rivet-cli 0.16.2

Rivet: PostgreSQL/MySQL/SQL Server → Parquet/CSV (local, S3, GCS, Azure). Crate name rivet-cli; binary rivet.
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
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
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
//! Pure chunk math — no I/O, no database access.
//!
//! Range generation, chunk-query SQL building, plan fingerprinting. All
//! functions in this module are deterministic and side-effect-free. Parsing the
//! DB scalars the planner turns into bounds (`min`/`max` strings) lives in the
//! shared `crate::scalar` leaf, not here.

// `strip_select_star_from` lives in the `sql` leaf (shared with preflight and
// plan::partition); re-exported here so `chunked`'s call sites and the
// `chunked::strip_select_star_from` re-export are unchanged.
pub(crate) use crate::sql::{strip_select_star_from, strip_simple_projection_from};

pub fn generate_chunks(min: i64, max: i64, chunk_size: i64) -> Vec<(i64, i64)> {
    if max < min || chunk_size <= 0 {
        return vec![];
    }
    let mut chunks = Vec::new();
    let mut start = min;
    loop {
        // Saturating arithmetic: when `start` is within `chunk_size` of i64::MAX
        // the naive `start + chunk_size - 1` would overflow and panic. We clamp
        // at i64::MAX and then `.min(max)` produces the correct inclusive end.
        let end = start.saturating_add(chunk_size - 1).min(max);
        chunks.push((start, end));
        // Two exit conditions: the chunk reached `max`, or `end` hit i64::MAX
        // (next start would overflow). Either way no more chunks remain.
        if end == max || end == i64::MAX {
            break;
        }
        start = end + 1;
    }
    chunks
}

/// Synthetic ordinal column for `chunk_dense`; stripped before writing files.
pub(crate) const RIVET_CHUNK_RN_COL: &str = "_rivet_chunk_rn";

pub(crate) fn build_chunk_query_sql(
    base_query: &str,
    order_column: &str,
    start: i64,
    end: i64,
    chunk_dense: bool,
    chunk_by_days: bool,
    source_type: crate::config::SourceType,
) -> String {
    let quoted_col = crate::sql::quote_ident(source_type, order_column);

    if chunk_dense {
        return format!(
            "SELECT * FROM (SELECT _rivet_i.*, ROW_NUMBER() OVER (ORDER BY _rivet_i.{oc}) AS {rn} FROM ({bq}) AS _rivet_i) AS _rivet_w WHERE _rivet_w.{rn} BETWEEN {s} AND {e}",
            bq = base_query,
            oc = quoted_col,
            rn = RIVET_CHUNK_RN_COL,
            s = start,
            e = end,
        );
    }

    // Fast path: when base_query is the `SELECT * FROM <simple-ident>` form
    // (typically produced by the `table:` YAML shortcut), append the WHERE
    // directly instead of wrapping it in a subquery. Three reasons:
    //
    // 1. The PG numeric catalog-hint parser only recognises the simple form,
    //    so declared `numeric(p,s)` types resolve automatically — wrapping
    //    in `SELECT * FROM (...) AS _rivet` would force every NUMERIC column
    //    to need a manual `columns:` override.
    // 2. PG's planner inlines the trivial subquery anyway, but skipping it
    //    keeps the EXPLAIN output (and chunk_plan_fingerprint) cleaner.
    // 3. The `table:` form is the only one we can reason about syntactically;
    //    arbitrary `query:` payloads still go through the subquery wrap so
    //    we don't accidentally rewrite the user's SQL.
    if let Some(table_ident) = strip_select_star_from(base_query) {
        if chunk_by_days {
            let epoch = chrono::NaiveDate::from_ymd_opt(1970, 1, 1).expect("epoch is valid");
            let start_date = epoch + chrono::Duration::days(start);
            let end_date = epoch + chrono::Duration::days(end + 1);
            return format!(
                "SELECT * FROM {table} WHERE {col} >= '{s}' AND {col} < '{e}'",
                table = table_ident,
                col = quoted_col,
                s = start_date.format("%Y-%m-%d"),
                e = end_date.format("%Y-%m-%d"),
            );
        }
        return format!(
            "SELECT * FROM {table} WHERE {col} BETWEEN {s} AND {e}",
            table = table_ident,
            col = quoted_col,
            s = start,
            e = end,
        );
    }

    if chunk_by_days {
        // start/end are days since epoch; end is inclusive so the exclusive upper bound is end+1.
        let epoch = chrono::NaiveDate::from_ymd_opt(1970, 1, 1).expect("epoch is valid");
        let start_date = epoch + chrono::Duration::days(start);
        let end_date = epoch + chrono::Duration::days(end + 1);
        return format!(
            "SELECT * FROM ({base}) AS _rivet WHERE {col} >= '{start}' AND {col} < '{end}'",
            base = base_query,
            col = quoted_col,
            start = start_date.format("%Y-%m-%d"),
            end = end_date.format("%Y-%m-%d"),
        );
    }

