Thanks Adrien:

Would the TestCompressingStoredFieldsFormat would eventually catch it?

This one seems to randomize its parameters, but perhaps it would be
good to explicitly add Test*StoredFieldsFormat's for the different
test codec modes we have:
CompressionMode.HIGH_COMPRESSION, CompressionMode.FAST_DECOMPRESSION,
CompressionMode.Fast ?


On Wed, Sep 4, 2013 at 12:06 PM, Adrien Grand <[email protected]> wrote:
> Actually this is a side-effect of LUCENE-5188. There is a bug in
> LZ4.compressHC (which I committed to test various trade-offs between
> compression speed and ratio but is not used in any official codec) on
> very compressible inputs which seems to be more easily triggered now
> that the inputs can be sliced. I have a fix that I'm testing and
> should be able to commit soon.
>
> --
> Adrien
>
> ---------------------------------------------------------------------
> 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]

Reply via email to