why not using commons-codec

[JensKapitza]

i'm just reading a little bit with your bloom filter integration but
i think i missed the part where codec dependency where discussed

org.apache.commons.codec.digest.MurmurHash3
for hashing?

[Claudenw]

Somehow I missed that implementation. I will change to it.

[Claudenw]

org.apache.commons.codec.digest.MurmurHash3 does not accept a ByteBuffer argument. And I can not guarantee that the ByteBuffer that I have has a buffer behind it, for very large objects this could be a problem.

…

On Mon, Oct 28, 2019 at 6:47 PM Jens ***@***.***> wrote: i'm just reading a little bit with your bloom filter integration but i think i missed the part where codec dependency where discussed org.apache.commons.codec.digest.MurmurHash3 for hashing? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AASTVHTBGISZY4RTXTW4XADQQ4XVDANCNFSM4JF6OJZQ> .

-- I like: Like Like - The likeliest place on the web <http://like-like.xenei.com> LinkedIn: http://www.linkedin.com/in/claudewarren

[JensKapitza]

i've not looked in detail, but there was a hash32(final byte[] data, ...) function and the buffer has a .array() function?

should'n this work?

[JensKapitza]

other task: you have a good reason to push a change to codecs with a merge of your code ( integration should be easier in small parts) your hasher is than in one commons project

[Claudenw]

the array() method only works if the ByteBuffer was constructed with an array of byte. There are several cases where this is not true. I am not certain what happens in the memory mapped file situation. However, it might make more sense for me to change the hasher implementation to take a byte[] and leave it at that. Claude

…

On Tue, Oct 29, 2019 at 11:41 AM Jens ***@***.***> wrote: i've not looked in detail, but there was a hash32(final byte[] data, ...) function and the buffer has a .array() function? should'n this work? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#1>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AASTVHVYGXWNXUHI5BA47NTQRAOPJANCNFSM4JF6OJZQ> .

-- I like: Like Like - The likeliest place on the web <http://like-like.xenei.com> LinkedIn: http://www.linkedin.com/in/claudewarren