project-metrics

Overview of projects and their health:

• popular repos
  
• repos without description
  
• repos grouped by language

Commits over time:

• code-vault
  
• fitness-tracker
  
• project-metrics

Repo sizes:

Current features

• pre commit hooks for:

  • Python auto-formatting with black
  • Python import sorting with isort
  • trailing-whitespace
  • end-of-file-fixer
  • check-yaml
  • check-added-large-files

• Statistics and health badges (commit activity, repo size, code test coverage, documentation status, CI health of workflows)

• Barplots grouping repos by size, sectioned into 4 categories (large, medium, small, smallest)

• Linecharts showing commits over time

• Sphinx auto doc, with doc hosting on readthedocs:
  project-metrics

• Automatically checking for updates using Dependabot:

  • pip (monthly)
  • GitHub Actions (weekly)

• GitHub Actions workflow wf.yml.
  NB: as the repo updates itself, the only trigger should be workflow_dispatch to avoid recursive workflow invocation.

• The workflow contains the following jobs:

  • qualify_code:

    • static type checking with mypy
    • code linting with flake8 and ruff
    • spell checking of README
    • linting of YAML files
    • software complexity metrics with wily
    • unit tests with pytest
    • code coverage reporting
    • caching of poetry and mypy dependencies

  • get_metrics:

    • updates this repo's README (links above) with following projects metrics:
      • top 10 most popular GitHub repos, by star count
      • repos lacking any description, across GitHub and GitLab
      • repos grouped by programming language

  • update_repos:

    • updates all GitHub repos:
      • All GitHub repos (including this one) get a repo size badge on top of their README
        NB: does not affect the GitHub profile page, which is a special repo.
      • support for both .md and .rst files
      • All GitHub repos gets a GitHub Actions workflow, .github/workflows/wf.yml, containing linting for Python and/or TypeScript code, as well as a CI badge on their README

the last job, update_repos, is further performance tested and a comparison of parallel and threading versions are provided alongside the serial (for-loop) version:

• Performance Benchmarks

Note on GraphQL

Useful explorer for testing out GraphQL queries:

• GitHub GraphQL API Explorer

Upcoming features

• linting of R scripts with lintr
• cleanup script: remove_badge.py
• setup a single, reusable base workflow (poetry setup etc.)
• identify public repos without any README (e.g. 'web-application-jquery-and-bootstrap')
• security scannings with bandit (searching for API keys etc.)
• filter repos by number of users, descendingly
• getting codecov above 50 %
• improve & extend documentation on readthedocs
• alphabetic sorting of words in config/.wordlist.txt
• group by category (using tags), e.g. health (nutrition, fitness, athlete profiler)
• group by created_at, display newest and oldest repos
• check if repos have tests
• if so, setup codecov badge on their README's
• if not, set up pytest, perhaps using cookiecutter / cruft or custom template