Windows 7.
Bash shell.
I've made a file called pre-commit with the following content but nothing happens.
#!/bin/bash
echo "PRE-COMMIT EXECUTE"
mvn compile
exit 1
First: I have seen references to .git/hooks and C:/GitInstalledDir/hooks for the file location so where exactly is this file suppose to go?
Second: If it's to go in ".git" what directory is this, my user directory?
Related
I've noticed that this will open a file with no project context:
$ code /path/to/some/file
and this will open the project context, but no file:
$ code /path/to/some/dir/
I want to write a bash function that does something like this:
vscode()
{
pushd $(dirname $1)
PROJECT_ROOT=$(git rev-parse --show-toplevel)
popd
code --file $1 \
--dir $PROJECT_ROOT
}
But that's not actually how vscode's command line interface works. Is there some other way to launch vscode from the command line so that it knows about both the project and the file when it opens?
Ideally the chosen project folder would always be whichever git root is the file's most immediate parent.
This is one way to do it
code /path/to/file --add /path/to/folder/
The --add argument adds the folder to the current workspace.
EDIT: Tested and works on Windows 10 Build 19042.746, Code Version 1.52.1, x64.
How can I change my working directory in bash script in windows. I have
~dp0 = C:\test\docker\windows and I want to change my directory to C:\test\build
So it means 2 levels up and then int o build folder
Thanks
Since C:\ is mounted by default into /mnt/c this will work.
Create a .bashrc in your home path by following command:
echo "BUILDDIR=/mnt/c/test/build" >> ~/.bashrc;source ~/.bashrc
cd $BUILDDIR
# Do your work below for example ./configure.
./configure
On my system in Git bash the C:\ root is just /c/ and other dirs from there are whatever they are, so it would be cd /c/test/build/.
You could also still say cd ../../build/.
Good luck.
To change the directory using a bash script is just like you would using normal bash.
cd "C:/test/build"
echo "You're now in the folder, do what you will."
Save the file as .sh and it can be used as such.
Note, when navigation your folders using a bash script remember the directory you'll be starting from is always the home directory.
I have the following code to checkout in a working directory in the hook post-receive:
#!/bin/sh
git --work-tree=d:/websites/__gitweb --git-dir=d:/_gitrepo.git/ checkout -f
Unfortunately it doesn't seem to work. However, the command does work when I enter just this in the windows command line (cms):
git --work-tree=d:/websites/__gitweb --git-dir=d:/_gitrepo.git/ checkout -f
I have checked the permissions and the executing attributes but nothing.
UPDATE:
I think I'm getting closer. Now I know what the problem is but I don't know why is this happening. The hook is actually being triggered but I receive this message:
remote: Starting copy from repository to work tree...
remote: fatal: Not a git repository: 'd:/_gitrepo.git/'
remote: Finished.
I have tried to change the path of d: to the whole network path but it still doesn't work. If I go to the remote repository and I do a git log, the changes are there and if I run the hook with sh, it works.
Why is it saying that it is not a git repository when clearly it is?
I finally got it working!
This is really weird. I had to type a pwd to see where actually is the batch being located and it showed the hook location on the server. However, in the next line I added a hostname and it showed me my local hostname.
Then I add the whole path just like:
#!/bin/sh
echo "Starting copy from repository to work tree..."
pwd
hostname
git --work-tree=//remotehost/d$/Webseiten/__gitweb --git-dir=//remotehost/d$/_gitr
epo.git checkout -f
echo "Finished."
I hope this solution works for someone
For a shell script (and those hook scripts will be executed as shell, even in Windows, through the msys layer of msysgit), you could use a different sort of path:
#!/bin/sh
git --work-tree=/d/websites/__gitweb --git-dir=/d/_gitrepo.git/ checkout -f
See also other possibilities with "Posix path conversion in MinGW"
I'm just starting to look into Git hooks, but I can't seem to get them to run.
I set up a local repository, so there is now a '.git' directory in my project folder. I have added a '.cmd' file into the C:/path/to/my/project/.git/hooks directory named 'pre-commit.cmd'. Here is the contents of this file:
echo "HOOK RUNNING"
echo. 2>C:/path/to/my/project/.git/hooks/EmptyFile.txt
This should echo the text "HOOK RUNNING" and create an empty text file in that directory. However, if I commit changes through my IDE (NetBeans) or use Git Bash to commit, neither of them seem to run my pre-commit hook, as no file is created.
My understanding is that all you have to do to get a hook to run is add an executable with the name of the hook (as I have done). Am I doing something wrong?
Note: This is on a Windows 7 PC.
Name your hook pre-commit (without any file extension).
And add #!/bin/sh on the first line or #!/bin/bash.
You probably don't have the permissions to run the pre-commit file
Run in your terminal:
chmod +x .git/hooks/pre-commit
Thanks to #vaughan for giving the idea
TL;DR
Git hooks work on Git for Windows by default assuming the Git hook script is simple.
