Improve -m to be more efficient, only lookup the current rev if it is pinned.

Improve -m to be more efficient, only lookup the current rev if it is pinned.
Use svn up --force if -m or -f is used.
Renamed AddAdditionalFlags to _AddAdditionalUpdateFlags to be clearer about its purpose.
Strings cleanup.

With this change, it becomes more manageable to use -m on the continuous build.

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=56394

[M-A Ruel, 2010-08-16 17:23:16]

That would have greatly reduced the downtime this morning. :/

[nsylvain, 2010-08-17 14:20:34]

http://codereview.chromium.org/3185006/diff/1/2
File gclient_scm.py (right):

http://codereview.chromium.org/3185006/diff/1/2#newcode951
gclient_scm.py:951: new_command.append('--force')
but i don't want to use force on my machine either. 

I like when it collapses the output and shows me only what really needs to be
updated.

Who wants --force? Why? Can't we specify --force if we want it?

[M-A Ruel, 2010-08-17 14:33:28]

svn up --force fixes issues when an old directory still exists and a new DEPS is
added at the same place, it happened in the past on the continuous builds and
required a full source clobber if --force wasn't used.

http://codereview.chromium.org/3185006/diff/1/2
File gclient_scm.py (right):

http://codereview.chromium.org/3185006/diff/1/2#newcode951
gclient_scm.py:951: new_command.append('--force')
On 2010/08/17 14:20:34, nsylvain wrote:
> but i don't want to use force on my machine either. 
> 
> I like when it collapses the output and shows me only what really needs to be
> updated.
> 
> Who wants --force? Why? Can't we specify --force if we want it?

I think you confuse "gclient sync --force" and "svn up --force". Here it's "svn
up --force". It's not affecting the collapse. Anyway I've changed this to not be
always be the case.

[nsylvain, 2010-08-17 14:34:50]

oh, i see.

lgtm either way