Issue 1316 - longrepr is a string when pytrace=False

[CarycaKatarzyna]

This pull request fixes #1316. I've added new _TracebackStyle - no_pytrace. For that style ReprExceptionInfo has as few information as possible. For example report.longrepr.chain is

[(ReprTraceback(reprentries=[ReprEntry(lines=['Hello World'], reprfuncargs=None, reprlocals=None, reprfileloc=None, style='no_pytrace')], extraline=None, style='no_pytrace'), None, None)]

for code contained in issue #1316.

[nicoddemus]

Thanks @CarycaKatarzyna for tackling this very old issue! 👍

Please take a look at my comments.

[nicoddemus]

Rebased to fix linting.

Thanks a lot @CarycaKatarzyna!

[nicoddemus]

We should squash all commits on merge btw. 👍

[bluetech]

Great cleanup, that str was really hacky. I left some minor comments but otherwise LGTM.

Also, we should consider adding value as a supported option to --tb - this will satisfy the use case in #6651 (and replace PR #6652). WDYT?

[bluetech]

Suggested change

	``TestReport.longrepr`` is now always an instance of ``ReprExceptionInfo``. Previously it was a ``str`` when a test failed with ``pytest.fail(..., pytrace=False)``.
	``TestReport.longrepr`` is now always an instance of ``ReprExceptionInfo``. Previously it was an ``str`` when a test failed with ``pytest.fail(..., pytrace=False)``.

[CarycaKatarzyna]

str starts with 's' which is a consonant. Why do you think that 'an' is a proper spelling?

[bluetech]

I read it as ESS-TEE-ARR, but maybe others don't. Not a big deal either way :)

[nicoddemus]

Also, we should consider adding value as a supported option to --tb - this will satisfy the use case in #6651 (and replace PR #6652). WDYT?

Great idea! We should consider doing that in a separate PR thought.

[bluetech]

Thanks @CarycaKatarzyna!