Is it possible to combine multiple branches that have the same steps within bitbucket pipelines?
ex: The teams I work on use one of two names for their review branches, either "rev" or "staging". Either way the same steps are used to publish to our review server. Right now the branches are called out separately.
pipelines:
branches:
rev:
steps:
- echo 'step'
staging:
steps:
- echo 'step'
but could it be something like
pipelines:
branches:
rev|staging:
steps:
- echo 'step'
A comma-separated list inside braces appears to work:
pipelines:
branches:
'{rev,staging}':
- step:
script:
- echo 'step'
This is a full example on how you can reuse some steps:
image: yourimage:latest
definitions:
services: ... # Service definitions go there
steps:
- step: &Test-step
name: Run tests
script:
- npm install
- npm run test
- step: &Deploy-step
name: Deploy to staging
deployment: staging
script:
- npm install
- npm run build
- fab deploy
pipelines:
default:
- step: *Test-step
- step: *Deploy-step
branches:
master:
- step: *Test-step
- step:
<<: *Deploy-step
deployment: production
trigger: manual
Read more about YAML anchors: https://confluence.atlassian.com/bitbucket/yaml-anchors-960154027.html
Instead of interpreting rev|staging
, a far more natural way of implementing that would be using a flow style sequence as a key:
pipelines:
branches:
[rev, staging]:
- step:
script:
- echo 'step'
That would alleviates the need for quoting and for making sure spaces, or an extra (trailing) comma, make no semantic difference. Depending on the library that bitbucket uses to process this, the above might parse correctly, but not load (e.g. PyYAML cannot handle the above, but ruamel.yaml
). I have not been able to verify if this preferable way actually works in bitbucket.
There are two ways that do work, one using the familiar YAML functionality of anchors and aliases to provide repeated (complex) data structures only once:
pipelines:
branches:
rev: &sharedsteps
- step:
script:
- echo 'step'
staging: *sharedsteps
The other possibility is, as others have indicated, to use some a non-standard, bitbucket specific, interpretation of scalar keys with embedded comma's. I have not found clear documentation on this, but the glob patterns seem applicable, so you can use {rev,staging}
as a key.
What is ugly about this is that {
is the flow-style sequence indicator in YAML, so that scalar needs to be quoted:
pipelines:
branches:
"{rev,staging}":
- step:
script:
- echo 'step'
The above was updated using the corrected step syntax that BlueM provided
As requested by Anthon in a comment to his answer, this is his perfect solution, but with the correct YAML structure as expected by Bitbucket Pipelines:
pipelines:
branches:
rev: &sharedsteps
- step:
script:
- echo 'step'
staging: *sharedsteps
With Bitbucket 5.8 in order to be able to manually trigger the pipeline I had to use this format:
pipelines:
branches:
rev,staging:
- step:
script:
- echo 'step'
So basically just comma separated branch list which need the same pipeline
来源:https://stackoverflow.com/questions/42304478/bitbucket-pipelines-multiple-branches-with-same-steps