Improve exceptions emitted and handled

[Jacob-Stevens-Haas]

No more naked exceptions, except in one place where it truly is intended (i.e. if something bad happens, warn about it and continue the loop). Exception testing rewritten in pytest, since that's where we're headed and I don't know how to modify robot tests.

• Changes _printErrorMessage to _createErrorMessage. Exception handling determines whether to print error message
• moved MaxRetriesException to _DataProductFile.py, its only user
• saveAsFile no longer handles error codes
• added some type annotations
• modify ONC to accept Path objects for outPath
• Modify tests/readme.md for pytest tests

Worth noting that documentation on wiki is unclear whether some error codes from server should be 127/129. (e.g. test_raw_bad_filters/test_bad_filters). Other services do not list error codes on the ONC wiki.

[kan-fu]

Monday and Tuesday are our regression test days before the minor release. We will come back to the PR on Wednesday.

BTW I think it would be better to merge your PR first. I need to add some commands in tox for the pytest.

[Jacob-Stevens-Haas]

...We will come back to the PR on Wednesday.

BTW I think it would be better to merge your PR first....

Sorry, I'm confused. Were you approving this PR and saying that we'll come back to your PR on Wednesday, or were you saying that we'll come back to this PR on Wednesday and then your PR?

[kan-fu]

...We will come back to the PR on Wednesday. BTW I think it would be better to merge your PR first....

Sorry, I'm confused. Were you approving this PR and saying that we'll come back to your PR on Wednesday, or were you saying that we'll come back to this PR on Wednesday and then your PR?

Sorry for not making myself clear. It is the latter one. Eli and I are expected to work fulltime on regression test (once every month) for those two days so we don't have time to review the PR. Just don't want to leave you confused when you don't see any comments from us during these two days.

[kan-fu]

From the wiki, it seems that rawdata service does not accept propertyCode unlike scaladata service. That is why it returns 129 (parameter name error) in test_raw_bad_filters, and 127 (parameter value error) in test_no_data_filters. If you want to test 129 in scaladata, you can change propertyCodeto propertyCode1in bad_data_filters.

BTW, did you run black, isort and ruff? There are some warnings, but should be easy to fix. Also I met some conflicts between isort and ruff check I rule. I think we can just remove the I rule from pyproject.toml.

[Jacob-Stevens-Haas]

Pushed new commits with correct linting and formatting. If you want more linting/autoformatting, I'd recommend waiting until #15 is done so I can use the config that PR adds.

[kan-fu]

LGTM! Thanks for your effort. This saves me a lot of trouble when reading except Exception: raise pattern.

One small thing, if you find splitting a long string makes the string look awkward, just add # noqa: E501 to disable the linting. I tend not to be too strict about this line-too-long rule by adding several noqa myself.

For the linting, don't worry about it. We will merge yours first. As I said earlier, something is not set up correctly between isort and ruff check I rule. I will fix it in #19.

[Jacob-Stevens-Haas]

Great, thanks Kan! I tend to be pretty strict about splitting strings unless it's something like a URL that will be copy/pasted a lot.

Am I able to merge this, or are we waiting on @spencerwplovie and @eliferguson?

[kan-fu]

Great, thanks Kan! I tend to be pretty strict about splitting strings unless it's something like a URL that will be copy/pasted a lot.

Am I able to merge this, or are we waiting on @spencerwplovie and @eliferguson?

In ONC we usually need two approvals to merge a PR. They are usually occupied by other tasks. Let's wait for them to review the PR.

[eliferguson]

looks good to me. Ill merge it.

[spencerwplovie]

Apologies for not seeing this sooner, been caught up with other tasks. LGTM.