Define inputs for configuration added with include

DETAILS: Tier: Free, Premium, Ultimate Offering: GitLab.com, Self-managed, GitLab Dedicated

• Introduced in GitLab 15.11 as a beta feature.
• Made generally available in GitLab 17.0.

Use inputs to increase the flexibility of CI/CD configuration files that are designed to be reused.

Inputs can use CI/CD variables, but have the same variable limitations as the include keyword.

Define input parameters with spec:inputs

Use spec:inputs to define parameters that can be populated in reusable CI/CD configuration files when added to a pipeline. Then use include:inputs to add the configuration to a project's pipeline and set the values for the parameters.

For example, in a file named custom_website_scan.yml:

spec:
  inputs:
    job-stage:
    environment:
---

scan-website:
  stage: $[[ inputs.job-stage ]]
  script: ./scan-website $[[ inputs.environment ]]

In this example, the inputs are job-stage and environment. Then, in a .gitlab-ci.yml file, you can add this configuration and set the input values:

include:
  - local: 'custom_website_scan.yml'
    inputs:
      job-stage: 'my-test-stage'
      environment: 'my-environment'

Specs must be declared at the top of a configuration file, in a header section separated from the rest of the configuration with ---. Use the $[[ inputs.input-id ]] interpolation format outside the header section to declare where to use the inputs.

With spec:inputs:

• Inputs are mandatory by default.
• Inputs are evaluated and populated when the configuration is fetched during pipeline creation, before the configuration is merged with the contents of the .gitlab-ci.yml file.
• A string containing an input must be less than 1 MB.
• A string inside an input must be less than 1 KB.

Additionally, use:

• spec:inputs:default to define default values for inputs when not specified. When you specify a default, the inputs are no longer mandatory.
• spec:inputs:description to give a description to a specific input. The description does not affect the input, but can help people understand the input details or expected values.
• spec:inputs:options to specify a list of allowed values for an input.
• spec:inputs:regex to specify a regular expression that the input must match.
• spec:inputs:type to force a specific input type, which can be string (default when not specified), array, number, or boolean.

Define inputs with multiple parameters

You can define multiple inputs per CI/CD configuration file, and each input can have multiple configuration parameters.

For example, in a file named scan-website-job.yml:

spec:
  inputs:
    job-prefix:     # Mandatory string input
      description: "Define a prefix for the job name"
    job-stage:      # Optional string input with a default value when not provided
      default: test
    environment:    # Mandatory input that must match one of the options
      options: ['test', 'staging', 'production']
    concurrency:
      type: number  # Optional numeric input with a default value when not provided
      default: 1
    version:        # Mandatory string input that must match the regular expression
      type: string
      regex: ^v\d\.\d+(\.\d+)$
    export_results: # Optional boolean input with a default value when not provided
      type: boolean
      default: true
---

"$[[ inputs.job-prefix ]]-scan-website":
  stage: $[[ inputs.job-stage ]]
  script:
    - echo "scanning website -e $[[ inputs.environment ]] -c $[[ inputs.concurrency ]] -v $[[ inputs.version ]]"
    - if $[[ inputs.export_results ]]; then echo "export results"; fi

In this example:

• job-prefix is a mandatory string input and must be defined.
• job-stage is optional. If not defined, the value is test.
• environment is a mandatory string input that must match one of the defined options.
• concurrency is an optional numeric input. When not specified, it defaults to 1.
• version is a mandatory string input that must match the specified regular expression.
• export_results is an optional boolean input. When not specified, it defaults to true.

Input types

You can specify that an input must use a specific type with the optional spec:inputs:type keyword.

The input types are:

• array
• boolean
• number
• string (default when not specified)

When an input replaces an entire YAML value in the CI/CD configuration, it is interpolated into the configuration as its specified type. For example:

spec:
  inputs:
    array_input:
      type: array
    boolean_input:
      type: boolean
    number_input:
      type: number
    string_input:
      type: string
---

test_job:
  allow_failure: $[[ inputs.boolean_input ]]
  needs: $[[ inputs.array_input ]]
  parallel: $[[ inputs.number_input ]]
  script: $[[ inputs.string_input ]]

When an input is inserted into a YAML value as part of a larger string, the input is always interpolated as a string. For example:

spec:
  inputs:
    port:
      type: number
---

test_job:
  script: curl "https://gitlab.com:$[[ inputs.port ]]"

Array type

• Introduced in GitLab 16.11.

