Access to index without SQL++ query

SQL++
,
19

Does it do a full index scan or does it use the values in the JSON vector bounds?

What does the query plan show?

If I use this (equivalent) query on the travel-sample, and create the recommended index on [ airportname, city]

select airport.* from travel-sample.inventory.airport where [ airportname, city] >= ['Albany Intl', 'Albany' ] and [ airportname, city] <= ['Alice Intl' ,'Alice' ]

The resulting query processes only the 9 matching documents. Even without specifying USE INDEX. (there are 1968 airport documents in the colllection).

And look at the range used in the query - it’s exactly what your “index API” would use.

"~children": [
      {
        "#operator": "IndexScan3",
        "#stats": {
          "#itemsIn": 9,
          "#itemsOut": 9,
          "#phaseSwitches": 39,
          "execTime": "24.127µs",
          "kernTime": "2.825µs",
          "servTime": "1.239574ms"
        },
        "bucket": "travel-sample",
        "index": "adv_airportnamecity",
        "index_id": "b24708afc103be3f",
        "index_projection": {
          "primary_key": true
        },
        "keyspace": "airport",
        "namespace": "default",
        "optimizer_estimates": {
          "cardinality": 1.000000000000183,
          "cost": 12.181079088630932,
          "fr_cost": 12.1810790886309,
          "size": 12
        },
        "scope": "inventory",
        "spans": [
          {
            "exact": true,
            "range": [
              {
                "high": "[\"Alice Intl\", \"Alice\"]",
                "inclusion": 3,
                "index_key": "[`airportname`, `city`]",
                "low": "[\"Albany Intl\", \"Albany\"]"
              }
            ]
          }
        ],