figuring out if we have an issue with unary datemath

[epsniff]

We may have an issue with unary support with datemath. This PR is just to investigate that issue and to possibly fix it if we find an issue. For example FILTER NOT last_event < "now-1d"

to run test:

go test --run TestDateMath/FILTER_NOT -v 

[epsniff]

@araddon I added indentation to the printing of the AST:

epsniff@hms-victory:...n/qlbridge/vm$ go test --run TestDateMath/FILTER_NOT -v 
=== RUN   TestDateMath
=== RUN   TestDateMath/FILTER_NOT_last_event_<_"now-1d"
node *expr.UnaryNode  NOT (last_event < "now-1d")  negated:false
  node *expr.BinaryNode  (last_event < "now-1d")  negated:true
    node *expr.IdentityNode  last_event  negated:true
node *expr.UnaryNode  NOT (last_event < "now-1d")  negated:false
  node *expr.BinaryNode  (last_event < "now-1d")  negated:true
    node *expr.IdentityNode  last_event  negated:true
filter      :FILTER NOT last_event < "now-1d"
t1:         :2019-05-31 17:49:28.480704699 -0700 PDT m=+0.002168089
boundary    :2019-06-01 17:49:28.480704699 -0700 PDT m=+86400.002168047
diff        :23h59m59.999999958s
timeStrings :[now-1d]
node *expr.IncludeNode  INCLUDE not_valid_lookup  negated:false
node *expr.BinaryNode  last_event > "now-3x"  negated:false
  node *expr.IdentityNode  last_event  negated:false
node *expr.BinaryNode  last_event == "now-3d"  negated:false
  node *expr.IdentityNode  last_event  negated:false

But Im not sure why it's printing off two ASTs yet....