Function utf8_setup

datafusion/pruning/src/pruning_predicate.rs:4456–4489

Creates a setup for chunk pruning, modeling a utf8 column "s1" with 5 different containers (e.g. RowGroups). They have [min, max]: s1 ["A", "Z"] s1 ["A", "L"] s1 ["N", "Z"] s1 [NULL, NULL] s1 ["A", NULL] s1 ["", "A"] s1 ["", ""] s1 ["AB", "A\u{10ffff}"] s1 ["A\u{10ffff}\u{10ffff}\u{10ffff}", "A\u{10ffff}\u{10ffff}"]

()

    /// s1 ["AB", "A\u{10ffff}"]
    /// s1 ["A\u{10ffff}\u{10ffff}\u{10ffff}", "A\u{10ffff}\u{10ffff}"]
    fn utf8_setup() -> (SchemaRef, TestStatistics) {
        let schema = Arc::new(Schema::new(vec![Field::new("s1", DataType::Utf8, true)]));

        let statistics = TestStatistics::new().with(
            "s1",
            ContainerStats::new_utf8(
                vec![
                    Some("A"),
                    Some("A"),
                    Some("N"),
                    Some("M"),
                    None,
                    Some("A"),
                    Some(""),
                    Some(""),
                    Some("AB"),
                    Some("A\u{10ffff}\u{10ffff}"),
                ], // min
                vec![
                    Some("Z"),
                    Some("L"),
                    Some("Z"),
                    Some("M"),
                    None,
                    None,
                    Some("A"),
                    Some(""),
                    Some("A\u{10ffff}\u{10ffff}\u{10ffff}"),
                    Some("A\u{10ffff}\u{10ffff}"),
                ], // max
            ),
        );
        (schema, statistics)
    }

    #[test]
    fn prune_utf8_eq() {