Github user mmiklavc commented on the issue:
https://github.com/apache/metron/pull/1021
+1 via inspection.
---
Github user cestella commented on the issue:
https://github.com/apache/metron/pull/1021
Ok, just ran this up in full-dev and it works as advertised. If this needs
to change the fundamental approach, let me know, but as it is submitted it
functions as advertised.
---
Github user cestella commented on the issue:
https://github.com/apache/metron/pull/1021
@ottobackwards As long as the underlying map is updated in
`MapVariableResolver` as per the `assign` function in that variable resolver,
it should be supported without any change as the approach
Github user ottobackwards commented on the issue:
https://github.com/apache/metron/pull/1021
@cestella re:assignment. If all assignment is in the language, and _ is
either or, then how can we assign to a message field when _ is active?
---
Github user simonellistonball commented on the issue:
https://github.com/apache/metron/pull/1021
This is a great feature from my perspective. I've seen a number of people
try and build solutions that require access to the whole message for a bunch of
custom functions, models, rules
Github user cestella commented on the issue:
https://github.com/apache/metron/pull/1021
[Multimaps](https://google.github.io/guava/releases/19.0/api/docs/com/google/common/collect/Multimap.html#entries())
don't have unique keys. Making the unique keys would necessitate copying the
Github user cestella commented on the issue:
https://github.com/apache/metron/pull/1021
@ottobackwards I don't see it really affecting assignment. `_` should
behave just like any other variable to my mind and return a map. For instance,
if you say `new_field := _`, then I'd expect