On Sun, 2017-04-30 at 23:02 +0200, Philippe Mouawad wrote: > Why not. > But why not just use the Java class instead of maintaining a custom > class > in HC ? >
Philippe I replaced the existing implementation [1] with that from JRE and it broke our deflate compression tests [2]. I also see little urgency in replacing custom DeflateInputStream given it uses standard Deflater class internally and is relatively small. However anyone willing to invest time into fixing the problem is very welcome to do so and raise a PR once the cause of test breakage is identified and resolved. Oleg [1] https://github.com/apache/httpclient/blob/trunk/httpclient5/src/mai n/java/org/apache/hc/client5/http/entity/DeflateInputStream.java [2] https://github.com/apache/httpclient/blob/trunk/httpclient5/src/tes t/java/org/apache/hc/client5/http/entity/TestDeflate.java > Thanks > > On Sun, Apr 30, 2017 at 10:22 PM, Gary Gregory <[email protected] > m> > wrote: > > > Maybe we should make it pluggable with an optional impl using > > Commons > > Compress? > > > > Gary > > > > On Apr 30, 2017 1:02 PM, "Philippe Mouawad" <[email protected]> > > wrote: > > > > > Hello, > > > For Deflate response encoded content, HC4 has a custom > > > implementation > > > org.apache.http.client.entity.DeflateInputStream > > > > > > Is there any reason for not using > > > java.util.zip.DeflaterInputStream > > > available since Java 6. > > > > > > > > > Thanks > > > Regards > > > @philmdot > > > > > > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