The content of the items in an array type can be any valid YAML map, sequence, or scalar. More complex YAML features like !reference cannot be used.

spec:
  inputs:
    rules-config:
      type: array
      default:
        - if: $CI_PIPELINE_SOURCE == "merge_request_event"
          when: manual
        - if: $CI_PIPELINE_SOURCE == "schedule"
---

test_job:
  rules: $[[ inputs.rules-config ]]
  script: ls

Multi-line input string values

Inputs support different value types. You can pass multi-string values using the following format:

spec:
  inputs:
    closed_message:
      description: Message to announce when an issue is closed.
      default: 'Hi {{author}} :wave:,

        Based on the policy for inactive issues, this is now being closed.

        If this issue requires further attention, please reopen this issue.'
---

Set input values when using include

• include:with renamed to include:inputs in GitLab 16.0.

Use include:inputs to set the values for the parameters when the included configuration is added to the pipeline.

For example, to include the scan-website-job.yml in the example above:

include:
  - local: 'scan-website-job.yml'
    inputs:
      job-prefix: 'some-service-'
      environment: 'staging'
      concurrency: 2
      version: 'v1.3.2'
      export_results: false

In this example, the inputs for the included configuration are:

Use include:inputs with multiple files

inputs must be specified separately for each included file. For example:

include:
  - component: $CI_SERVER_FQDN/the-namespace/the-project/the-component@1.0
    inputs:
      stage: my-stage
  - local: path/to/file.yml
    inputs:
      stage: my-stage

Use inputs with child pipelines

You can pass inputs to child pipelines, if the child pipeline's configuration file uses spec:inputs. For example:

trigger-job:
  trigger:
    strategy: depend
    include:
      - project: my-group/my-project
        file: ".gitlab-ci.yml"
        inputs:
          job-name: "defined"
  rules:
    - if: $CI_PIPELINE_SOURCE == 'merge_request_event'

Include the same file multiple times

You can include the same file multiple times, with different inputs. However, if multiple jobs with the same name are added to one pipeline, each additional job overwrites the previous job with the same name. You must ensure the configuration prevents duplicate job names.

For example, including the same configuration multiple times with different inputs:

include:
  - local: path/to/my-super-linter.yml
    inputs:
      linter: docs
      lint-path: "doc/"
  - local: path/to/my-super-linter.yml
    inputs:
      linter: yaml
      lint-path: "data/yaml/"

The configuration in path/to/my-super-linter.yml ensures the job has a unique name each time it is included:

spec:
  inputs:
    linter:
    lint-path:
---
"run-$[[ inputs.linter ]]-lint":
  script: ./lint --$[[ inputs.linter ]] --path=$[[ inputs.lint-path ]]

Reuse configuration in inputs

To reuse configuration with inputs, you can use YAML anchors.

For example, to reuse the same rules configuration with multiple components that support rules arrays in the inputs:

.my-job-rules: &my-job-rules
  - if: $CI_PIPELINE_SOURCE == "merge_request_event"
  - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH

include:
  - component: $CI_SERVER_FQDN/project/path/component1@main
    inputs:
      job-rules: *my-job-rules
  - component: $CI_SERVER_FQDN/project/path/component2@main
    inputs:
      job-rules: *my-job-rules

You cannot use !reference tags in inputs, but issue 424481 proposes adding this functionality.

inputs examples

Use inputs with needs

You can use array type inputs with needs for complex job dependencies.

For example, in a file named component.yml:

spec:
  inputs:
    first_needs:
      type: array
    second_needs:
      type: array
---

test_job:
  script: echo "this job has needs"
  needs:
    - $[[ inputs.first_needs ]]
    - $[[ inputs.second_needs ]]

In this example, the inputs are first_needs and second_needs, both array type inputs. Then, in a .gitlab-ci.yml file, you can add this configuration and set the input values:

include:
  - local: 'component.yml'
    inputs:
      first_needs:
        - build1
      second_needs:
        - build2

When the pipeline starts, the items in the needs array for test_job get concatenated into:

test_job:
  script: echo "this job has needs"
  needs:
  - build1
  - build2

Allow needs to be expanded when included

You can have needs in an included job, but also add additional jobs to the needs array with spec:inputs.

