Dev list: Yeah I've definitely learned that you might be much better off using the 'create branches for features and delete them when done, or make them public to allow people to help' method rather than the old CVS 'make changes and then a patch' method of development. It's a transition, but I think it works really well once you get used to it. And I'm one of the biggest CVS curmudgeons in Drupal. Dave Reid dave@davereid.net On Mon, Apr 11, 2011 at 4:42 PM, Bob Hutchinson <hutchlists@midwales.com>wrote:
On Monday 11 April 2011, Dave Cohen wrote:
I'm trying, and failing, to create a patch that adds new files to a module.
Here is a simple way to pull together a patch on a drupal project.
Create a new directory, let's call it 'git-sandbox' cd into it and run something like this:
git clone --branch 6.x-3.x http://git.drupal.org/project/mymodule.git
This will create a directory 'mymodule' (or whatever you fetched)
copy your changed files into it and run
git diff > mymodule.patch
if you are adding new files copy them in and run 'add .'
cd in and run
git diff --staged > mymodule.patch
To do both at the same time
git diff HEAD > mymodule.patch
As Sam Boyer explains earlier in this thread.
open the patchfile in your editor and eyeball to see if it makes any sense
If that's OK copy the patchfile away somewhere and delete the whole mymodule directory and run the git clone command again, copy the patchfile back in and run
git apply mymodule.patch or patch -p1 < mymodule.patch
if it applies cleanly you are ready to upload the patch to the issue queue or wherever.
Of course a git clone run using http cannot be used for commits so it is 'safer' as it were.
You can also do all this and a lot more using a tool like Smartgit though many would argue that learning it on the command line is better for understanding it, true enough.
Another *big* help is to install the git docs (as well as php, mysql etc) on your devbox.
Hope this helps.
According to http://drupal.org/node/707484, I should simply run `git add ...` then "Next time you run git diff, these new files will be included
as
well." In my experience, this statement is false.
When I run `git status`, I see something like this:
[dave@starbuck fb-3]$ git status # On branch 6.x-3.x # Changes to be committed: # (use "git reset HEAD <file>..." to unstage) # # new file: fb.process.inc # new file: fb_tab.js # # Changes not staged for commit: # (use "git add <file>..." to update what will be committed) # (use "git checkout -- <file>..." to discard changes in working directory) # # modified: README.txt # modified: contrib/fb_example.module # modified: fb.module # modified: fb_canvas.admin.inc # modified: fb_canvas.module # modified: fb_devel.module # modified: fb_settings.inc # modified: fb_tab.admin.inc # modified: fb_tab.info # modified: fb_tab.module # modified: fb_url_rewrite.inc
When I run `git diff`, I see changes to all the modified files, but I do not see the new files.
What am I missing? Thanks,
-Dave
-- ----------------- Bob Hutchinson Midwales dot com -----------------