I am getting my feet wet with Git and have the following issue:
My project source tree:
/
|
+--src/
+----refs/
+----...
|
+--vendor/
+----...
Here's how to print the complete list of files in the working tree which match patterns located anywhere in Git's multiple gitignore sources (if you're using GNU find
):
$ cd {your project directory}
$ find . -path ./.git -prune -o -print \
| git check-ignore --no-index --stdin --verbose
It will check all the files in the current branch of the repository (unless you've deleted them locally).
And it identifies the particular gitignore source lines, as well.
Git continues to track changes in some files which match gitignore patterns, simply because those files were added already. Usefully, the above command displays those files, too.
Negative gitignore patterns are also matched. However, these are easily distinguishable in the listing, because they begin with !
.
If you're using Windows, Git Bash includes GNU find
(as revealed by find --version
).
If the list is long (and you have rev
), you can display them by extension (somewhat), too:
$ cd {your project directory}
$ find . -path ./.git -prune -o -print \
| git check-ignore --no-index --stdin --verbose \
| rev | sort | rev
For more details, see man find
, man git-check-ignore
, man rev
, and man sort
.
The point of this whole approach is that Git (the software) is changing rapidly and is highly complex. By contrast, GNU's find
is extremely stable (at least, in its features used here). So, anyone who desires to be competitive by displaying their in-depth knowledge of Git will answer the question in a different way.
What's the best answer? This answer deliberately minimizes its reliance on Git knowledge, toward achieving the goal of stability and simplicity through modularity (information isolation), and is designed to last a long time.