Selective generic text file testing based upon file type

[stevehose]

This request is for improving testing. Currently, integration testing analyzes text files ('.gitignore', '.gitattributes', '.ps1', '.psm1', '.psd1', '.json', '.xml', '.cmd', '.mof', '.md', '.js', '.yml') with several tests such as looking for empty files, Unicode encoding, content with tab characters, and files that do not end with a newline character. A better approach is to select which tests are appropriate for each file type. For example, a .xml file doesn't have to end with a newline and it is irrelevant if it contains a tab character, but it shouldn't be an empty file or be Unicode-encoded. No files should be empty (that I'm aware of).

[PlagueHO]

Thank you for raising this @stevehose. This certainly could be done. I'd recommend if such changes were made we would use a data driven approach (e.g. a PSD1 file defining each file type and the test types that should apply). But we should leave this up for community discussion.

[stevehose]

Awesome! Thanks for the reply. I'll start working on a fork and proposal and we can review it.