nyc-diff

Make sure every addition and modification is covered with nyc-diff by reporting coverage for changed lines from a unified diff format.

Installation

npm i -g nyc-diff

Usage

You can pipe any unified diff output into the nyc-diff via stdin. An example of usage with git-diff is shown below:

git diff -w <commit-ish>...<commit-ish> | nyc-diff [options]

Output

A JSON file named nyc-diff-<unix timestamp> will be generated when the diff is calculated succesfully in the format shown below. The path of the file can be modified as well by modifying the --output option.

[
	[<path from git diff>, <array of line numbers>],
	... and more
]

Options

--project-dir

This is the location in the local file system where code was instrumented by nyc. In most common situations this would be the root of your repository. Otherwise, wherever the .nyc_output folder of your nyc run is located.

--output / -o

Path at which the resultant JSON will be written to as .nyc-diff-\<unix-timestamp\>. This defaults next to the same directory as .nyc_output.

Limitations

1. Currently only supports checking coverage of individual statements and not the coverage of their branches.
2. Can only be used with pipe and does not support cli args.
3. This tool is built around nyc v14.1.1. This tool's functionality is subject to nyc's output format. This might be changed in future versions of nyc, to ensure complete compatibility remain on the same major version.

Motivation

Working towards 100% coverage can be a troublesome task, especially at SaaS and product businesses where development on a code base never stops. nyc-diff was created to support development teams so as to ensure that any code that has been added or modified has unit-tests written for it.

Credits

Idea was all @KlentyBoogi.