Skip to content

Releases: Tanbouz/codepipeline-artifact-merge

CodePipeline policy and Nodejs10.x upgrade

14 Jul 17:27
94ea90e
Compare
Choose a tag to compare
  • Upgrade to Nodejs10.x
  • Use CodePipelineLambdaExecutionPolicy to allow Lambda function to report job status back to CodePipeline without the need to add custom IAM policy manually.
  • Rename CloudFormation/SAM deploy.yml template to template.yaml
  • Update node dependencies

gzip support & NodeJs 8.10

11 Aug 15:59
cafa02e
Compare
Choose a tag to compare
  • Accepts gzipped tar balls as input artifacts
  • Upgraded Lambda function to use NodeJS 8.10
  • Lambda default memory size is now 512 mb
  • Default timeout for this lambda function is now 60 seconds.

Thanks @RainaWLK for this update.

Fix to handle file/folder permissions

05 Apr 17:53
ae05e79
Compare
Choose a tag to compare

Explicitly state the platform when generating the zip so it knows how to handle and maintain file/folder permissions within the resulting zip file.

Thanks @maxholman

Revisions

17 Feb 10:37
346eb73
Compare
Choose a tag to compare

Revisions
Creates a file with the revision-id of the input artifact in the output. This allows the succeeding codepipeline stages such as CodeBuild to use the git/S3 revision number for versioning. To set this option, enter the JSON string { "revisions": true } as an input parameter to lambda.

Thanks @ts33 for this addition

Subfolder merge & tests

02 Feb 17:07
Compare
Choose a tag to compare
  • Subfolder merge mode: Option to have each input artifact contained in its own folder within the output artifact zip, with the folder name being the input artifact name itself. To set this option, enter the JSON string '{ "subfolder": true }' as an input parameter to the lambda.

  • Tests
    mergeArtifacts test
    mergeArtifactsWithSubFolder test
    Options parameter validation test

Credits for the new feature & tests goes to @ts33 , thanks!

Handle missing input/output artifacts

22 Jan 06:29
Compare
Choose a tag to compare
  • If no input artifacts are configured, an output artifact won't be uploaded and the pipeline will continue successfully.
  • If a single input artifact is defined, it will be passed as an output artifact.
  • An output artifact must be configured in CodePipeline. Otherwise, the pipeline will fail.

SSE S3 detection

14 Jan 14:31
Compare
Choose a tag to compare
  • Encrypt output artifacts using the same S3 Server Side Encryption type of the input artifacts. CodePipelines created through AWS Console restrict their S3 buckets policies to aws:kms encryption.
  • Removed lambda parameter based encryption configuration.