uni-plugin-rhai 2.2.4

Rhai-script loader for the uni-db plugin framework
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146

//! Vectorized scalar fn — column userdata path.

#![cfg(feature = "rhai-runtime")]

use std::sync::Arc;

use arrow_array::{Array, Float64Array};
use datafusion::logical_expr::ColumnarValue;
use uni_plugin::{Capability, CapabilitySet, PluginId, PluginRegistrar, PluginRegistry, QName};
use uni_plugin_rhai::RhaiLoader;

#[test]
fn vectorized_scalar_fn_round_trips() {
    let script = r#"
        fn uni_manifest() {
            #{
                id: "ai.test.vec",
                version: "0.1.0",
                scalar_fns: [
                    #{ name: "score_v", vectorized: true,
                       args: ["float","float"], returns: "float" },
                ],
            }
        }
        fn score_v(xs, ys) {
            let n = xs.len();
            let out = uni_float_column(n);
            for i in 0..n {
                out[i] = xs[i] * 0.7 + ys[i] * 0.3;
            }
            out
        }
    "#;

    let registry = PluginRegistry::new();
    let caps = CapabilitySet::from_iter_of([Capability::ScalarFn]);
    let mut r = PluginRegistrar::new(PluginId::new("rhai.loading"), &caps, &registry);
    let outcome = RhaiLoader::new()
        .load(script, &mut r, &caps)
        .expect("loads");
    assert_eq!(outcome.scalars_registered.len(), 1);
    r.commit_to_registry().expect("commits");

    let qn = QName::new("ai.test.vec", "score_v");
    let entry = registry.scalar_fn(&qn).expect("registered");

    let xs = Arc::new(Float64Array::from(vec![1.0, 2.0, 3.0]));
    let ys = Arc::new(Float64Array::from(vec![10.0, 20.0, 30.0]));
    let args = vec![ColumnarValue::Array(xs), ColumnarValue::Array(ys)];
    let out = entry.function.invoke(&args, 3).expect("invokes");
    let arr = match out {
        ColumnarValue::Array(a) => a,
        other => panic!("expected Array, got {other:?}"),
    };
    let f = arr.as_any().downcast_ref::<Float64Array>().unwrap();
    assert!((f.value(0) - 3.7).abs() < 1e-9);
    assert!((f.value(1) - 7.4).abs() < 1e-9);
    assert!((f.value(2) - 11.1).abs() < 1e-9);
}

#[test]
fn vectorized_matches_row_mode_for_haversine() {
    // Run both row-mode and vectorized haversine; compare cells.
    let script = r#"
        fn uni_manifest() {
            #{
                id: "ai.test.hv",
                version: "0.1.0",
                scalar_fns: [
                    #{ name: "h_row", args: ["float","float","float","float"], returns: "float" },
                    #{ name: "h_vec", vectorized: true,
                       args: ["float","float","float","float"], returns: "float" },
                ],
            }
        }
        const R = 6371.0;
        fn hav(lat1, lon1, lat2, lon2) {
            let rlat1 = lat1.to_radians();
            let rlat2 = lat2.to_radians();
            let dlat = (lat2 - lat1).to_radians();
            let dlon = (lon2 - lon1).to_radians();
            let a = (dlat / 2.0).sin() ** 2
                  + rlat1.cos() * rlat2.cos() * (dlon / 2.0).sin() ** 2;
            global::R * 2.0 * a.sqrt().asin()
        }
        fn h_row(lat1, lon1, lat2, lon2) { hav(lat1, lon1, lat2, lon2) }
        fn h_vec(la1, lo1, la2, lo2) {
            let n = la1.len();
            let out = uni_float_column(n);
            for i in 0..n {
                out[i] = hav(la1[i], lo1[i], la2[i], lo2[i]);
            }
            out
        }
    "#;

    let registry = PluginRegistry::new();
    let caps = CapabilitySet::from_iter_of([Capability::ScalarFn]);
    let mut r = PluginRegistrar::new(PluginId::new("rhai.loading"), &caps, &registry);
    RhaiLoader::new()
        .load(script, &mut r, &caps)
        .expect("loads");
    r.commit_to_registry().expect("commits");

    let lat1 = Arc::new(Float64Array::from(vec![40.7128, 51.5074]));
    let lon1 = Arc::new(Float64Array::from(vec![-74.0060, -0.1278]));
    let lat2 = Arc::new(Float64Array::from(vec![37.7749, 48.8566]));
    let lon2 = Arc::new(Float64Array::from(vec![-122.4194, 2.3522]));
    let args = vec![
        ColumnarValue::Array(lat1.clone()),
        ColumnarValue::Array(lon1.clone()),
        ColumnarValue::Array(lat2.clone()),
        ColumnarValue::Array(lon2.clone()),
    ];

    let row_out = match registry
        .scalar_fn(&QName::new("ai.test.hv", "h_row"))
        .unwrap()
        .function
        .invoke(&args, 2)
        .unwrap()
    {
        ColumnarValue::Array(a) => a,
        _ => panic!(),
    };
    let vec_out = match registry
        .scalar_fn(&QName::new("ai.test.hv", "h_vec"))
        .unwrap()
        .function
        .invoke(&args, 2)
        .unwrap()
    {
        ColumnarValue::Array(a) => a,
        _ => panic!(),
    };
    let r_arr = row_out.as_any().downcast_ref::<Float64Array>().unwrap();
    let v_arr = vec_out.as_any().downcast_ref::<Float64Array>().unwrap();
    for i in 0..2 {
        assert!(
            (r_arr.value(i) - v_arr.value(i)).abs() < 1e-9,
            "row vs vec mismatch at {i}: row={} vec={}",
            r_arr.value(i),
            v_arr.value(i)
        );
    }
}