Models

Every output has an independent model. A caller supplies only the published rsm-schema metadata needed by that file.

The documentation models are ordinary Markdown file models:

• DocumentationOverviewModel writes overview.md.

• DocumentationUserModel writes usage.md.

• DocumentationDeploymentModel writes deployment.md.

• DocumentationDeveloperModel writes developer.md.

• DocumentationReferenceModel writes reference.md.

• DocumentationLegalModel writes legal.md.

The models do not select or configure a documentation builder. Callers choose the pages they need and provide any builder configuration.

from rs_files_templates import (
    DocumentationOverviewModel,
    DocumentationUserModel,
    render_many,
)

render_many(
    [
        DocumentationOverviewModel(
            project_name="Example",
            project_slug="example",
            project_short_description="Reusable research software.",
        ),
        DocumentationUserModel(
            project_name="Example",
            project_slug="example",
            project_manager="uv",
            programming_languages={"entries": [{"name": "Python"}]},
        ),
    ],
    "docs",
)

ReadmeModel produces a complete metadata-only README. Installation, usage, citation, support, licensing, and development guidance are resolved from its rsm-schema fields. Repository generators can post-process the Markdown to add repository-specific content, such as a badge for a generated workflow. Top-level EDAM topics appear as project research domains. CodeMetaModel maps them to applicationSubCategory, while BiotoolsModel maps them to bio.tools EDAM topics and combines them with function-level EDAM operations, inputs, and outputs.

from rs_files_templates import ReadmeModel, render_readme

render_readme(
    ReadmeModel(
        project_name="Example",
        project_slug="example",
        project_manager="uv",
        topics={"entries": [{"term": "Data analysis"}]},
        software_functions={"entries": [{"cmd": "example --help"}]},
    ),
    "output",
)

The other models are BiotoolsModel, CitationModel, CodeMetaModel, ZenodoModel, LicenseModel, ChangelogModel, CodeOfConductModel, ContributingModel, GovernanceModel, SecurityModel, SupportModel, PullRequestTemplateModel, and IssueTemplateModel.

All inherit from FileTemplateModel, which provides JSON loading, JSON writing, and render(). Use render_many() for independent files rendered into one directory.

License text

LicenseModel.render() creates LICENSE. SPDX identifiers resolve to their license text. Multiline custom terms and unrecognized single-line terms are used verbatim.

Archive models

Most models render one template to one output file. IssueTemplateModel writes the multi-file GitHub issue-template set to .github/ISSUE_TEMPLATE.zip.