Difference between author and committer in Git?

笑着哭i 提交于 2019-11-26 00:36:20

问题


I am trying to make a commit like

git commit --author=\"John Doe <john@doe.com>\" -m \"<the usual commit message>\"

where John Doe is some user in whose name I want to make the commit.

It appears all right in git log. However, when I do a gitk, the author name is correct, but the committer name is picked from my global git config settings (and is thus set to my name/email).

Questions

  1. What is the difference between the two (committer vs author)?

  2. Should I be setting the committer as well to the other user?

  3. If yes, how?


回答1:


The original poster asks:

What is the difference between the two (Committer vs author)?

The author is the person who originally wrote the code. The committer, on the other hand, is assumed to be the person who committed the code on behalf of the original author. This is important in Git because Git allows you to rewrite history, or apply patches on behalf of another person. The FREE online Pro Git book explains it like this:

You may be wondering what the difference is between author and committer. The author is the person who originally wrote the patch, whereas the committer is the person who last applied the patch. So, if you send in a patch to a project and one of the core members applies the patch, both of you get credit — you as the author and the core member as the committer.

The original poster asks:

Should I be setting the committer as well to the other user?

No, if you want to be honest, you should not be setting the committer to the author, unless the author and the committer are indeed the same person.




回答2:


Mailing list + git format-patch + git apply can generate author != committer

In projects like the Linux kernel where patches are:

  • generated by git format-patch
  • sent by email, either by copy pasting, or more commonly with git send-email
  • applied by another person with either git apply or git am: How to use git am to apply patches from email messages?

generating a single new commit with different author and committer:

  • the author is who wrote the patch
  • the committer is who is a project maintainer, and who merged the patch

See for example this randomly selected patch and the corresponding commit:

  • https://lkml.org/lkml/2018/1/25/568
  • https://github.com/torvalds/linux/commit/5beda7d54eafece4c974cfa9fbb9f60fb18fd20a

Git web interfaces like GitHub and GitLab may or may not generate author != committer

Since Git(Hub|Lab) hold both the upstream and the fork repositories on a the same machine, they can automatically do anything that you can do locally as well, including either of:

  • Create a merge commit.

    Does not generate author != committer.

    Keeps the SHA or the new commit intact, and creates a new commit:

    * Merge commit (committer == author == project maintainer)
    |\
    | * Feature commit (committer == author == contributor)
    |/
    * Old master (random committer and author)
    

    Historically, this was the first available method on GitHub.

    Locally, this is done with git merge --no-ff.

    This produces two commits per pull request, and keeps a fork in the git history.

  • rebase on top of master

    GitHub also hacks the commits to set committer == whoever pressed the merge button. This is not mandatory, and not even done by default locally by git rebase, but it gives accountability to the project maintainer.

    The git tree now looks like:

    * Feature commit (committer == maintainer, author == contributor)
    |
    * Old master (random committer and author)    
    

    which is exactly like that of the git apply email patches.

On GitHub currently:

  • you choose the method when merging via the dropdown on the merge button
  • methods can be enabled or disabled on the repo settings by the owner

https://help.github.com/articles/about-merge-methods-on-github/

How to set the committer of a new commit?

The best I could find was using the environment variables to override the committer:

GIT_COMMITTER_NAME='a' GIT_COMMITTER_EMAIL='a' git commit --author 'a <a>'

How to get the committer and commit date of a given commit?

Only author data shows by default on git log.

To see the committer date you can either:

  • format the log specifically for that:

    git log --pretty='%cn %cd' -n1 HEAD
    

    where cn and cd stand for Committer Name and Committer Date

  • use the fuller predefined format:

    git log --format=fuller
    

    See also: How to configure 'git log' to show 'commit date'

  • go low level and show the entire commit data:

    git cat-file -p HEAD
    

How to set the committer date of a new commit?

git commit --date only sets the author date: for the committer date the best I could find was with the environment variable:

GIT_COMMITTER_DATE='2000-01-01T00:00:00+0000' git commit --date='2000-01-01T00:00:00+0000'

See also: What is the difference between author and committer in Git?

How Git stores author vs committer internally?

See: What is the file format of a git commit object?

Basically, the commit is a text file, and it contains two line separated fields:

author {author_name} <{author_email}> {author_date_seconds} {author_date_timezone}
committer {committer_name} <{committer_email}> {committer_date_seconds} {committer_date_timezone}

This makes it clear that both are two completely independent data entries in the commit object.




回答3:


@Ciro Santilli 新疆改造中心 六四事件 法轮功 proposed to use

GIT_COMMITTER_NAME='a' GIT_COMMITTER_EMAIL='a' git commit --author 'a <a>'

To avoid repeating the name and email, you can reuse them

GIT_COMMITTER_NAME='a'; GIT_COMMITTER_EMAIL='a'; git commit --author "$GIT_COMMITTER_NAME <$GIT_COMMITTER_EMAIL>"

which first sets the variables in separate commands, then uses them for the git commit call (note the double parentheses).



来源:https://stackoverflow.com/questions/18750808/difference-between-author-and-committer-in-git

易学教程内所有资源均来自网络或用户发布的内容,如有违反法律规定的内容欢迎反馈
该文章没有解决你所遇到的问题?点击提问,说说你的问题,让更多的人一起探讨吧!