fix(unpivot): _field should not be added as a group key by default

[skartikey]

if _field is in otherColumns parameter, _field should not be made group key by default
The issue was discovered during the iox group push down work

Checklist

Dear Author 👋, the following checks should be completed (or explicitly dismissed) before merging.

• ✏️ Write a PR description, regardless of triviality, to include the value of this PR
• 🔗 Reference related issues
• 🏃 Test cases are included to exercise the new code
• 🧪 If new packages are being introduced to stdlib, link to Working Group discussion notes and ensure it lands under experimental/
• 📖 If language features are changing, ensure docs/Spec.md has been updated

Dear Reviewer(s) 👋, you are responsible (among others) for ensuring the completeness and quality of the above before approval.

[wolffcm]

I think this is pretty close, but the logic is not quite right.

[wolffcm]

For this to be a representative test case, there shouldn't be a _field column in the input to unpivot, since pivoted data does generally not have a _field column.

[wolffcm]

Thinking it through, if there is a _field column in the input an error seems to be appropriate? Since there is no way to produce output that does not overwrite the data that already exists in _field. Thoughts?

[skartikey]

I think it makes sense to show an error if the input to unpivot contains _field in it. wondering if we have to do the same for _value column?

The reason I have to put up this PR is that the pushdown group test was failing because it contains _field in the input.

The fix would be to throw a user error when there is _field or _value column in the input to unpivot.

At the moment, if a column is present in the otherColumns and missing in the input column, the code throws an error.

import "array"
import "experimental"
import "testing"

 array.from(
    rows: [
        {
            _measurement: "m",
            tag: "t1",
            f0: 10.1,
            f1: 10.2,
            _time: 2018-12-01T00:00:00Z,
        },
        {
            _measurement: "m",
            tag: "t1",
            f0: 20.1,
            f1: 20.2,
            _time: 2018-12-01T00:00:10Z,
        },
    ],
)
    |> group(columns: ["_measurement"])
    |> experimental.unpivot(otherColumns: ["_time", "tag", "_field"])

Result: _result
Error: runtime error @24:8-24:70: unpivot: unpivot could not find column named "_field"

[wolffcm]

This logic doesn't seem quite right to me. If there is no _field column in the input, then the _field column will never get created. We want to create it regardless, we just want to omit the column from groupCols and groupValues below if _field is in the otherCols parameter.

[skartikey]

No, you’ve got it wrong. _field will only 'not' get created when it is present in the otherColumns, rest of the cases it will always get created.

import "array"
import "experimental"
import "testing"

 array.from(
    rows: [
        {
            _measurement: "m",
            tag: "t1",
            f0: 10.1,
            f1: 10.2,
            _time: 2018-12-01T00:00:00Z,
        },
        {
            _measurement: "m",
            tag: "t1",
            f0: 20.1,
            f1: 20.2,
            _time: 2018-12-01T00:00:10Z,
        },
    ],
)
    |> group(columns: ["_measurement"])
    |> experimental.unpivot(otherColumns: ["_time", "tag"])

Result: _result
Table: keys: [_measurement, _field]
   _measurement:string           _field:string                      _time:time              tag:string                  _value:float
----------------------  ----------------------  ------------------------------  ----------------------  ----------------------------
                     m                      f0  2018-12-01T00:00:00.000000000Z                      t1                          10.1
                     m                      f0  2018-12-01T00:00:10.000000000Z                      t1                          20.1
Table: keys: [_measurement, _field]
   _measurement:string           _field:string                      _time:time              tag:string                  _value:float
----------------------  ----------------------  ------------------------------  ----------------------  ----------------------------
                     m                      f1  2018-12-01T00:00:00.000000000Z                      t1                          10.2
                     m                      f1  2018-12-01T00:00:10.000000000Z                      t1                          20.2

[wolffcm]

Looks good, thanks.