monkey VS go-txdb

Code that's hard to test is just bad code. And unlike most dynamic languages you can't just override some internals in Go (though there are some arcane hacks you can pull of but only at full moon https://github.com/bouk/monkey).

This person did not read the license of the original library https://github.com/bouk/monkey/blob/master/LICENSE.md

I am a new gopher. I was looking into the Monkey Patching module and it is archived now. I was wondering if there is an alternative for that.

Here is an example of this in Go - but as he says, don't actually do this. https://github.com/bouk/monkey

I discovered this lib a few days ago. https://github.com/bouk/monkey it allows you to monkey patch entire functions, replacing them by whatever you want. Perfect for mocking. It's simple to use. The program is hard patching the code using assembly to replace the function address at runtime. You should not use this lib out of your tests since it's absolutely not safe. It's only compatible with linux and windows. But it works great!

3 years ago I created a internal library where I worked to setup the testing with the db's. One param was the schema directory and the other one was the path of the data insertion file. It worked really well but was a little slow. It was a little bit weird to configure the container to run the tests. As I would have to give access to go container that ran the unit tests access to docker. But back then I was running boot2docker and made things very cumbersome. I think with a local docker things would be way simpler. I ended up using and normal DB from the CI/CD and https://github.com/DATA-DOG/go-txdb If I had to do it again I might do an hybrid in-between the two solutions

The first one is TestMain, which could improve the lifecycle management of the container. The second one is DATA-DOG/go-txdb, this allows you to open a new db connection running in a transaction, so each test could have its own isolated DB connection.