saya-cli 0.3.1

Database-aware AI agent for the terminal: full-screen TUI, schema discovery, and bounded read-only SQL over PostgreSQL, MySQL, SQLite, DuckDB, and Snowflake.
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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237

//! Generates self-contained interactive Chart.js HTML documents.

mod kind;
mod render;
mod spec;

#[cfg(test)]
use saya_types::QueryResult;

pub(crate) use kind::{ChartKind, ChartSpec};
pub(crate) use render::render_html;
pub(crate) use spec::suggest_spec;

pub(super) fn is_numeric_column(rows: &[Vec<serde_json::Value>], col_idx: usize) -> bool {
    let mut non_null_count = 0;
    for row in rows {
        if let Some(cell) = row.get(col_idx).filter(|c| !c.is_null()) {
            non_null_count += 1;
            if cell_to_f64(cell).is_none() {
                return false;
            }
        }
    }
    non_null_count >= 1
}

/// Interprets a cell as an `f64`, accepting both JSON numbers and numeric
/// *strings*. Postgres `NUMERIC`/`DECIMAL` and MySQL `DECIMAL` (e.g. `SUM`/`AVG`,
/// money columns) decode to JSON strings, so a number-only check would silently
/// drop them from charts.
pub(super) fn cell_to_f64(value: &serde_json::Value) -> Option<f64> {
    match value {
        serde_json::Value::Number(_) => value.as_f64(),
        serde_json::Value::String(s) => s.trim().parse::<f64>().ok(),
        _ => None,
    }
}

pub(super) fn normalize_row(row: &serde_json::Value, col_count: usize) -> Vec<serde_json::Value> {
    let mut cells = match row {
        serde_json::Value::Array(arr) => arr.clone(),
        scalar => vec![scalar.clone()],
    };
    cells.resize(col_count, serde_json::Value::Null);
    cells
}

/// Writes the HTML document string to the specified path.
#[allow(dead_code)]
pub(crate) fn write_html(html: &str, path: &std::path::Path) -> Result<(), String> {
    std::fs::write(path, html).map_err(|e| format!("failed to write chart file: {e}"))
}

/// Opens `path` in the user's default application (browser for .html).
pub(crate) fn open_file(path: &std::path::Path) -> Result<(), String> {
    use std::process::Command;
    #[cfg(target_os = "macos")]
    let mut cmd = {
        let mut c = Command::new("open");
        c.arg(path);
        c
    };
    #[cfg(all(unix, not(target_os = "macos")))]
    let mut cmd = {
        let mut c = Command::new("xdg-open");
        c.arg(path);
        c
    };
    #[cfg(target_os = "windows")]
    let mut cmd = {
        let mut c = Command::new("cmd");
        c.args(["/C", "start", ""]).arg(path);
        c
    };
    cmd.spawn()
        .map(|_| ())
        .map_err(|e| format!("could not open {}: {e}", path.display()))
}

#[cfg(test)]
mod chart_gen_tests {
    use super::*;
    use serde_json::json;

    #[test]
    fn test_chart_kind_parse() {
        assert_eq!(ChartKind::parse("bar"), Some(ChartKind::Bar));
        assert_eq!(ChartKind::parse("BAR"), Some(ChartKind::Bar));
        assert_eq!(ChartKind::parse("Line"), Some(ChartKind::Line));
        assert_eq!(ChartKind::parse("AREA"), Some(ChartKind::Area));
        assert_eq!(ChartKind::parse("pie"), Some(ChartKind::Pie));
        assert_eq!(ChartKind::parse("Doughnut"), Some(ChartKind::Doughnut));
        assert_eq!(ChartKind::parse("scatter"), Some(ChartKind::Scatter));
        assert_eq!(ChartKind::parse("unknown"), None);
        assert_eq!(ChartKind::parse("123"), None);
    }

    #[test]
    fn test_suggest_spec() {
        // text + numeric => Bar
        let text_num_result = QueryResult {
            columns: vec!["category".to_string(), "count".to_string()],
            rows: vec![json!(["A", 10]), json!(["B", 20])],
            row_count: 2,
            truncated: false,
            executed_sql: "SELECT category, count FROM t".to_string(),
        };
        let spec = suggest_spec(&text_num_result);
        assert_eq!(spec.kind, ChartKind::Bar);
        assert_eq!(spec.x, Some("category".to_string()));
        assert_eq!(spec.y, vec!["count".to_string()]);
        assert_eq!(spec.title, None);

