azure.com/pandas-dev/pandas/_build/results?buildId=12078

We should be using the same for tests, and also "upload" to Azure pipelines the summary of the tests.

This is implemented in this pytest plugin: https://github.com/tonybaloney/pytest-azurepipelines

I guess the best option is to install the plugin (may be install in the azure yaml files, as the dependency is not useful outside of azure). The plugin is quite simple, if it makes sense we may prefer to implement something specific to our needs based on the plugin code: https://github.com/tonybaloney/pytest-azurepipelines/blob/master/pytest_azurepipelines.py

[TomAugspurger]

I use this for dask-ml and it seems to work well.

…

On Jun 1, 2019, at 05:47, Marc Garcia ***@***.***> wrote: Azure Pipelines can show errors and warnings clearer, by adding some specific syntax to the log. We are already using this for the linting, where we output flake8 errors like ##vso[task.logissue type=error;sourcepath=%(path)s;linenumber=%(row)s;columnnumber=%(col)s;code=%(code)s;]%(text)s (see ci/code_checks.sh). This makes it easier to find the errors in the logs, as they appear in the summary. See for example the Checks_and_docs build in https://dev.azure.com/pandas-dev/pandas/_build/results?buildId=12078 We should be using the same for tests, and also "upload" to Azure pipelines the summary of the tests. This is implemented in this pytest plugin: https://github.com/tonybaloney/pytest-azurepipelines I guess the best option is to install the plugin (may be install in the azure yaml files, as the dependency is not useful outside of azure). The plugin is quite simple, if it makes sense we may prefer to implement something specific to our needs based on the plugin code: https://github.com/tonybaloney/pytest-azurepipelines/blob/master/pytest_azurepipelines.py — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.