From 3f184159c83db61a089e7ae9db3bb81446301e13 Mon Sep 17 00:00:00 2001 From: Pablo Moreno Date: Sun, 11 Feb 2024 10:38:09 +0000 Subject: [PATCH 1/3] Update README.md --- README.md | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/README.md b/README.md index 9b38e7e5..2fc39da6 100644 --- a/README.md +++ b/README.md @@ -20,6 +20,22 @@ For development installation, we suggest following the github actions python-pac Currently, tests run on python 3.9, so those are the recommended versions if not installing via conda. BKNN doesn't currently install on Python 3.10 due to a skip in Bioconda. +## How to update + +Updating the stack that depends on this script is unfortunately complex, it gets all the way to Galaxy tools using this for production pipelines in at least two different institutions. + +1. Open a branch here from develop +2. Change the conda dependencies versions desired in the test-env.yaml file (if any) +3. Make your changes in the branch (including adding any tests to the bats files for new areas or functionalities). +4. Iterate with the CI on the branch until all test pass with the new changes and dependencies. +5. Get reviews and merge to develop. +6. Open a bump branch in bioconda, making sure that it points to the develop branch of this repo rather than the pypi release and that all dependencies in bioconda reflect what is currently in the test-env.yaml file of this repo (which was used to test in the feature branch here in points 2 to 4). +7. Once the tests pass in bioconda, ask the bot to fetch the artifacts, download the linux artifact. Leave the bioconda branch in waiting, do not merge it (and probably add a message saying so) +8. Use the image .tar.gz inside the artifact to create a new local docker container on the linux machine where you intend to test the Galaxy tools. +9. Open a local feature branch on container-galaxy-sc repo to start trying the changes. +10. Change the scanpy-scripts macro2 requirements part to point to the newly created container. +11. Run planemo test, it should use the local container. + ## Test installation There is an example script included: From 34d917352497346d00e57ae066adb421fb2ea736 Mon Sep 17 00:00:00 2001 From: Pablo Moreno Date: Sun, 11 Feb 2024 10:49:59 +0000 Subject: [PATCH 2/3] Precisions on steps --- README.md | 24 +++++++++++++++++++++--- 1 file changed, 21 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index 2fc39da6..7f083b08 100644 --- a/README.md +++ b/README.md @@ -32,9 +32,27 @@ Updating the stack that depends on this script is unfortunately complex, it gets 6. Open a bump branch in bioconda, making sure that it points to the develop branch of this repo rather than the pypi release and that all dependencies in bioconda reflect what is currently in the test-env.yaml file of this repo (which was used to test in the feature branch here in points 2 to 4). 7. Once the tests pass in bioconda, ask the bot to fetch the artifacts, download the linux artifact. Leave the bioconda branch in waiting, do not merge it (and probably add a message saying so) 8. Use the image .tar.gz inside the artifact to create a new local docker container on the linux machine where you intend to test the Galaxy tools. -9. Open a local feature branch on container-galaxy-sc repo to start trying the changes. -10. Change the scanpy-scripts macro2 requirements part to point to the newly created container. -11. Run planemo test, it should use the local container. + ``` + # once downloaded and unzipped the artifact, you can do (using the correct image name) + gzip -dc LinuxArtifacts/images/scanpy-scripts:1.9.0--pyhdfd78af_0.tar.gz | docker load + # this will create locally + Loaded image: quay.io/biocontainers/scanpy-scripts:1.9.0--pyhdfd78af_0 + ``` +10. Open a local feature branch on container-galaxy-sc repo to start trying the changes. +11. Change the scanpy-scripts macro2 requirements part to point to the newly created container. + ``` + + + + quay.io/biocontainers/scanpy-scripts:1.9.0--pyhdfd78af_0 + + ... + ``` + +13. Run planemo test, it should use the local container. ## Test installation From fc1daaf8c83147f4a858479bf636b102eb7c20cc Mon Sep 17 00:00:00 2001 From: Pablo Moreno Date: Fri, 14 Feb 2025 11:49:55 +0000 Subject: [PATCH 3/3] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 96a96450..1289313b 100644 --- a/README.md +++ b/README.md @@ -30,7 +30,7 @@ Updating the stack that depends on this script is unfortunately complex, it gets 4. Iterate with the CI on the branch until all test pass with the new changes and dependencies. 5. Get reviews and merge to develop. 6. Open a bump branch in bioconda, making sure that it points to the develop branch of this repo rather than the pypi release and that all dependencies in bioconda reflect what is currently in the test-env.yaml file of this repo (which was used to test in the feature branch here in points 2 to 4). -7. Once the tests pass in bioconda, ask the bot to fetch the artifacts, download the linux artifact. Leave the bioconda branch in waiting, do not merge it (and probably add a message saying so) +7. Once the tests passes in bioconda, ask the bot to fetch the artifacts, download the linux artifact (to run the tests against that artifact or inside the artifact container). Leave the bioconda branch in draft and do not merge it (and probably add a message saying so) until you can verify that the created artifact can be used to run the tests (see steps below). 8. Use the image .tar.gz inside the artifact to create a new local docker container on the linux machine where you intend to test the Galaxy tools. ``` # once downloaded and unzipped the artifact, you can do (using the correct image name)