Background of Git and Windows
Please Note: Git was made for shell interpretation; thus, using Git hooks on a Windows command prompt or Windows-made PowerShell will inherently have its flaws, and complete interoperability is not to be expected.
Using Git hooks on Windows is possible, but it has many drawbacks.
Git for Windows uses a shell emulator that makes Bash and shell commands possible. This means that when a Git hook is activated by Git, the Windows version will run the command using the shell emulator. Which in turn, will convert those shell commands to commands that the Windows operating system can understand. Meaning, simple shell scripts will work right off the bat. For example, the Git hook pre-commit that ships with an initialization of a Git repository can be run without any modification.
Example of default behavior
Initialize a Git repository with the command git init
Navigate to the Git hooks directory with the command cd .git\hooks
This directory holds all the Git hook scripts. Create a file named pre-commit. Note:
The name of the file is important
Replace the contents with the following shell script
#!/bin/sh
echo "Hello, World!"
Navigate back to your root directory of the project and create a file named test.txt using the command echo "something" > text.txt
Stage the file to commit using the command git add test.txt
Commit the change and watch the pre-commit hook activate using the command git commit -m "test commit"
Verify the output to look like the following
git commit -m "test commit"
Hello, World!
[master f00ccea] test commit
Example of bad behavior
When using a very advanced shell script to do things in Git hooks, Windows shell interpretation doesn't always stack up. For example, when using the Husky Git hook plugin for NPM, along with the Prettier formatter, the commands do not map 1-1. Meaning that your pre-commit Git hook will fail on Windows.
Answering user1578653's question
A Git hook is an executable script; however, you are using a command prompt script (.cmd) and not a shell script (.sh). If you would like this behavior you described on a Windows operating system then create the file named pre-commit and place it in the .git\hooks directory (relative to the project you are working on). Then place the following content in that file.
.git\hooks\pre-commit
#!/bin/sh
echo "HOOK RUNNING"
thisCausesError 2> .git/hooks/EmptyFile.txt
You will notice that the Git hook works and outputs the words HOOK RUNNING to the console, and the thisCauseError will print an error message to standard error that will be printed to the file EmptyFile.txt.
In my case, I had set core.hooksPath to the wrong directory.
Resetting it with git config --global core.hooksPath '~/.githooks' solved the issue :)
You can verify your hooks path using git config --get core.hooksPath
For me, I tried to run a .bat file.
I discovered that backslashes need to be escaped:
For example:
#!/bin/bash
C:\\somefolder\\somefile.bat
For me, none of the previous solutions worked. I moved the pre-commit file from hooks to some other location, effectively deleting the file from hooks.
That worked for me :D
In my case, where I did npm install and accidentally deleted the .git folder, npm install pre-commit --save worked.
If it helps anyone:
I was getting the following error:
error: cannot spawn .git/hooks/pre-commit: No error
It turned out that in my pre-commit file I did not have 'newline' character after last exit command:
#!/bin/sh
# From gist at https://gist.github.com/chadmaughan/5889802
# Stash any unstaged changes
git stash -q --keep-index
# Run the tests with the Gradle wrapper
./gradlew test --daemon
# Store the last exit code in a variable
RESULT=$?
# Unstash the unstashed changes
git stash pop -q
# Return the './gradlew test' exit code
exit $RESULT
# << must have a newline after the above command >>
I was running a Gradle project on Windows and Gradle commands in the Cmder shell and cmd.
I tried solutions suggested in other answers and it didn't help to fix this problem:
cannot spawn .git/hooks/pre-commit: No such file or directory
The solution, which worked for me, was to rename the file .git/pre-commit.sample to .git/pre-commit and insert the script for formatting changed files with Prettier. The file with the name 'pre-commit' which I have created manually must have had some problems (encoding or end-line symbols remains unclear).
I have some script and I have no idea what it is doing, will be happy if somebody will explain me:
#!/bin/tcsh
if (-d test) then
svn up test
else
svn checkout http:some address test
endif
cd tests
python test_some.py $argv
P.S can't find info about functions cd and svn
thanks in advance for any help
The script runs a second revision-controlled test script
This script runs a python program which appears to run some tests. The script understands that the test directory is stored in a subversion repository.
If there is a test directory, it updates it in case it has been changed in the repository, perhaps by another svn user or by the same user in a different working directory.
If there is no test directory, it checks it out.
Then it changes its current directory to the working directory.
Then it runs the test script.
I'm a bit confused about one thing. It checks out "test" but then changes its directory to "tests". So either there is a transcription error in the original post or something slightly more complex is going on, like, it somehow assumes that tests exists but not test.
cd is the "Change Directory" command.
svn is a source code repository client.
The script does the following:
if the test folder exists
update it through subversion
else
check it out from subversion repository
go into the tests directory // interestingly enough, it doesn't match the checked out directory name?
run the test_some.py python file, passing the script arguments.
cd, and svn, and python are executable names. cd is the command for changing current directory. svn is the command (executable name) for Subversion source control system. python is the Python language interpreter.