spark-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Luis Ángel Vicente Sánchez <langel.gro...@gmail.com>
Subject Re: proposal: replace lift-json with spray-json
Date Sun, 09 Feb 2014 20:26:30 GMT
I could take a look on how lift-json is used in spark and compare it to
argonaut.io. That would be a nice exercise as I still haven't played with
json in spark.
On 9 Feb 2014 20:12, "Will Benton" <willb@redhat.com> wrote:

> ----- Original Message -----
>
> > spray-json only relies on parboiled, a library mantained by the
> spray-team
> > itself. Right now it's performance is much worse than lift-json (
> > http://engineering.ooyala.com/blog/comparing-scala-json-libraries) but
> that
> > would change when they finished parboiled2.
>
> Thanks for this feedback, Luis!
>
> > And alternative could be argonaut.io but you would bring scalaz as a
> > dependency.
>
> Yes, I thought about mentioning Argonaut, but was concerned about bringing
> in scalaz as an immediate dependency.  (From a packaging perspective,
> Argonaut is equivalent to Lift in this regard, since IIRC scalaz is a build
> and runtime dependency for lift-json.)  More generally, though, I wasn't
> sure how well the Argonaut style would fit in to Spark as a replacement for
> lift-json.  If it's not a concern to bring in scalaz or that style of code,
> then Argonaut is worth considering as well.
>
> (It looks like json4s and jackson-module-scala are excellent candidates in
> any case.)
>
>
> best,
> wb
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message