![\"Screenshot](\"https://user-images.githubusercontent.com/9599/178353732-77e58ddd-c78c-4032-aeea-cb388bac8ec6.jpg\")
My simonwillisonblog-backup workflow periodically creates a JSON backup of my blog's PostgreSQL database, using db-to-sqlite and sqlite-diffable. It then commits any changes back to the repo using this pattern:
\n - name: Commit any changes\n run: |-\n git config user.name \"Automated\"\n git config user.email \"actions@users.noreply.github.com\"\n git add simonwillisonblog\n timestamp=$(date -u)\n git commit -m \"Latest data: ${timestamp}\" || exit 0\n git pushI decided to upgrade it to also build and deploy a SQLite database of the content to datasette.simonwillison.net - but only if a change had been detected.
\nI figured out the following pattern for doing that.
\nFirst, I added a line to the above block that set a change_detected output variable for that step if it made it past the || exit 0. I also added an id to the step so I could reference it later on:
- name: Commit any changes\n id: commit_and_push\n run: |-\n git config user.name \"Automated\"\n git config user.email \"actions@users.noreply.github.com\"\n git add simonwillisonblog\n timestamp=$(date -u)\n git commit -m \"Latest data: ${timestamp}\" || exit 0\n git push\n echo \"::set-output name=change_detected::1\"This next piece took me a while to figure out: I also had to declare that output variable at the top of the initial job, copying the result of the named step:
\njobs:\n backup:\n runs-on: ubuntu-latest\n outputs:\n change_detected: ${{ steps.commit_and_push.outputs.change_detected }}Without this, the output is not visible to the second job.
\nMy second job started like this:
\n build_and_deploy:\n runs-on: ubuntu-latest\n needs: backup\n if: ${{ inputs.force_deploy || needs.backup.outputs.change_detected }}\n steps:This second job is called build_and_deploy and specify that it needs: backup - so it should run directly after that backup job completes.
That new job has an if: expression which looks at needs.backup.outputs.change_detected to read the variable that was set by my echo \"::set-output line.
I'm also checking inputs.force_deploy here. That's a separate mechanism, which allows me to trigger the workflow with manually and specify that a deploy should happen even if no changes were detected - useful for when I alter the code that configures the deployed Datasette instance.
The force_deploy variable comes from this section at the start of the YAML:
on:\n push:\n branches:\n - main\n workflow_dispatch:\n inputs:\n force_deploy:\n description: 'Deploy even if no changes detected'\n required: false \n type: booleanThis configuration adds the following UI which I can use to manually trigger the workflow:
\n