You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I encountered the same problem, and I noticed that others PR cannot pass the check as well.
I switch the runner image of doctest to ubuntu-24.04, and get the pass result.
I guess maybe there is some version misalign.
Yes I have the same problem as you, and the error message from log are all identical with me, so I create another branch on my repo and run all test as same with PR.
Maybe you can run all test on your another test branch to check result.
For some reason it can't find blurb when running blurb help, even though the previous step had created a venv: The venv has been created in the ./venv directory.
diff --git a/Doc/Makefile b/Doc/Makefile
index a090ee5ba92..5ad04120b9f 100644
--- a/Doc/Makefile+++ b/Doc/Makefile@@ -55,6 +55,7 @@ build:
# Look first for a Misc/NEWS file (building from a source release tarball
# or old repo) and use that, otherwise look for a Misc/NEWS.d directory
# (building from a newer repo) and use blurb to generate the NEWS file.
+ $(BLURB) help
@if [ -f ../Misc/NEWS ] ; then \
echo "Using existing Misc/NEWS file"; \
cp ../Misc/NEWS build/NEWS; \
❯ make -C Doc/ PYTHON=../python.exe SPHINXERRORHANDLING="-W --keep-going" doctestmkdir -p buildPATH=./venv/bin:$PATH blurb help/bin/sh: ./venv/bin/blurb: /Users/hugo/github/python/cpython/main/Doc/venv/bin/python@EXE_SUFFIX@: bad interpreter: No such file or directorymake[1]: *** [build] Error 126Testing of doctests in the sources finished, look at the results in build/doctest/output.txtmake: *** [doctest] Error 1
That EXE_SUFFIX was a change added in e924bb6 / #125699.
Reverting this commit fixes the doctest locally and on CI:
That EXE_SUFFIX was a change added in e924bb6 / #125699.
Reverting this commit fixes the doctest locally and on CI:
Let's do the revert; I can land it. We've tried to rename configure/Make variables before, and it's always trickier than you'd expect. I've got a better idea of how to solve gh-125698.
@hugovk, it could be a caching thing. Are we caching the docs venv in CI? If so, that could explain the breakage. Is there a way to invalidate the CI cache?
Activity
Damien-Chen commentedon Oct 20, 2024
I encountered the same problem, and I noticed that others PR cannot pass the check as well.
I switch the runner image of doctest to
ubuntu-24.04, and get the pass result.I guess maybe there is some version misalign.
Damien-Chen commentedon Oct 20, 2024
BTW, please pull and rebase before testing.
rruuaanng commentedon Oct 20, 2024
How should I implement it specifically? Have you solved the problem?
Damien-Chen commentedon Oct 20, 2024
I tested it on my branch here, but the result is difference between my branch and my PR check.
rruuaanng commentedon Oct 20, 2024
But this is just your branch. It seems that we cannot modify the CI server of the main branch.
Damien-Chen commentedon Oct 20, 2024
Yes you are right, so I abort all modification and use original CI yml file to test again, and the result is pass(attached in my last comment).
rruuaanng commentedon Oct 20, 2024
So what should I do in this situation? You should go to those links to review.
Damien-Chen commentedon Oct 20, 2024
Yes I have the same problem as you, and the error message from log are all identical with me, so I create another branch on my repo and run all test as same with PR.
Maybe you can run all test on your another test branch to check result.
rruuaanng commentedon Oct 20, 2024
I mean, how can I eliminate this error and get it to pass CI? If you find a solution, you should review it in my list above.
hugovk commentedon Oct 20, 2024
For some reason it can't find
blurbwhen runningblurbhelp, even though the previous step had created a venv:The venv has been created in the ./venv directory.The last passing build is: https://github.com/python/cpython/actions/runs/11422034235/job/31779554757?pr=125739
The first failing build is: https://github.com/python/cpython/actions/runs/11422219191/job/31779951386?pr=125571
rruuaanng commentedon Oct 20, 2024
What should we do about this problem? From the list I made, it seems that I'm not the only one who has this problem :-(
hugovk commentedon Oct 20, 2024
We should figure out the cause and fix it :)
I've confirmed it's not caused by yesterday's release of blurb 1.3.0 (https://github.com/python/blurb/releases/tag/v1.3.0) and also not by yesterday's bump of the min Sphinx version to 7.2.6 (2bb7ab7).
hugovk commentedon Oct 20, 2024
Editing this locally:
And running:
And then:
That
EXE_SUFFIXwas a change added in e924bb6 / #125699.Reverting this commit fixes the doctest locally and on CI:
cc @erlend-aasland
Damien-Chen commentedon Oct 20, 2024
Should that change need to be merge to main branch ?
erlend-aasland commentedon Oct 20, 2024
@hugovk, is
EXEEXTused by blurb?hugovk commentedon Oct 20, 2024
No: https://github.com/search?q=repo%3Apython%2Fblurb%20EXEEXT&type=code
erlend-aasland commentedon Oct 20, 2024
Hugo:
Let's do the revert; I can land it. We've tried to rename configure/Make variables before, and it's always trickier than you'd expect. I've got a better idea of how to solve gh-125698.
erlend-aasland commentedon Oct 20, 2024
@hugovk, it could be a caching thing. Are we caching the docs venv in CI? If so, that could explain the breakage. Is there a way to invalidate the CI cache?
AA-Turner commentedon Oct 20, 2024
We cache pip's cache:
https://github.com/python/cpython/blob/main/.github/workflows/reusable-docs.yml#L102-L107
It seems you can use
gh actions-cacheto delete the cache, or just changerequirements.txtso that the cache key changes.A
erlend-aasland commentedon Oct 20, 2024
I reverted the
EXEEXTrename. Closing this as completed.