        // two numeric => Scatter
        let two_num_result = QueryResult {
            columns: vec!["x_val".to_string(), "y_val".to_string()],
            rows: vec![json!([1.0, 2.0]), json!([3.0, 4.0])],
            row_count: 2,
            truncated: false,
            executed_sql: "SELECT x_val, y_val FROM t".to_string(),
        };
        let spec2 = suggest_spec(&two_num_result);
        assert_eq!(spec2.kind, ChartKind::Scatter);
        assert_eq!(spec2.x, Some("x_val".to_string()));
        assert_eq!(spec2.y, vec!["y_val".to_string()]);
        assert_eq!(spec2.title, None);
    }

    #[test]
    fn test_render_html_bar_chart() {
        let result = QueryResult {
            columns: vec!["rating".to_string(), "films".to_string()],
            rows: vec![json!(["G", 178]), json!(["R", 195])],
            row_count: 2,
            truncated: false,
            executed_sql: "SELECT rating, films FROM film".to_string(),
        };
        let spec = ChartSpec {
            kind: ChartKind::Bar,
            x: Some("rating".to_string()),
            y: vec!["films".to_string()],
            title: None,
        };
        let html = render_html(&result, &spec).unwrap();
        assert!(html.contains("\"type\":\"bar\""));
        assert!(html.contains("178"));
        assert!(html.contains("195"));
        assert!(html.contains("\"G\""));
        assert!(html.contains("\"R\""));
        assert!(html.contains("Chart.js v4.4.4"));

        let no_num_result = QueryResult {
            columns: vec!["col1".to_string(), "col2".to_string()],
            rows: vec![json!(["a", "b"]), json!(["c", "d"])],
            row_count: 2,
            truncated: false,
            executed_sql: "SELECT col1, col2 FROM t".to_string(),
        };
        let err_spec = ChartSpec {
            kind: ChartKind::Bar,
            x: None,
            y: Vec::new(),
            title: None,
        };
        assert!(render_html(&no_num_result, &err_spec).is_err());
    }

    // Postgres NUMERIC/DECIMAL and MySQL DECIMAL decode to JSON *strings*, not
    // numbers (e.g. SUM(amount) -> "3094.78"). They must still chart.
    #[test]
    fn numeric_string_values_are_plotted() {
        let result = QueryResult {
            columns: vec!["month".to_string(), "revenue".to_string()],
            rows: vec![
                json!(["2022-01", "3094.78"]),
                json!(["2022-02", "10164.97"]),
            ],
            row_count: 2,
            truncated: false,
            executed_sql: "SELECT month, SUM(amount) AS revenue FROM payment GROUP BY 1"
                .to_string(),
        };
        let spec = suggest_spec(&result);
        assert_eq!(spec.y, vec!["revenue".to_string()]);
        let html = render_html(&result, &spec).unwrap();
        assert!(html.contains("3094.78"), "numeric-string value was dropped");
        assert!(
            html.contains("10164.97"),
            "numeric-string value was dropped"
        );
    }

    #[test]
    fn suggest_spec_detects_numeric_string_columns() {
        // Two numeric columns encoded as strings should be recognized as numeric
        // (and so picked as a scatter), not treated as categorical text.
        let result = QueryResult {
            columns: vec!["length".to_string(), "avg_rate".to_string()],
            rows: vec![json!(["46", "2.59"]), json!(["47", "2.70"])],
            row_count: 2,
            truncated: false,
            executed_sql: "SELECT length, AVG(rate) FROM film GROUP BY 1".to_string(),
        };
        assert_eq!(suggest_spec(&result).kind, ChartKind::Scatter);
    }

    #[test]
    fn test_render_html_script_injection_prevention() {
        let result = QueryResult {
            columns: vec!["label".to_string(), "val".to_string()],
            rows: vec![
                json!(["</script><img src=x onerror=alert(1)>", 10]),
                json!(["</SCRIPT>", 20]),
            ],
            row_count: 2,
            truncated: false,
            executed_sql: "SELECT label, val FROM t".to_string(),
        };
        let spec = ChartSpec {
            kind: ChartKind::Bar,
            x: Some("label".to_string()),
            y: vec!["val".to_string()],
            title: None,
        };
        let html = render_html(&result, &spec).unwrap();

        assert!(html.contains("id=\"saya-chart-config\""));
        assert!(html.contains("type=\"application/json\""));
        assert!(html.contains("JSON.parse"));

        assert!(!html.contains("</script><img"));
        assert!(!html.contains("</SCRIPT>"));

        assert!(html.contains("\\u003c/script"));
        assert!(html.contains("\\u003c/SCRIPT"));
    }
}