Bash script to automate the Git Flow tag/release process

release.sh

#!/bin/bash

# current Git branch
branch=$(git symbolic-ref HEAD | sed -e 's,.*/\(.*\),\1,')

# v1.0.0, v1.5.2, etc.
versionLabel=v$1

# establish branch and tag name variables
devBranch=develop
masterBranch=master
releaseBranch=release-$versionLabel
 
# create the release branch from the -develop branch
git checkout -b $releaseBranch $devBranch
 
# file in which to update version number
versionFile="version.txt"
 
# find version number assignment ("= v1.5.5" for example)
# and replace it with newly specified version number
sed -i.backup -E "s/\= v[0-9.]+/\= $versionLabel/" $versionFile $versionFile
 
# remove backup file created by sed command
rm $versionFile.backup
 
# commit version number increment
git commit -am "Incrementing version number to $versionLabel"
 
# merge release branch with the new version number into master
git checkout $masterBranch
git merge --no-ff $releaseBranch
 
# create tag for new version from -master
git tag $versionLabel
 
# merge release branch with the new version number back into develop
git checkout $devBranch
git merge --no-ff $releaseBranch
 
# remove release branch
git branch -d $releaseBranch

Hi @JoneSabino,
A lot of programs out there have one or more files that have the version number in them. For instance, most Unix programs you can run "--version" or "-V" and they will print out the current version. For bclinkinbeard's case, it appears he has this stored in a file named 'version.txt'. This could be convenient for a program to quickly query if needed, but I bet there are a lot of other schemas people use to store the version number (if they do this at all)

Thanks for sharing.

isn't simpler to do
echo "$versionLabel" > $versionFile
instead of

# find version number assignment ("= v1.5.5" for example)
# and replace it with newly specified version number
sed -i.backup -E "s/\= v[0-9.]+/\= $versionLabel/" $versionFile $versionFile 
# remove backup file created by sed command
rm $versionFile.backup