Git¹

Low Level Commands

Git has a set of commands that translate between the git hash-based organization of content and the repository file system. This is possible because of the magical one-to-one correspondence between a file’s contents and its SHA1 key.

1. To view the contents of a blob in the repository from its SHA1 key:

    $ git cat-file -p <SHA1-key>
   

2. The inverse lookup, to find a SHA1 key from a tag, file-name? etc.

    $ git rev-parse <name>
   

3. To run the SHA1 hash on content

    $ git hash-object <file-name>
   

4. To see the correspondence in the index between files and their keys.

    $ git ls-files --stage
   

Day-to-day commands are composed of lower-level commands such as these that are exposed to create git objects.

1. To create a tree node from the current index:

    $ git write-tree 
   

2. The SHA1 key returned for the tree is used to write a commit node to stdout:

    $ git commit-tree git write-tree -m"Here is a commit node"
   

3. Human readable git objects:

    $ git show <some-object>
   

Useful obscurities

1. The inverse of git add, to unstage a file (remove it from the index) without reverting to the previous committed version:

    $ git rm --cached <file-name>
   

In general the --cached switch applies the command to the index. Along these lines, this shows how to restore .gitignore to its proper function

	$ git rm -r --cached .
	$ git add .
	$ git commit -m"fixed gitignore"

1. To see the sequence in the log on a file that has been moved or renamed:

    $ git log --follow <file-name>