Execute commands from shell file, storing last successful execution in detached git branch.
You have some bash file(-s) with few commands inside. Each on it's own line. If you will run it in shell -- it will execute all of them.
During deploy you need to run only new commands, that was added to this
file. Think of it as a one way DB migrations that will get commands to
execute from git diff.
Git to the resque! System will store successfull commands executions in
detached branch named .gitmigrate. Then if we will have 5 commands
and only 3 of them was successful, on next run we will not run them.
First lines until two new lines will be always executed. So when you have shebang and then 2 new lines (as it usual) -- it will be executed. Also you can add some includes after your 1st line, that will allow you to have DSL-like functions inside your script. After which you shoudl have two new lines.
Configuration values are in file .gitmigrate.
See .gitmigrate.dist as an example
- Detached branch name
_gitmigrate. - Path to command file(s) is
.gitmigrate.*. It could be both.gitmigrate.shor.gitmigrate.pyor.gitmigrate.d/directory.