JSON schema: Detaxizer

• Fork the pipeline to your user account, if you have not done so before (Make sure you deselect Copy the master branch only)
• Go to the dev branch and sync the upstream changes (Sync fork -> Update branch)
• Create a branch called json-schema-improvements with the dev branch as base in the fork
• Open the json-schema-improvements branch of the fork either locally (git clone) or online (via Github Codespaces, GitPod...)
• Open the nextflow_schema.json file
• Check that the value in "$schema" is "https://json-schema.org/draft/2020-12/schema". The pipeline should be updated to nf-schema if this is not the case. See the migration guide on how to do this
• Check the schema for insufficient validation keywords and error messages. See the guide below for more information on this
• Do the same thing for assets/schema_input.json
• Run nextflow run main.nf -profile test,docker --outdir results -stub to check for any failures in validation with the basic test setup
• Update the changelog of the pipeline
• Push all changes to your fork (if you have not done so yet)
• Open a pull request to the dev branch of the nf-core repository
• Request a review in #hackathon-mar-2025-json-schema-improvements or in the pipeline channel (e.g. #sarek for nf-core/sarek) and set the issue status to Ready to review (PR: )
• Wait for approval and merge the pull request once everything works again
• Set the issue status to Done

Checking the validation keywords

This section contains options for each type of parameter (the type can be found in the "type" keyword)

IMPORTANT: If an errorMessage is present, please update it so that it covers all possible validation errors for the type (the more it covers the better)

ALSO IMPORTANT: Check that the help_text and description keywords are clear enough

Type-independent keywords

• enum: A list of allowed values for the value
• default: The default value, make sure this is filled in if the default value is not null, "" or false
• required: A list of required parameters for each section

string

• minLength and maxLength: specify the minimal and/or maximum length of the string value
• pattern: A regular expression the string value has to adhere to
• format: This is mainly used to define if a value should be a "file-path", "directory-path", "path" or "file-path-pattern". Other formats are allowed that are part of the official JSON schema specs, but are rarely used in Nextflow JSON schemas
• exists: Make sure this is true if the file should exist (Be careful, some times files or directories do not have to exist for a pipeline to succesfully start e.g. --outdir)
• schema: The relative path from the root of the pipeline to another schema to validate the file specified by that value. Please also check this schema very carefully

number and integer

• multipleOf: The value needs to be a multiple of this number
• minimum, exclusiveMinimum, maximum and exclusiveMaximum: Specify a specific range in which the numeric value should be

boolean

The boolean type has no type-specific keywords

array

WARNING: This type is not supported in nextflow_schema.json, but is supported in other schemas

• uniqueEntries: A list of keys that need to be unique in the objects present in this array
• uniqueItems: Set to true if you want each entry in the array to be unique

object

WARNING: This type is not supported in nextflow_schema.json (apart from the root "type" keyword of the file and of each section in "$defs"), but is supported in other schemas

This type can be very complex and will not be used in 99% of the cases. You can take a look at the official object documentation or ask for help in #hackathon-mar-2025-json-schema-improvements if you need to check object types

Start with nextflow_schema.json and assets/schema_input.json, then read the linked nf-schema migration guide to understand the expected schema version and validation keywords. Run nextflow run main.nf -profile test,docker --outdir results -stub before and after reviewing the schemas. Done means the schemas and changelog are updated and the test pipeline validates successfully.