On 03/19/2012 08:50 PM, Gary Gregory wrote:
> On Mon, Mar 19, 2012 at 3:08 PM, Thomas Neidhart
> <thomas.neidh...@gmail.com>wrote:
> 
>> On 03/19/2012 05:38 PM, Gary Gregory wrote:
>>
>>> Ok, can you: (1) back out the QuotedPrintableCodec changes and (2)
>> complete
>>> Codec-130?
>>
>> (1) I am working on it, svn seems to be dead over here in EU
>>    The reverted diff will be attached to the issue
>>
>> (2) I managed to commit it, would be nice if you could review the
>>    changes, the CC for the new methods is 100% (missing some coverage
>>    for read though)
>>
> 
> On (2): Looks good. I committed a couple of tweaks.
> 
> Why should we have the almost same impl InflaterIS WRT the limit on the
> size of the input skip param? What if I want to skip in some giant
> multi-gig file (unlikely perhaps)?

This was my first guess. InputStream, e.g. uses a skip buffer of 2048.

The assumption is that nobody will call this method anyway, just in case
somebody does, it should work correctly (though slow maybe).

But I would really avoid creating a buffer the size of the number of
bytes to skip, as this could cause serious problems (e.g. DOS attacks).

Thomas

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org

Reply via email to