snog

Experimental hybrid of a snapshot and a log.

How does it work?

First add it to your project with npm install --save snog

Then add a few snogs here and there in your code. Eg.

const snog = require('snog')
 
 
// ... lots of code ...
 
snog({ a, b, c })
 
// ... more code ...
 
snog({ d, e })

Whenever this code runs, it will record a snapshot of whatever you give it (eg. { a, b, c}).

Then boot up the snog-diff viewer to watch the log messages as your code runs:

npx snog-diff --watch

Now, whenever you run your code, we compare the new logs against the recorded ones. So you can make a change and see what changed!

For extra ✨ try running snog-diff in one terminal, and nodemon yourscript.js in another.

Updating the reference

There are 2 sets of files, which are counterparts laid out like this:

__snog__/
├── latest
│   └── <srcdir>
│       └── <srcfile>.js
│           ├── <func>-<line>.txt
│           └── <func>-<line>.txt
└── ref
    └── <srcdir>
        └── <srcfile>.js
            ├── <func>-<line>.txt
            └── <func>-<line>.txt

The files in __snog__/latest represent the logs from the most recent run, and will be compared with the file of the same name in __snog__/ref.

When the snog-diff cli notifies you of a diff, it also asks whether you'd like to update the reference. This is your way of saying "I like what I see". So subsequent runs will be compared against the new reference.

How does this relate to traditional snapshot tests or unit tests?

snog is intended for quick feedback in development. But it's quite easy to start with a snog and turn it into a test. For example, use the first snog to record the args being passed into a function, and the second one to record the result. Once it looks right in your snog-logs, paste it into your test suite.

Does snog have any bugs?

Oh my word yes it does. And there's still heaps of improvements to make. If this project sounds interesting to you I'd love to hear your feedback!

License

MIT