Linting is a crucial step in the development process to ensure code quality and adherence to coding standards. However, as projects grow, linting checks can become time-consuming, affecting the overall development workflow. GitHub Actions provides a powerful and flexible CI/CD platform, and optimizing your linter checks can significantly improve development speed.

In this article, we’ll explore how to optimize PHP linting checks using GitHub Actions, potentially speeding up the process and it can reach 10x based on the size of your project.

How The Linter Optimize The Running Time

Every linting library has its techniques to optimize the running time of the checks and this can happen in multiple ways such as:-

1. Parsing and Lexical Analysis:

   • Linters quickly scan the code to identify tokens, keywords, and syntax errors. This initial parsing allows them to flag obvious mistakes without delving deeper into semantic analysis.

   • By catching basic syntax errors early, linters avoid wasting time on further processing of invalid code.

2. Heuristics and Static Analysis:

   • Linters utilize pre-defined rules and patterns to detect potential issues like unused variables, code smells, and stylistic inconsistencies. This avoids the need for complex, resource-intensive analysis like full program execution.

   • By focusing on specific rules and avoiding unnecessary calculations, linters can identify many problems efficiently.

3. Incremental Processing:

   • Some linters can track changes in the codebase and only focus on analyzing the modified portions instead of re-evaluating everything from scratch. This significantly reduces processing time for iterative development.

   • By caching previous analysis results and only updating them for changed code, linters can significantly improve runtime.

4. Caching and Pre-computation:

   • Linters can store frequently used data, like rule patterns or language constructs, in memory to avoid repeated lookups and calculations. This reduces the overhead of each analysis run.

   • By caching commonly accessed information, linters can avoid repetitive processing and further enhance efficiency.

5. Asynchronous Processing:

   • Advanced linters can leverage multiple cores or threads to parallelize tasks like parsing and rule evaluation. This helps distribute the workload and speeds up the analysis process for large codebases.

   • Linters can run much faster using parallel processing techniques, especially on modern multi-core systems.

Our main focus in this article will be on the caching mechanism to optimize the running speed and to reduce the github deployment minutes as well.

GitHub Action Setup

This is the typical linter before optimization

This listing check used to take about 8 minutes on average

This is the linter workflow after cache optimization

What we have done here just caching the results of the linting check and using it later to check only the modified files instead of scanning the whole codebase each time.

After this optimization, it takes about 1 minute on average

The key thing in the caching is to know the right path of the thing that you want to cache and this actions/cache will do the rest for you.

References

• https://docs.github.com/en/actions/using-workflows/caching-dependencies-to-speed-up-workflows

• https://github.com/phpstan/phpstan/discussions/9301

• https://cs.symfony.com/doc/usage.html#caching