Introduce Readable Readme

A CI/CD pipeline to help maintain high-quality docs

Tianhao Zhou
Readable Readme
3 min readFeb 6, 2020

--

credit: logo from tianhaoz95/readable-readme

In software engineering, code readability is essential for keeping a project maintainable.

To achieve code readability, tools like static analyzer and linter (e.g. TSLint, DeepScan, and DartAnalyzer) were developed.

By contrast, though documentation is equally important, the focus on documentation readability is lacking.

Readable Readme is an NLP based documentation analyzer

To fill in the gap, I built Readable Readme, an automated documentation static analyzer plus linter, and here is how it works:

credit: workflow diagram from tianhaoz95/readable-readme

Readable Readme is built with:

  • GitHub Actions
  • TypeScript
  • Rule-based language analyzer
  • Deep learning-based language analyzer

Readable Readme uses GitHub Action for better reusability

To add Readable Readme into your repository, simply create a .github/workflows/readable-readme.yml with the following content:

Once the workflow configuration above is added, upon workflow triggering (e.g. code pushes, pull requests), Readable Readme will generate a report with issues and suggested fixes:

Clicking on a report expands the full report:

The full report of a documentation file looks like:

As of the writing of this article, Readable Readme supports grammar checking with a rule-based analyzer and sentiment analysis with NLP neural network as shown above.

Let’s make README readable again together!

If you like the idea of Readable Readme:

--

--