    format!(
        "SELECT * FROM ({base}) AS _rivet WHERE {col} BETWEEN {start} AND {end}",
        base = base_query,
        col = quoted_col,
        start = start,
        end = end,
    )
}

pub(crate) fn chunk_plan_fingerprint(
    base_query: &str,
    chunk_column: &str,
    chunk_size: usize,
    chunk_count: Option<usize>,
    chunk_dense: bool,
    chunk_by_days: Option<u32>,
) -> String {
    use xxhash_rust::xxh3::xxh3_64;
    let mut buf = String::with_capacity(base_query.len() + chunk_column.len() + 32);
    buf.push_str(base_query);
    buf.push('\x1f');
    buf.push_str(chunk_column);
    buf.push('\x1f');
    match chunk_by_days {
        Some(d) => buf.push_str(&format!("date_{}d", d)),
        None if chunk_dense => buf.push_str("dense_rn"),
        None if chunk_count.is_some() => buf.push_str(&format!("count_{}", chunk_count.unwrap())),
        None => {
            buf.push_str(&chunk_size.to_string());
            buf.push('\x1f');
            buf.push_str("range");
        }
    }
    format!("{:016x}", xxh3_64(buf.as_bytes()))
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_generate_chunks() {
        let chunks = generate_chunks(1, 100, 30);
        assert_eq!(chunks, vec![(1, 30), (31, 60), (61, 90), (91, 100)]);
    }

    #[test]
    fn test_generate_chunks_exact() {
        let chunks = generate_chunks(0, 99, 50);
        assert_eq!(chunks, vec![(0, 49), (50, 99)]);
    }

    #[test]
    fn test_generate_chunks_single() {
        let chunks = generate_chunks(1, 10, 100);
        assert_eq!(chunks, vec![(1, 10)]);
    }

    #[test]
    fn test_generate_chunks_empty() {
        assert!(generate_chunks(10, 5, 100).is_empty());
    }

    #[test]
    fn test_build_chunk_query_range_mode() {
        let q = build_chunk_query_sql(
            "SELECT id FROM t",
            "id",
            1,
            100,
            false,
            false,
            crate::config::SourceType::Postgres,
        );
        // Column name is quoted; numeric bounds are not
        assert!(q.contains("WHERE \"id\" BETWEEN 1 AND 100"), "got: {}", q);
        assert!(!q.contains("ROW_NUMBER()"), "got: {}", q);
    }

    // ── Fast-path: `SELECT * FROM <ident>` is rewritten in place ──────────

    #[test]
    fn chunk_query_fast_path_select_star_from_simple_table() {
        let q = build_chunk_query_sql(
            "SELECT * FROM public.users",
            "id",
            1,
            100,
            false,
            false,
            crate::config::SourceType::Postgres,
        );
        // Fast path: no subquery wrapper — the catalog-hint parser can see
        // `FROM public.users` and resolve declared numeric(p,s) types.
        assert_eq!(
            q,
            "SELECT * FROM public.users WHERE \"id\" BETWEEN 1 AND 100"
        );
    }

    #[test]
    fn chunk_query_fast_path_unqualified_table() {
        let q = build_chunk_query_sql(
            "SELECT * FROM orders",
            "id",
            1,
            10,
            false,
            false,
            crate::config::SourceType::Postgres,
        );
        assert_eq!(q, "SELECT * FROM orders WHERE \"id\" BETWEEN 1 AND 10");
    }

    #[test]
    fn chunk_query_wraps_when_query_is_not_select_star() {
        // Anything that's not the exact `SELECT * FROM <ident>` form falls
        // back to the subquery wrap.
        for not_fast in [
            "SELECT id, name FROM users",
            "SELECT * FROM users u",
            "SELECT * FROM users WHERE deleted_at IS NULL",
            "SELECT * FROM users JOIN orders ON users.id = orders.user_id",
            "SELECT * FROM (SELECT * FROM users) sub",
        ] {
            let q = build_chunk_query_sql(
                not_fast,
                "id",
                1,
                10,
                false,
                false,
                crate::config::SourceType::Postgres,
            );
            assert!(
                q.starts_with("SELECT * FROM ("),
                "fallback wrap expected for `{not_fast}`; got: {q}"
            );
        }
    }

    #[test]
    fn chunk_query_fast_path_by_days() {
        let q = build_chunk_query_sql(
            "SELECT * FROM public.events",
            "created_at",
            18000, // 2019-04-14
            18001,
            false,
            true,
            crate::config::SourceType::Postgres,
        );
        assert!(
            q.starts_with("SELECT * FROM public.events WHERE \"created_at\" >= '"),
            "expected fast-path date range; got: {q}"
        );
        assert!(
            !q.contains("AS _rivet"),
            "fast path should not wrap in subquery; got: {q}"
        );
    }

