Re: branching for 9.2devel

Alvaro Herrera wrote:
> Excerpts from Tom Lane's message of lun abr 25 20:48:42 -0300 2011:
> > Andrew Dunstan <andrew@dunslane.net> writes:
> 
> > > Well, that way you'll have a handful of -Ttypdef parameters for each 
> > > invocation of indent instead of a gazillion of them. No more command 
> > > line length issues.
> > 
> > Well, -Ttypedef is wrong on its face.  Right would be a switch
> > specifying the name of the file to read the typedef list from.
> > Then you don't need massive script-level infrastructure to try
> > to spoonfeed that data to the program doing the work.
> 
> I gather that Andrew will be working on replacing the pgindent shell
> script with a Perl script, but this new script will still rely on our
> patched BSD indent, right?
> 
> Of course, it would make sense to further patch indent so that it
> accepts typedefs in a file instead of thousands of -T switches, but that
> seems orthogonal.

I have done as you suggested, modifying our version of BSD indent to
allow a file of typedefs to be processed.  I also renamed the download
and binary to 'pg_bsd_indent' so it can be installed on a system that
already has 'indent'.  It should propogate to the ftp mirrors in a few
hours.  Even after we go to Perl, this is still a necessary change.

I have modified the pgindent script to use this new flag, and applied
those changes, attached.

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + It's impossible for everything to be true. +