Bail on activation when CONDA_BUILD is set

[maresb]

Otherwise, the GDB symlinks will be included in packages built with libarrow as a host dependency.

Closes #1478

Checklist

• Used a personal fork of the feedstock to propose changes
• Bumped the build number (if the version is unchanged)
• Reset the build number to 0 (if the version changed)
• Re-rendered with the latest conda-smithy (Use the phrase @conda-forge-admin, please rerender in a comment in this PR for automated rerendering)
• Ensured the license file is being packaged.

[conda-forge-admin]

Hi! This is the friendly automated conda-forge-linting service.

I wanted to let you know that I linted all conda-recipes in your PR (recipe/meta.yaml) and found some lint.

Here's what I've got...

For recipe/meta.yaml:

• ❌ This recipe is using a compiler, which now requires adding a build dependence on {{ stdlib("c") }} as well. Note that this rule applies to each output of the recipe using a compiler. For further details, please see META: {{ stdlib("c") }} migration conda-forge.github.io#2102.

For recipe/meta.yaml:

• ℹ️ The boost-cpp output has been superseded by libboost-devel (as of 1.82), which now comes with a run-export (on libboost) as well. In case you only needed the boost headers, you should use libboost-headers.
• ℹ️ The recipe is not parsable by parser conda-souschef (grayskull). This parser is not currently used by conda-forge, but may be in the future. We are collecting information to see which recipes are compatible with grayskull.
• ℹ️ The recipe is not parsable by parser conda-recipe-manager. The recipe can only be automatically migrated to the new v1 format if it is parseable by conda-recipe-manager.

_{This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/12934897643. Examine the logs at this URL for more detail.}

[conda-forge-admin]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe/meta.yaml) and found it was in an excellent condition.

I do have some suggestions for making it better though...

For recipe/meta.yaml:

• ℹ️ The recipe is not parsable by parser conda-souschef (grayskull). This parser is not currently used by conda-forge, but may be in the future. We are collecting information to see which recipes are compatible with grayskull.
• ℹ️ The recipe is not parsable by parser conda-recipe-manager. The recipe can only be automatically migrated to the new v1 format if it is parseable by conda-recipe-manager.

_{This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/12957715354. Examine the logs at this URL for more detail.}

[h-vetinari]

Thank you very much! 🙏 🙂

[h-vetinari]

Ah wait, CI is deep red, failing in

arrow-cpp-feedstock/recipe/meta.yaml

Lines 282 to 283 in b271605

	- test -f $PREFIX/share/gdb/auto-load/$PREFIX/lib/libarrow.so.{{ so_version }}-gdb.py # [linux]
	- test -f $PREFIX/share/gdb/auto-load/$PREFIX/lib/libarrow.{{ so_version }}.dylib-gdb.py # [osx]

Is -n the right argument here?

[h-vetinari]

It seems that CONDA_BUILD is still set during the test phase for some reason (so probably your implementation was correct on the first go!). I had been under the impression that this is not the case, but looking at the docs, perhaps I was wrong...? 🤔

In this case, just add something like

# need to unset this to test gdb-integration despite https://github.com/conda-forge/arrow-cpp-feedstock/issues/1478
- unset CONDA_BUILD  # [unix]

before

arrow-cpp-feedstock/recipe/meta.yaml

Lines 282 to 283 in b271605

	- test -f $PREFIX/share/gdb/auto-load/$PREFIX/lib/libarrow.so.{{ so_version }}-gdb.py # [linux]
	- test -f $PREFIX/share/gdb/auto-load/$PREFIX/lib/libarrow.{{ so_version }}.dylib-gdb.py # [osx]

[maresb]

Indeed.

In this case, just add something like

# need to unset this to test gdb-integration despite https://github.com/conda-forge/arrow-cpp-feedstock/issues/1478
- unset CONDA_BUILD  # [unix]

I wouldn't expect that to have any effect post-initialization though.

I'm a bit stumped at the moment. (Or do you still think your suggestion might work?)

[h-vetinari]

I wouldn't expect that to have any effect post-initialization though.

Ah, true!

@jaimergp @jakirkham @beckermr, would you have any idea whether it's intentional that CONDA_BUILD is set also during the test phase? I seemed to remember that that's not the case, but perhaps that was just my imagination?

[xhochy]

You should be able to lookup the testing phase using "${CONDA_BUILD_STATE:-0}" = "TEST".

[maresb]

Thanks @xhochy, trying that now.

TODO: remove the debug commit... converting to draft

[h-vetinari]

I took the liberty to clean up the debug line and add a guard against unset variables. Since the resulting changes

git diff d1919bbf00cd928e036d0987ab86d25fa1a7b119..197c2cb181665e0bc9059d654a1fdeadb076c045
diff --git a/recipe/activate.sh b/recipe/activate.sh
index bc796802..2ec4ad24 100755
--- a/recipe/activate.sh
+++ b/recipe/activate.sh
@@ -6,7 +6,7 @@

 # doesn't come with a deactivate script, because the symlink
 # is benign and doesn't need to be deleted.
-CF_LIBARROW_ACTIVATE_LOGGING=1
+
 _la_log() {
     if [ "${CF_LIBARROW_ACTIVATE_LOGGING:-}" = "1" ]; then
         # The following loop is necessary to handle multi-line strings
@@ -21,9 +21,9 @@ _la_log() {
 # Skip activation if CONDA_BUILD environment variable is set.
 # (CONDA_BUILD is also set in the test stage, and we don't want to skip there.)
 # Otherwise, the symlinks will be included in packages built with libarrow as a host dependency.
-# <https://github.com/prefix-dev/rattler-build/issues/979#issuecomment-2243070530>
-if [ -n "$CONDA_BUILD" ] && [ "$CONDA_BUILD_STATE" != "TEST" ]; then
-    _la_log "CONDA_BUILD is set to $CONDA_BUILD and CONDA_BUILD_STATE is $CONDA_BUILD_STATE, skipping libarrow activation."
+# see https://github.com/conda-forge/arrow-cpp-feedstock/issues/1478
+if [ -n "$CONDA_BUILD" ] && [ "${CONDA_BUILD_STATE:-0}" != "TEST" ]; then
+    _la_log "CONDA_BUILD is set to $CONDA_BUILD (and CONDA_BUILD_STATE != \"TEST\"), skipping libarrow activation."
     return 0
 fi

do not change anything functionally, I'm further taking the liberty of merging this. Hope that's OK. :)

[h-vetinari]

Thanks a lot for the PR, and thanks for the tip @xhochy! 🥳

[maresb]

Of course, thanks so much @h-vetinari!