RML Implementation Report

Static implementation report for GitHub Pages using ReSpec and CSV files.

Structure

• dev.html: ReSpec source document
• css/style.css: visual tweaks
• js/app.js: loads module inventories and engine CSV files, or uses embedded snapshot data in docs/index.html
• scripts/export-docs.mjs: exports docs/index.html from dev.html via ReSpec CLI and embeds a data snapshot
• data/modules.csv: list of RML modules and the CSV to fetch test cases from
• data/processors.csv: implementations metadata
• data/testcases/*.csv: example module test-case inventories using the upstream format
• data/results/*.csv: one result file per engine

CSV formats

modules.csv

module_id,module_name,specification_slug,testcases_csv

Each row points to a CSV with the module test cases. By default this project points directly to the raw files in the kg-construct GitHub repositories. The app expects the CSV to include at least:

ID,title,description,specification,...,error

You can keep these CSVs local for testing, but the default setup uses the raw GitHub URLs directly.

processors.csv

processor_id,name,version,release_date,contact,homepage,results_file

results_file is optional. If omitted, the app looks for data/results/<processor_id>.csv.

data/results/.csv

testcase_id,status,notes

You can also keep a processor_id column in each result file; when it is missing, the app fills it from the engine declared in processors.csv.

Allowed status values:

• passed
• failed
• inapplicable

Publish on GitHub Pages

1. Run node scripts/export-docs.mjs.
2. Commit the generated docs/ output.
3. Enable GitHub Pages from docs/.
4. Open the published URL.

The export script uses ReSpec's own HTML export through npx respec, then embeds the resolved data snapshot into docs/index.html so the published page does not fetch CSV files at runtime.

Suggested next step

Replace the example engine CSVs with your real results. If needed, adjust data/modules.csv to point to different branches or repositories.