While working on this repository, a couple of issues sprung up on me that required me to modify my current git configuration to make sure that specs run properly:
Bundler depends on the fact that gems generated for the purpose of a test have a master branch. If someone has an init.defaultBranch = main configuration in git, this will cause those temporary repositories with temporary gems to not have a master branch, failing a test.
To create a proper repository, tests create an initial commit using git commit, which launches a pre-commit hook by default. If someone has a core.hooksPath variable set in their global git config, and if they have a global pre-commit hook, it will run on those repositories.
Since we're not testing how git works, and we're merely using it to create entities that we can use as stubs and dummies for testing, we don't need to trigger those features.
This PR edits dependency helpers to make it so that aforementioned local setup differences don't influence Appraisal testing.
Summary
While working on this repository, a couple of issues sprung up on me that required me to modify my current
git
configuration to make sure that specs run properly:master
branch. If someone has aninit.defaultBranch = main
configuration in git, this will cause those temporary repositories with temporary gems to not have amaster
branch, failing a test.git commit
, which launches apre-commit
hook by default. If someone has acore.hooksPath
variable set in their global git config, and if they have a globalpre-commit
hook, it will run on those repositories.Since we're not testing how git works, and we're merely using it to create entities that we can use as stubs and dummies for testing, we don't need to trigger those features.
This PR edits dependency helpers to make it so that aforementioned local setup differences don't influence Appraisal testing.