[PATCH 1 of 2] revlog: return offset from _chunkraw()

Martin von Zweigbergk martinvonz at google.com
Tue Jan 5 23:14:17 CST 2016


On Tue, Jan 5, 2016 at 9:11 PM Gregory Szorc <gregory.szorc at gmail.com>
wrote:

> On Tue, Jan 5, 2016 at 8:52 PM, Martin von Zweigbergk <
> martinvonz at google.com> wrote:
>
>>
>> On Tue, Jan 5, 2016 at 8:47 PM Gregory Szorc <gregory.szorc at gmail.com>
>> wrote:
>>
>>> On Tue, Jan 5, 2016 at 8:39 PM, Martin von Zweigbergk <
>>> martinvonz at google.com> wrote:
>>>
>>>>
>>>>
>>>> On Tue, Jan 5, 2016 at 7:52 PM Gregory Szorc <gregory.szorc at gmail.com>
>>>> wrote:
>>>>
>>>>> +        Callers will need to call ``self.start(rev)`` and
>>>>> ``self.length(rev)``
>>>>> +        to determine where each revision's data begins and ends.
>>>>
>>>>
>>>> Is that still true? Is the returned offset/start different from
>>>> self.start(rev)?
>>>>
>>>
>>> The function returns data for 1 or more revisions. The first revision
>>> will have offset == self.start(rev), so you don't need to call anything if
>>> you are only getting a single revision.
>>>
>>
>> I see. Will test and push to the clowncopter soon. Thanks for cleaning
>> up! And great commit message on the next patch. I'm glad someone seems to
>> remember the (obsolete) reason for it. I would probably have given up
>> trying to understand that piece of code.
>>
>
> This should be in the commit message, but
> https://selenic.com/repo/hg/rev/f13728d59c0e the changed code in patch 2.
> Since `hg serve` is referenced, I assume it is related to thread safety.
> Actually, https://selenic.com/repo/hg/rev/4ab287c2d337 could probably be
> reverted too...
>

Want me to hold off queuing and wait for a V2? I don't mind queuing as is
and you can send a patch 3/2 later if you feel like it.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://selenic.com/pipermail/mercurial-devel/attachments/20160106/f13962b4/attachment.html>


More information about the Mercurial-devel mailing list