threads / discuss / 15090

Stange diff in "mailinfo: re-fix MIME multipart boundary parsing"

Subject: Stange diff in "mailinfo: re-fix MIME multipart boundary parsing"

## tl;dr

4 messages between Aug 19, 2008 and Aug 20, 2008.

replies: 3people: 3as markdown or json

Marco Costalba· Aug 19, 2008, 17:38 UTC · lore

Looking the diff at

http://git.kernel.org/?p=git/git.git;a=commitdiff;h=289796dd29dd656734cfd59b657deb943a71cf6a

the part of the patch applied to /t/t5100/sample.mbox file seems very strange, is it correct ?

Marco

Marcus Griep· Aug 19, 2008, 17:51 UTC · re: Marco Costalba · lore

Re: Stange diff in "mailinfo: re-fix MIME multipart boundary parsing"

Marco Costalba wrote:

Show 6 quoted lines

> Looking the diff at
> 
> http://git.kernel.org/?p=git/git.git;a=commitdiff;h=289796dd29dd656734cfd59b657deb943a71cf6a
> 
> the part of the patch applied to /t/t5100/sample.mbox file seems very
> strange, is it correct ?

It looks like an empty line added to the end of the file. Unsanitary, but ok since it comes after the last MIME multipart boundary. Junio could probably just ax the change to sample.mbox from the patch before applying it.

-- 
Marcus Griep
GPG Key ID: 0x5E968152
——
http://www.boohaunt.net
את.ψο´

Marcus Griep· Aug 19, 2008, 17:57 UTC · re: Marcus Griep · lore

Re: Stange diff in "mailinfo: re-fix MIME multipart boundary parsing"

Marcus Griep wrote:

> It looks like an empty line added to the end of the file.
> Unsanitary, but ok since it comes after the last MIME
> multipart boundary.  Junio could probably just ax the change
> to sample.mbox from the patch before applying it.

Or I could have misunderstood the intent of the change, and it is necessary to fully test the change, since it appears to operate line-by-line, an empty line at the end would be necessary to trigger the code path that stops mailinfo from looking for another boundary.

-- 
Marcus Griep
GPG Key ID: 0x5E968152
——
http://www.boohaunt.net
את.ψο´

Don Zickus· Aug 20, 2008, 13:10 UTC · re: Marcus Griep · lore

Re: Stange diff in "mailinfo: re-fix MIME multipart boundary parsing"

On Tue, Aug 19, 2008 at 1:57 PM, Marcus Griep <neoeinstein@gmail.com> wrote:

Show 5 quoted lines

> Marcus Griep wrote:
>> It looks like an empty line added to the end of the file.
>> Unsanitary, but ok since it comes after the last MIME
>> multipart boundary.  Junio could probably just ax the change
>> to sample.mbox from the patch before applying it.

No the change was intentional. In fact, because that extra line was not there, a bug was hidden. Adding the extra line exposed the bug and hopefully will catch any more bugs in the future with regards to the boundary code.

Show 6 quoted lines

>
> Or I could have misunderstood the intent of the change, and it
> is necessary to fully test the change, since it appears to
> operate line-by-line, an empty line at the end would be necessary
> to trigger the code path that stops mailinfo from looking for
> another boundary.

Yes, exactly.

Cheers, Don