Bug 15201 – Enhancement for prepare-ChangeLog script

• Component: WebKit Misc.
• Status: RESOLVED FIXED

[Mike Hommey]

I made some enhancement for prepare-ChangeLog script:
- Properly parse GECOS field.
- Use git configuration for user name and email when appropriate.

[Mike Hommey]

Created attachment 16276 [details]
Proposed patch

[Adam Roben (:aroben)]

Comment on attachment 16276 [details]
Proposed patch

+    my $_name = `$GIT config user.name`;
+    chomp $_name;
+    $name = $_name if $_name;

I'm not sure it's right for a git config variable to override the REAL_NAME and EMAIL_ADDRESS environment variables. Perhaps we should try to check this information before checking the environment variables?

[Mike Hommey]

(In reply to comment #2)
> I'm not sure it's right for a git config variable to override the REAL_NAME and
> EMAIL_ADDRESS environment variables. Perhaps we should try to check this
> information before checking the environment variables?

I'm not sure either, but I, for one, would expect the most specific configuration to apply. While the environment variables are more specific than the GECOS entry, the git configuration is more specific than the environment variables.

I would actually be bothered if environment variables, that are global to many applications, would override the preferences I specifically use for git.

[Adam Roben (:aroben)]

(In reply to comment #3)
> (In reply to comment #2)
> > I'm not sure it's right for a git config variable to override the REAL_NAME and
> > EMAIL_ADDRESS environment variables. Perhaps we should try to check this
> > information before checking the environment variables?
> 
> I'm not sure either, but I, for one, would expect the most specific
> configuration to apply. While the environment variables are more specific than
> the GECOS entry, the git configuration is more specific than the environment
> variables.
> 
> I would actually be bothered if environment variables, that are global to many
> applications, would override the preferences I specifically use for git.

Well, the nice thing about environment variables is that they're easy to change for a single invocation:

EMAIL_ADDRESS=[e-mail] prepare-ChangeLog

It's much more of a hassle to change a git config variable for a one-time thing like that.

[Mike Hommey]

> Well, the nice thing about environment variables is that they're easy to change
> for a single invocation:
> 
> EMAIL_ADDRESS=[e-mail] prepare-ChangeLog
> 
> It's much more of a hassle to change a git config variable for a one-time thing
> like that.

Mmmmm you got a point here.

[Mike Hommey]

Created attachment 16280 [details]
New patch

Here you are

[Adam Roben (:aroben)]

Comment on attachment 16280 [details]
New patch

+    $result;

We normally use the "return" keyword in our scripts, even though it isn't strictly necessary.

I like this fix!

I think this could be committed as-is, or the committer can add the "return" as I said above.

r=me

[David Kilzer (:ddkilzer)]

Committed revision 25616.

Added 'return' gitConfig() subroutine.  Also added a subroutine declaration at the top of the file and moved the gitConfig() sub out of the main body of the script and into the other subroutines.