Re: [PATCH] Issue 1074 (Take 3)

John Szakmeister <john@szakmeister.net> writes:

> Here is the next attempt at a patch for Issue 1074 (Need a svnadmin
> command to verify that the repository is not corrupted).
>
> There were a couple of things I was unsure about. I went ahead and added
> subcommand_verify to svnadmin, but it really calls subcommand_dump after
> setting a couple of options.

Make sense. Perhaps a cleaner way would just be to core the guts out
of subcommand_dump into a helper function that is called by both
subcommand_dump and subcommand_verify.

> It sets one new one called 'verify' so that the dump command knows
> to create a NULL dumpstream. Is it okay to have this 'verify' option
> in opt_state, or should I find a different way to do this?

Oh, no, please don't add options to opt_state that don't reflect real
commandline options. See the above suggestion.

> I also set the incremental option. Since we're walking through the
> entire repository, I didn't feel it was necessary to have it create
> the full-text at every revision. Any problems with doing this?

Nope, that's all good, as long as this subcommand doesn't accept the
-r option.

> C-Mike, I've cc'd you on this because you seem to be my patch
> reviewer/mentor for this issue. :-)

I'll try to review the patch itself (instead of just your comments
about it above) Real Soon. Thanks, John.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Aug 25 01:35:39 2003