This discussion pretty much proves my point. It's all too damn
complicated. Perl scripts, sysread, bypass C library, etc.
when it would be a complete no-brainer if the dump file format
were a 'text' based one - you know, capable of being read by
any line oriented tool. Oh well.
roland
"Dale Worley"<dworley@pingtel.com> writes:
> Well, I just loaded a 115M test file into my Emacs. But yes, in general,
> that's not a particularly good solution. My choice would be to use Perl,
> disable all the complex character set interpretation and use "sysread" and
> "syswrite" to access the file without interference from the C library. Then
> do some minimal level of parsing, so parts of versioned files that look like
> control information get skipped over.
>
> Dale
>
> -----Original Message-----
> From: roland@kanaha.am.mot.com [mailto:roland@kanaha.am.mot.com]On
> Behalf Of Roland Besserer
> Sent: Thursday, December 30, 2004 2:24 PM
> To: users@subversion.tigris.org
> Subject: Re: SVN Book Method for Splitting Repos doesn't work
>
>
>
> Using GNU sed solves the problem. Loading it in either Emacs or vim
> wasn't really an option as the dump file is 115MB and the database
> is expected to grow and I really want to process these huge dump files
> with stream oriented tools, not an editor :-)
>
> roland
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Thu Dec 30 23:48:10 2004