companion object & extra schema

[xhudik]

Hi, 2 newbies questions:

1. why did you use companion object for each EvalFunc[Tuple] in your files? I mean why those functions in companion objects are not defined within the class extending EvalFunc? Is there some performance adavantage?
2. outputSchema: (e.g. here https://github.com/bearrito/scala-pig/blob/master/src/main/scala/barrett/udf/SimpleWeatherParser.scala). There is defined a new schema: val tuple = new Schema(), everything is set up and then at the end another schema (s) is created and returned. Why this extra s schema had to be defined?

thanks!

[bearrito]

I wanted to be able to test those methods without creating an instance of the parser. If I had to do it over I would have put them in a trait that is probably more scala idiomatic.

The extra tuple is just a bug. I believe I had an error and was doing some debugging and pulled apart the tuple to make it easier to see what was happening.

All of this code was created for a demo, so it isn't necessarily the cleanest.