Detail change after Git pull

问题

After a Git pull, its output gives a summary on the change amount.

How can I see each or some of the files detailed changes?

Okay, here is my question to Jefromi:

How do I know if I was pulling to master? All I did is "git pull".
What does master point to and what is the difference between master and HEAD, the two default heads of Git?
How do I see the detailed change in a specific file?
How do I see the change in the summary output by the last git pull again?
What's difference between git diff and git whatchanged?

回答1:

Suppose you're pulling to master. You can refer to the previous position of master by master@{1} (or even master@{10.minutes.ago}; see the specifying revisions section of the git-rev-parse man page), so that you can do things like

See all of the changes: git diff master@{1} master
See the changes to a given file: git diff master@{1} master <file>
See all the changes within a given directory: git diff master@{1} master <dir>
See the summary of changes again: git diff --stat master@{1} master

As for your question of "how do I know if I'm on master"... well, using branches is an important part of the Git workflow. You should always be aware of what branch you're on - if you pulled changes, you want to pull them to the right branch! You can see a list of all branches, with an asterisk by the currently checked-out one, with the command git branch. The current branch name is also printed along with the output of git status. I highly recommend skimming the man pages of commands to use - it's a great way to slowly pick up some knowledge.

And your last question: HEAD is the name for the currently checked out branch. You can indeed use HEAD and HEAD@{1} in this context as well, but it's a bit more robust to use the branches, since if you go and check out another branch. HEAD is now that second branch, and HEAD@{1} is now master - not what you want!

To save having to ask a lot of little questions like this, you should probably have a look at a Git tutorial. There are a million on the web, for example:

The Pro Git book
Git Magic
and the 4.5 million hits on Google for "Git tutorial"

回答2:

Say you do a git pull like this:

$ git pull
remote: Counting objects: 10, done.
remote: Compressing objects: 100% (6/6), done.
remote: Total 6 (delta 4), reused 0 (delta 0)
Unpacking objects: 100% (6/6), done.
From git@dev.example.com:reponame
   a407564..9f52bed  branchname   -> origin/branchname
Updating a407564..9f52bed
Fast forward
 .../folder/filename          |  209 ++++++++-----
 .../folder2/filename2        |  120 +++++++++++---------
 2 files changed, 210 insertions(+), 119 deletions(-)

You can see the diff of what changed by using the revision numbers:

$ git diff a407564..9f52bed

回答3:

1. How do I know if I was pulling to master? All I did is "git pull".

The command itself works like this:

git pull [options] [<repository> [<refspec>…]]

and per default refers to the current branch. You can check your branches by using

git branch -a

This will list your local and remote branches like for e.g so (Added a --- as divider between local and remote to make it more clear)

*master
foo
bar
baz
---
origin/HEAD -> origin/master
origin/deploy
origin/foo
origin/master
origin/bar
remote2/foo
remote2/baz

When you then take a look at one remote repo, you will see what you are referring to:

git remote show origin

will list like the following:

* remote origin
  Fetch URL: ssh://git@git.example.com:12345/username/somerepo.git
  Push  URL: ssh://git@git.example.com:12345/username/somerepo.git
  HEAD branch: master
  Remote branches:
    foo    tracked
    master tracked
  Local refs configured for 'git push':
    foo    pushes to foo    (up to date)
    master pushes to master (fast-forwardable)

So it's quite easy to be sure where to pull from and push to.

3. how to see the detail change in a specific file?

4. how to see the change in summary output by last git pull again?

The easiest and most elegant way (imo) is:

git diff --stat master@{1}..master --dirstat=cumulative,files

This will give you two blocks of information about the changes in between your last pull an the current state of work. Example output (I added a --- as divider between --stat and --dirstat output to make it more clear):

 mu-plugins/media_att_count.php                     |  0
 mu-plugins/phpinfo.php                             |  0
 mu-plugins/template_debug.php                      |  0
 themes/dev/archive.php                             |  0
 themes/dev/category.php                            | 42 ++++++++++++++++++
 .../page_templates/foo_template.php                |  0
 themes/dev/style.css                               |  0
 themes/dev/tag.php                                 | 44 +++++++++++++++++++
 themes/dev/taxonomy-post_format.php                | 41 +++++++++++++++++
 themes/dev/template_parts/bar_template.php         |  0
 themes/someproject/template_wrappers/loop_foo.php  | 51 ++++++++++++++++++++++
---
 11 files changed, 178 insertions(+)
  71.3% themes/dev/
  28.6% themes/someproject/template_wrappers/
 100.0% themes/
  27.2% mu-plugins/
   9.0% themes/dev/page_templates/
   9.0% themes/dev/template_parts/
  63.6% themes/dev/
   9.0% themes/someproject/template_wrappers/
  72.7% themes/

回答4:

This way's kind of hacky, but it'll allow you to use graphical tools like gitk or gitg or git-gui:

git pull
git reset HEAD@{1}
gitg (or gitk or whatever tool you like)

The answer with the most upvotes gives the best way using the git tool, but I use this method because I can then utilize tools with GUI to see the changes :P

I'd then have the extra step of doing a git checkout . and then doing git pull again so that I properly pull and merge, but I value the ability to examine differences in a GUI enough to deal with the extra two steps.

来源：https://stackoverflow.com/questions/1362952/detail-change-after-git-pull

标签

git

git-pull