Exploring History

Overview

Teaching: 25 min
Exercises: 0 min

Questions

• How can I identify old versions of files?

• How do I review my changes?

• How can I recover old versions of files?

Objectives

• Explain what the HEAD of a repository is and how to use it.

• Identify and use Git commit numbers.

• Compare various versions of tracked files.

• Restore old versions of files.

As we saw in the previous episode, we can refer to commits by their identifiers. You can refer to the most recent commit of the working directory by using the identifier HEAD.

We’ve been adding one line at a time to mars.txt, so it’s easy to track our progress by looking, so let’s do that using our HEADs. Before we start, let’s make a change to mars.txt, adding yet another line.

$ nano mars.txt
$ cat mars.txt

Cold and dry, but everything is my favorite color
The two moons would make for interesting tides, if the planet had oceans
The lack of humidity is good for my hair
An ill-considered change

Now, let’s see what we get.

$ git diff HEAD mars.txt

diff --git a/mars.txt b/mars.txt
index b36abfd..0848c8d 100644
--- a/mars.txt
+++ b/mars.txt
@@ -1,3 +1,4 @@
 Cold and dry, but everything is my favorite color
 The two moons would make for interesting tides, if the planet had oceans
 The lack of humidity is good for my hair
+An ill-considered change.

which is the same as what you would get if you leave out HEAD (try it). The real goodness in all this is when you can refer to previous commits. We do that by adding ~1 (where “~” is “tilde”, pronounced [til-duh]) to refer to the commit one before HEAD.

$ git diff HEAD~1 mars.txt

If we want to see the differences between older commits we can use git diff again, but with the notation HEAD~1, HEAD~2, and so on, to refer to them:

$ git diff HEAD~2 mars.txt

diff --git a/mars.txt b/mars.txt
index df0654a..b36abfd 100644
--- a/mars.txt
+++ b/mars.txt
@@ -1 +1,4 @@
 Cold and dry, but everything is my favorite color
+The two moons would make for interesting tides, if the planet had oceans
+The lack of humidity is good for my hair
+An ill-considered change

We could also use git show which shows us what changes we made at an older commit as well as the commit message, rather than the differences between a commit and our working directory that we see by using git diff.

$ git show HEAD~2 mars.txt

commit f22b25e3233b4645dabd0d81e651fe074bd8e73b
Author: FIRST_NAME LAST_NAME <your_email@example.com>
Date:   Thu Aug 22 09:51:46 2013 -0400

    Start notes on Mars as a base

diff --git a/mars.txt b/mars.txt
new file mode 100644
index 0000000..df0654a
--- /dev/null
+++ b/mars.txt
@@ -0,0 +1 @@
+Cold and dry, but everything is my favorite color

In this way, we can build up a chain of commits. The most recent end of the chain is referred to as HEAD; we can refer to previous commits using the ~ notation, so HEAD~1 means “the previous commit”, while HEAD~123 goes back 123 commits from where we are now.

We can also refer to commits using those long strings of digits and letters that git log displays. These are unique IDs for the changes, and “unique” really does mean unique: every change to any set of files on any computer has a unique 40-character identifier. Our first commit was given the ID f22b25e3233b4645dabd0d81e651fe074bd8e73b, so let’s try this:

$ git diff f22b25e3233b4645dabd0d81e651fe074bd8e73b mars.txt

diff --git a/mars.txt b/mars.txt
index df0654a..93a3e13 100644
--- a/mars.txt
+++ b/mars.txt
@@ -1 +1,4 @@
 Cold and dry, but everything is my favorite color
+The two moons would make for interesting tides, if the planet had oceans
+The lack of humidity is good for my hair
+An ill-considered change

That’s the right answer, but typing out random 40-character strings is annoying, so Git lets us use just the first few characters (typically seven for normal size projects):

$ git diff f22b25e mars.txt

diff --git a/mars.txt b/mars.txt
index df0654a..93a3e13 100644
--- a/mars.txt
+++ b/mars.txt
@@ -1 +1,4 @@
 Cold and dry, but everything is my favorite color
+The two moons would make for interesting tides, if the planet had oceans
+The lack of humidity is good for my hair
+An ill-considered change

All right! So we can save changes to files and see what we’ve changed. Now, how can we restore older versions of things? Let’s suppose we change our mind about the last update to mars.txt (the “ill-considered change”).

git status now tells us that the file has been changed, but those changes haven’t been staged:

$ git status

On branch main
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:   mars.txt

no changes added to commit (use "git add" and/or "git commit -a")

We can put things back the way they were by using git checkout:

$ git checkout HEAD mars.txt
$ cat mars.txt

Cold and dry, but everything is my favorite color
The two moons would make for interesting tides, if the planet had oceans
The lack of humidity is good for my hair

As you might guess from its name, git checkout checks out (i.e., restores) an old version of a file. In this case, we’re telling Git that we want to recover the version of the file recorded in HEAD, which is the last saved commit. If we want to go back even further, we can use a commit identifier instead:

$ git checkout f22b25e mars.txt

$ cat mars.txt

Cold and dry, but everything is my favorite color

$ git status

On branch main
Changes to be committed:
  (use "git reset HEAD <file>..." to unstage)

    modified:   mars.txt

Notice that the changes are currently in the staging area. Again, we can put things back the way they were by using git checkout:

$ git checkout HEAD mars.txt

Don’t Lose Your HEAD

Above we used

$ git checkout f22b25e mars.txt

to revert mars.txt to its state after the commit f22b25e. But be careful! The command checkout has other important functionalities and Git will misunderstand your intentions if you are not accurate with the typing. For example, if you forget mars.txt in the previous command.

$ git checkout f22b25e

Note: checking out 'f22b25e'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by performing another checkout.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -b with the checkout command again. Example:

 git checkout -b <new-branch-name>

HEAD is now at f22b25e Start notes on Mars as a base

The “detached HEAD” is like “look, but don’t touch” here, so you shouldn’t make any changes in this state. After investigating your repo’s past state, reattach your HEAD with git checkout main.

It’s important to remember that we must use the commit number that identifies the state of the repository before the change we’re trying to undo. A common mistake is to use the number of the commit in which we made the change we’re trying to discard. In the example below, we want to retrieve the state from before the most recent commit (HEAD~1), which is commit f22b25e:

So, to put it all together, here’s how Git works in cartoon form:

Simplifying the Common Case

If you read the output of git status carefully, you’ll see that it includes this hint:

(use "git checkout -- <file>..." to discard changes in working directory)

As it says, git checkout without a version identifier restores files to the state saved in HEAD. The double dash -- is needed to separate the names of the files being recovered from the command itself: without it, Git would try to use the name of the file as the commit identifier.

The fact that files can be reverted one by one tends to change the way people organize their work. If everything is in one large document, it’s hard (but not impossible) to undo changes to the introduction without also undoing changes made later to the conclusion. If the introduction and conclusion are stored in separate files, on the other hand, moving backward and forward in time becomes much easier.

Key Points

• git diff displays differences between commits.

• git checkout recovers old versions of files.