For example:

spec:
  inputs:
    test_job_needs:
      type: array
      default: []
---

build-job:
  script:
    - echo "My build job"

test-job:
  script:
    - echo "My test job"
  needs:
    - build-job
    - $[[ inputs.test_job_needs ]]

In this example:

• test-job job always needs build-job.
• By default the test job doesn't need any other jobs, as the test_job_needs: array input is empty by default.

To set test-job to need another job in your configuration, add it to the test_needs input when you include the file. For example:

include:
  - component: $CI_SERVER_FQDN/project/path/component@1.0.0
    inputs:
      test_job_needs: [ my-other-job ]

my-other-job:
  script:
    - echo "I want build-job` in the component to need this job too"

Add needs to an included job that doesn't have needs

You can add needs to an included job that does not have needs already defined. For example, in a CI/CD component's configuration:

spec:
  inputs:
    test_job:
      default: test-job
---

build-job:
  script:
    - echo "My build job"

"$[[ inputs.test_job ]]":
  script:
    - echo "My test job"

In this example, the spec:inputs section allows the job name to be customized.

Then, after you include the component, you can extend the job with the additional needs configuration. For example:

include:
  - component: $CI_SERVER_FQDN/project/path/component@1.0.0
    inputs:
      test_job: my-test-job

my-test-job:
  needs: [my-other-job]

my-other-job:
  script:
    - echo "I want `my-test-job` to need this job"

Specify functions to manipulate input values

• Introduced in GitLab 16.3.

You can specify predefined functions in the interpolation block to manipulate the input value. The format supported is the following:

$[[ input.input-id | <function1> | <function2> | ... <functionN> ]]

Details:

• Only predefined interpolation functions are permitted.
• A maximum of 3 functions may be specified in a single interpolation block.
• The functions are executed in the sequence they are specified.

spec:
  inputs:
    test:
      default: 'test $MY_VAR'
---

test-job:
  script: echo $[[ inputs.test | expand_vars | truncate(5,8) ]]

In this example, assuming the input uses the default value and $MY_VAR is an unmasked project variable with value my value:

1. First, the function expand_vars expands the value to test my value.
2. Then truncate applies to test my value with a character offset of 5 and length 8.
3. The output of script would be echo my value.

Predefined interpolation functions

expand_vars

• Introduced in GitLab 16.5.

Use expand_vars to expand CI/CD variables in the input value.

Only variables you can use with the include keyword and which are not masked can be expanded. Nested variable expansion is not supported.

Example:

spec:
  inputs:
    test:
      default: 'test $MY_VAR'
---

test-job:
  script: echo $[[ inputs.test | expand_vars ]]

In this example, if $MY_VAR is unmasked (exposed in job logs) with a value of my value, then the input would expand to test my value.

truncate

• Introduced in GitLab 16.3.

Use truncate to shorten the interpolated value. For example:

• truncate(<offset>,<length>)

Example:

$[[ inputs.test | truncate(3,5) ]]

Assuming the value of inputs.test is 0123456789, then the output would be 34567.

Troubleshooting

YAML syntax errors when using inputs

CI/CD variable expressions in rules:if expect a comparison of a CI/CD variable with a string, otherwise a variety of syntax errors could be returned.

You must ensure that expressions remain properly formatted after input values are inserted into the configuration, which might require the use of additional quote characters.

For example:

spec:
  inputs:
    branch:
      default: $CI_DEFAULT_BRANCH
---

job-name:
  rules:
    - if: $CI_COMMIT_REF_NAME == $[[ inputs.branch ]]

In this example:

• Using include: inputs: branch: $CI_DEFAULT_BRANCH is valid. The if: clause evaluates to if: $CI_COMMIT_REF_NAME == $CI_DEFAULT_BRANCH, which is a valid variable expression.
• Using include: inputs: branch: main is invalid. The if: clause evaluates to if: $CI_COMMIT_REF_NAME == main, which is invalid because main is a string but is not quoted.

Alternatively, add quotes to resolve some variable expression issues. For example:

spec:
  inputs:
    environment:
      default: "$ENVIRONMENT"
---

$[[ inputs.environment | expand_vars ]] job:
  script: echo
  rules:
    - if: '"$[[ inputs.environment | expand_vars ]]" == "production"'

In this example, quoting the input block and also the entire variable expression ensures valid if: syntax after the input is evaluated. The internal and external quotes in the expression must not be the same character. You can use " for the internal quotes and ' for the external quotes, or the inverse. On the other hand, the job name does not require any quoting.