Re: Why common-compress skip data by reading and discard?

Another way you could support this is add an adapter that uses one 
approach or the other depending on the JDK version of the user. Use 
reflection to detect whether the new API exists or not.

Gili

On 2024-05-24 10:41, sebb wrote:
> On Fri, 24 May 2024 at 14:51, iseki zero <[email protected]> wrote:
>> Thank you. It will better if we can do a detection before we call it, to
>> improve the performance.
> Is there a performance issue here?
> If so, is it enough to be worth the effort of updating the code?
>
> Alternatively, it would be easy to update the Javadoc to advise users
> who are on Java12+ to update their code to use the new methods...
>
>
>> On 2024/05/23 11:21:33 Gary Gregory wrote:
>>   > Hello iseki,
>>   >
>>   > The Javadoc explains all of this in detail.
>>   >
>>   > Java 12 adds skipNBytes and IO is still based on Java 8.
>>   >
>>   > Gary
>>   >
>>   > On Thu, May 23, 2024, 6:30 AM iseki zero <[email protected]> wrote:
>>   >
>>   > > Hello,
>>   > >
>>   > > When I reading the code of common-compress, I found it skip data using
>>   > > the `IOUtils.skip` in `commons-io` module.
>>   > >
>>   > > I found this method just reading the data and discard it directly. But
>>   > > in JDK there's an more effeciency method `InputStream.skip` and
>>   > > `InputStream.skipNBytes`.
>>   > >
>>   > > So why we read & discard data instead of call `skip*` directly?
>>   > >
>>   > > Thank you.
>>   > >
>>   > > iseki zero.
>>   > >
>>   > >
>>   > > ---------------------------------------------------------------------
>>   > > To unsubscribe, e-mail: [email protected]
>>   > > For additional commands, e-mail: [email protected]
>>   > >
>>   > >
>>   >
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [email protected]
>> For additional commands, e-mail: [email protected]
>>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>