A simple bash script to automatically generate a
CHANGELOG.md
file with every git commit.
Copy the file post-commit
into the directory .git/hooks/
of your git project and make it executable.
chmod +x post-commit
cp post-commit ${YOUR_GIT_PROJECT_PATH}/.git/hooks
This hook is only triggered locally, so if you use this with your team, everybody has to store the hook inside his or her local copy, since the .git
directory is not pushed or tracked with the files from your repository.
To learn more about the .git
folder and what it really contains, I suggest you the brief explanation of the various files and directories inside the .git
directory.
You can change the name of the generated CHANGELOG file inside the post-commit
file by modifying the variable OUTPUT_FILE
to whatever you want. Some projects also use HISTORY.md
, NEWS.md
or RELEASE.md
, but I suggest you stick to the convention.
The default setting of the script generates markdown, which can be seen in the CHANGELOG.md.
If you want to change the output format, you should read the Git documentation for pretty-formats
and change the --format
argument in post-commit
file according to your desired format.
You can test your output format by executing the following command inside your git repository folder:
git --no-pager log --no-merges --format="YOUR_FORMAT_GOES_HERE"
If you are happy with the result, just replace the command in the post-commit
file with your command.
Don't forget to pipe the output in you're CHANGELOG with > $OUTPUT_FILE
at the end of the line.
The script uses Git Hooks to execute the git log
command after every commit and write the output inside the CHANGELOG.md
file.
Since we want to add the changes in our CHANGELOG.md
as well, we have to use Git's --amend
feature to change the previous commit and apply the changes from the script, too.
By changing the last commit, we are triggering the post-commit
hook another time, since the --ammend
commit is a commit as well. To prevent an infite execution of this hook, we have to check, whether our generated CHANGELOG.md
have have changed at all. If it is executed the second time, the file should not be different.
We can check this by using the git status
command with --porcelain
and parse the output for our $OUTPUT_FILE
. If the process returns a code higher than zero, we had a match and know that we have a changed CHANGELOG.md
, which we have to append to our previous commit.
To understand the idea behind a change log, I highly suggest reading Keep a CHANGELOG and / or listening to the episode 127: Keep a CHANGELOG with Olivier Lacan.
It says
A change log is a file which contains a curated, chronologically ordered list of notable changes for each version of an open source project.
Even dough, this matches partly the generated change log by this script, since there are changes in chronologically order, the section about What’s the point of a change log? does not fit at all.
To make it easier for users and contributors to see precisely what notable changes have been made between each release (or version) of the project.
This script stupidly dumps all the commit logs into a single file, which doesn't help anybody in any way.
You should realy consider if this is the right way to keep a changelog for your project and what the benefits are by using this script, despite the fact that someone does not have to execute git log
by itself in the command line.
Because log diffs are full of noise — by nature. They could not make a suitable change log even in a hypothetical project run by perfect humans who never make typos, never forget to commit new files, never miss any part of a refactoring. The purpose of a commit is to document one atomic step in the process by which the code evolves from one state to another. The purpose of a change log is to document the noteworthy differences between these states.
As is the difference between good comments and the code itself, so is the difference between a change log and the commit log: one describes the why, the other the how.