#2378 data tab operations need equivalents in Jython

[Jeremy Faden]

Sadie noticed that the data operation "|setDepend0Units('ms since 1970-01-01T00:00')" doesn't have an obvious analog in scripting. This makes me realize there are a number of functions in the data operations which are implemented with three or fours lines of code, which are easily composed in a script by me, but not by others who aren't so familiar with the libraries. The goal should be that any operation "|doXXX(arg)" should have a Jython function "doXXX(ds,arg)".

Short of that, script authors can work around this problem by calling the OperationsProcessor themselves, like so:

ds= dataset( findgen(40), sin( linspace( 0, 2*PI, 40 ) ) )

from org.das2.qds import OperationsProcessor
ds= OperationsProcessor.process( ds, "|setDepend0Units('ms since 1970-01-01T00:00')", None )

plot( ds )

(See https://github.com/autoplot/dev/demos/2021/20210818/demoOperationsProcessor.jy)

[Jeremy Faden]

It's ironic that sprocess is sprocess( str, ds ) not sprocess( ds, str )! Whoops...

[Jeremy Faden]

I've gone ahead and added process( ds, str ) and deprecated sprocess. I'll update the above example to use this.

[Jeremy Faden]

I made new setDepend0Units, setDepend1Units, and setValidRange in Ops (which is imported automatically into Jython scripts). Other functions are fairly simple wrappers around existing code.

[Jeremy Faden]

I see setDepend0Cadence but not setDepend0Units.

[Jeremy Faden]

We also saw that netcdf didn't allow "hrs+since+2020-01-01" for units, but does allow "hr since 2020-01-01T00:00Z"

[Jeremy Faden]

Also another script showed where the timetags are dropped in some operations.