Closed edmorley closed 4 years ago
@edmorley Got it, thanks.
@edmorley tracked this down to our open source changelogs repo: https://github.com/pyupio/changelogs/blob/0cdb929ac4546c766cd7eef9ae4eb4baaa08f452/changelogs/parser.py#L35
I should get to this in the next few days.
cc @jayfk for context.
@edmorley I have created a PR with a fix here: https://github.com/pyupio/changelogs/pull/139/files
@jayfk did you want to strip #'s from GitHub issue numbers? If so, I can update the regex accordingly.
@Jwomers The thing is that links to GitHub issues create quite some noise. There were a couple of issues in the past that had hundred of references from unrelated issues created by the bot.
In most cases this is not a problem because the issues are closed. But if there's an issue where people are still working on a fix and are discussing changes, it's distracting to see all these pull request references.
That's why I stripped these comments out.
But if you are working on a better regex, it should be pretty simple to just make sure that the line does not contain github.com.
@jayfk got it, so I will add a regex to replace#[0-9]*
with [0-9]*
👍
Thanks for all information shared, guys. Fixed with https://github.com/pyupio/changelogs/pull/139
Should be available with next release to PyPi.
Hi
In this PR: mozilla/treeherder#3784
The changelog section contains this URL:
https://django-filter.readthedocs.io/en/master/guide/migration.htmlmigrating-to-2-0
Whereas the actual URL in the original changelog is:
https://django-filter.readthedocs.io/en/master/guide/migration.html#migrating-to-2-0
I'm presuming the
#
was removed by the process that sanitizes the GitHub issue numbers (ie converts#123
to123
).Please could the regex be tightened up, to avoid false positives?