-
-
Notifications
You must be signed in to change notification settings - Fork 38
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Stale *.egg-info/SOURCES.txt can hide problems with MANIFEST.in #1
Comments
Running
Which python version are you using? |
Please tell me how I can reproduce the issue, and I'll do my best to fix it. As Maurits already said, |
You are right, just running egg_info doesn't help, I have to actually delete SOURCES.txt to trigger the update. The problem is that egg_info will read SOURCES.txt and then MANIFEST.in and merge them, meaning that removing something from MANIFEST.in will not update SOURCES.txt unless you first delete SOURCES.txt. For example, pyroma has a |
Yes, I've been bitten by this behavior of setuptools before. I've a plan to detect this:
This will solve your problem (not noticing that a MANIFEST.in is incomplete because the required files are still included thanks to a stale *.egg-info/SOURCES.txt) and some other problems too (not noticing that a MANIFEST.in is incomplete because you've got setuptools-git installed). I was procrastinating on this because
and now suddenly I have come up with the perfect solution:
Thanks for the inspiration! |
Right, a checkout/export would also not solve this if you check in .egg-info in the VCS (yes, I've seen this done several times). So that seems like the right way to go. Thanks! |
I've no words to describe the feeling I got when you told me people actually commit .egg-info into the VCS. I'll add an explicit check for that. |
This if fixed in master. I'll release 0.9 later today. |
0.9 is out. |
Well, check-manifest did take a lot of pain out of making MANIFEST.in files. Thanks!
However, one pain is still there. I've noticed that when changing MANIFEST.in it doesn't make any difference unless I make sure the .egg-info is updated. I assume it's the SOURCES.txt that is the culprit.
I'm not sure where the bug is that causes that, but check-release doesn't change it. I'd suggest that check-release verifies that if there is a .egg-info/SOURCES.txt is also runs setup.py egg_info to rebuild it , at least if it's timestamp is before the timestamp of MANIFEST.in.
The text was updated successfully, but these errors were encountered: