Philip Martin wrote:
>Julian Foad <julianfoad@btopenworld.com> writes:
>
>
>
>>Philip Martin wrote:
>>
>>
>>>Martin Furter <mf@rola.ch> writes:
>>>
>>>
>>>>The config is stored in apr hashes so i assume that the value of the
>>>>former one was replaced by that of the latter one when reading the
>>>>config.
>>>>
>>>>
>>>We could check when adding to the hash that the key does not already
>>>exist. It could even be a hard error, just like a config parse
>>>error.
>>>
>>>
>>Yes, I'd say that should be considered a parse error regardless of
>>whether any additional clarification is needed in the autoprops
>>documentation. Are you volunteering?
>>
>>
>
>Was allowing ~/.subversion/config to contain things like
>
> # the second one takes precedence
> global-ignores = *.o
> global-ignores = *.o *.obj
>
>an conscious decision, or is that simply the way it happened to turn
>out?
>
>
It was a conscious decision. It simplifies the parser immensely, and it
especially simplifies the implementation of svn_config_merge. If anyone
wants to go to the trouble of forbidding duplicate option (and section)
names while still making svn_config_merge work, be my guest. I don't
think it's woth the trouble; it'd be much easier to document that a
later occurrence of the same option overrides the previous value.
--
Brane Čibej <brane_at_xbc.nu> http://www.xbc.nu/brane/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Oct 17 23:17:58 2003