[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Testing equality between svnrdump and svnadmin dump

From: Julian Foad <julianfoad_at_btopenworld.com>
Date: Mon, 26 Jan 2015 18:30:05 +0000

Branko Čibej wrote:
> On 26.01.2015 15:50, Julian Foad wrote:
>>  Excellent. Is anyone able to lend a hand with one of these two of the tasks
>> I listed:
>>
>>    - Bypass authz when dumping with svnrdump, otherwise a partial dump is
>>      obtained for tests using authz, making those tests fail.
>
> Hmmm, this will involve changing the DAV and svnserve autocheck
> configurations, I think. Not really my home turf, but I can try.

I wonder, rather than set up the authz with a back door from the start (that sounds like what you're thinking), could this extra test code switch off all authorization for the particular repository? Is that feasible, for the different kinds of server? It has plain disk access to the repos directory, but I don't know if it knows exactly where to find the authz file or other config files in every case.

>>    - Make optional -- enabled by a configure option?
>
> I can certainly do this bit, I had a go at similar tweaks when I added
> the option to skip the C tests and to run with pre-cooked v4 and v6 FSFS
> repositories.

Thank you!

>>  if I do this one:
>>
>>    - Put back the dumpfile parser's ability to check for well-formed dumpfile
>>      contents, which has been temporarily stripped out here in order to
>>      achieve more generic parsing, as some dump/load tests rely on that.
>
> May I suggest we do this on a branch?

Sure. Committed onto branch '^/subversion/branches/dump-load-cross-check' in r1654853.

- Julian
Received on 2015-01-26 19:31:03 CET

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.