Timo and Dawid,
Registering my UDF via the deprecated *registerFunction()* instead of the
new *createTemporarySystemFunction()* worked. So it would appear there is
some incompatibility with my implementation and the new registration
system. I will wait for the Flip to be completed and retry then.
Sorry for jumping in so late. I think Dawid gave a nice summary.
As he said, integration of the DataStream <> Table integration is still
under development. Until then I would suggest to option 3) which means
don't upgrade the functions and use the old registration function
`registerFunction`.
Hi Steve,
Unfortunately the information you posted still does not explain how you
ended up with *RAW('java.util.Map', ?)* for your input type. Would be
best if you could share an example that I could use to reproduce it.
I tried putting down some potential approaches:
I tested it with a class ge
Hi Dawid,
Just wanted to bump this thread in case you had any thoughts.
Thanks,
Steve
On Thu, Oct 29, 2020 at 2:42 PM Steve Whelan wrote:
> For some background, I am upgrading from Flink v1.9 to v1.11. So what I am
> about to describe is our implementation on v1.9, which worked. I am trying
>
For some background, I am upgrading from Flink v1.9 to v1.11. So what I am
about to describe is our implementation on v1.9, which worked. I am trying
to achieve the same functionality on v1.11.
I have a DataStream whose type is an avro generated POJO, which contains a
field *UrlParameters* that is
Sorry for a late reply.
Could you share a complete, reproducible example? I am mostly interested
in where do you get the input RAW('java.util.Map', '...') type that you
are passing into your UDF.
Raw types are equal/equivalent only if both the class and the serializer
are equal.
A side note: Hav
Hi Dawid,
I added `*bridgedTo = Map.class*` as you suggested and got a slightly
different exception. I also tried passing a rawSerializer (an
implementation similar to MapSerializer[1] with String type key and value)
but got the same exception as without it. I am using Flink v1.11 for
reference.
Hey Steve,
You should be able to do via the bridgedTo parameter. You can
additionally specify a serializer you want to use via rawSerializer
parameter:
@FunctionHint(
input = {
@DataTypeHint(value = "RAW", bridgedTo =
Map.class[, rawSerializer = ...
Hi Steve,
Thanks for reaching out to the Flink community. I am pulling in Timo who
might be able to help you with this question.
Best,
Jark
On Mon, 26 Oct 2020 at 23:10, Steve Whelan wrote:
> Hi,
>
> I have a column of type *RAW('java.util.Map', ?)* that I want to pass to
> a scalar function
