Restore WebFrame::GetPrintPageShrink() method originally added in rev 14639

Restore the WebFrame::GetPrintPageShrink() method originally added in rev 14639
and later removed in rev 19669. From the original commit message:

Rev 12100 changed the way that we spool and render printed output, most notably
introducing use of the PrintContext class.  The existing
PrintContext::spoolPage() method applies a webkit scaling factor before
rendering output to the graphics context.  ChromePrintContext::spoolPage() (in
webframe_impl.cc), which is used by chromium instead of
PrintContext::spoolPage(), does not apply this scaling factor, but instead
eventually returns the scaling factor via WebFrame::PrintPage().  This is a
problem for the Chromium Embedded Framework (CEF) because, unlike chromium, the
CEF renders directly to the printer device context.  It is therefore important
for CEF that we retrieve and apply the webkit scaling factor before calling
PrintPage().  In order to support this capability the following adds a
WebFrame::GetPrintPageShrink() method.

[Marshall Greenblatt, 2009-07-05 01:55:59]

Please review this patch to restore the WebFrame::GetPrintPageShrink() method
that was removed by Darin in rev 19669.  This method is required for the
Chromium Embedded Framework.

darin = primary
maruel = fyi

[M-A Ruel, 2009-07-05 12:25:55]

lgtm

You hit a strange bug in gcl/upload because the file path include the whole
path. I'll get back to you about that.

[Marshall Greenblatt, 2009-07-05 15:12:01]

On 2009/07/05 12:25:55, Marc-Antoine Ruel wrote:
> lgtm
> 
> You hit a strange bug in gcl/upload because the file path include the whole
> path. I'll get back to you about that.

My fault because I specified the full path on 'gcl change'.  I guess gcl could
strip off the path component that matches the SVN root, or I could just pay more
attention next time ;-).  It would be nice, though, if gcl listed files that
already exist in other patch sets under a section similar to how files that
exist in no patch sets are listed.  There have been a number of occasions where
I've had the same file in multiple patch sets.