    #[test]
    fn test_build_chunk_query_range_mode_mysql() {
        let q = build_chunk_query_sql(
            "SELECT id FROM t",
            "id",
            1,
            100,
            false,
            false,
            crate::config::SourceType::Mysql,
        );
        assert!(q.contains("WHERE `id` BETWEEN 1 AND 100"), "got: {}", q);
    }

    #[test]
    fn test_build_chunk_query_dense_mode() {
        let q = build_chunk_query_sql(
            "SELECT id FROM t",
            "id",
            1,
            5000,
            true,
            false,
            crate::config::SourceType::Postgres,
        );
        assert!(q.contains("ROW_NUMBER()"), "got: {}", q);
        assert!(q.contains(RIVET_CHUNK_RN_COL), "got: {}", q);
        assert!(q.contains("BETWEEN 1 AND 5000"), "got: {}", q);
    }

    // `parse_date_flexible` and `parse_scalar_i64` now live in the shared
    // `crate::scalar` leaf; their tests moved there too.

    #[test]
    fn test_build_chunk_query_date_mode() {
        // days since epoch: 2023-01-01 = 19358, 2023-01-07 = 19364
        // chunk start=19358, end=19364 → >= '2023-01-01' AND < '2023-01-08'
        let epoch = chrono::NaiveDate::from_ymd_opt(1970, 1, 1).unwrap();
        let start = (chrono::NaiveDate::from_ymd_opt(2023, 1, 1).unwrap() - epoch).num_days();
        let end = (chrono::NaiveDate::from_ymd_opt(2023, 1, 7).unwrap() - epoch).num_days();
        let q = build_chunk_query_sql(
            "SELECT * FROM orders",
            "created_at",
            start,
            end,
            false,
            true,
            crate::config::SourceType::Postgres,
        );
        assert!(q.contains(">= '2023-01-01'"), "got: {q}");
        assert!(q.contains("< '2023-01-08'"), "got: {q}"); // end+1 day exclusive
        assert!(!q.contains("BETWEEN"), "should use >= AND <, got: {q}");
    }

    #[test]
    fn test_build_chunk_query_date_mode_single_day() {
        let epoch = chrono::NaiveDate::from_ymd_opt(1970, 1, 1).unwrap();
        let day = (chrono::NaiveDate::from_ymd_opt(2024, 3, 15).unwrap() - epoch).num_days();
        let q = build_chunk_query_sql(
            "SELECT * FROM t",
            "ts",
            day,
            day,
            false,
            true,
            crate::config::SourceType::Postgres,
        );
        assert!(q.contains(">= '2024-03-15'"), "got: {q}");
        assert!(q.contains("< '2024-03-16'"), "got: {q}");
    }

    // ─── Property-style invariants (QA backlog Task 3.2) ────────────────────
    //
    // Core invariants checked on every non-empty `generate_chunks` output.
    // Encoded as a helper + wide grid sweep so every boundary condition
    // the planner encounters in practice is exercised deterministically.

    fn assert_chunk_invariants(min: i64, max: i64, size: i64, chunks: &[(i64, i64)]) {
        if max < min || size <= 0 {
            assert!(chunks.is_empty(), "invalid input must yield []: {chunks:?}");
            return;
        }

        assert!(
            !chunks.is_empty(),
            "valid range must yield at least one chunk"
        );
        assert_eq!(chunks.first().unwrap().0, min, "first chunk starts at min");
        assert_eq!(chunks.last().unwrap().1, max, "last chunk ends at max");

        for (i, (s, e)) in chunks.iter().enumerate() {
            assert!(s <= e, "chunk {i} well-formed");
            if i > 0 {
                let (_, pe) = chunks[i - 1];
                assert_eq!(*s, pe + 1, "chunks are adjacent, no gap or overlap");
            }
        }

        for (i, (s, e)) in chunks.iter().enumerate() {
            let len = e - s + 1;
            if i + 1 < chunks.len() {
                assert_eq!(len, size, "non-final chunks have exactly size elements");
            } else {
                assert!(len >= 1 && len <= size, "final chunk is 1..=size");
            }
        }

        let total: i64 = chunks.iter().map(|(s, e)| e - s + 1).sum();
        assert_eq!(total, max - min + 1, "sum of chunks equals total range");
    }

    #[test]
    fn generate_chunks_invariants_over_grid() {
        let mins: [i64; 6] = [-50, -1, 0, 1, 100, 1_000_000];
        let sizes: [i64; 8] = [1, 2, 3, 7, 10, 33, 100, 10_000];
        for &min in &mins {
            for &size in &sizes {
                for delta in 0..=(4 * size) {
                    let max = min + delta - 1;
                    let chunks = generate_chunks(min, max, size);
                    assert_chunk_invariants(min, max, size, &chunks);
                }
            }
        }
    }

    #[test]
    fn generate_chunks_empty_on_invalid_inputs() {
        assert!(generate_chunks(10, 5, 10).is_empty());
        assert!(generate_chunks(1, 100, 0).is_empty());
        assert!(generate_chunks(1, 100, -5).is_empty());
    }

    #[test]
    fn generate_chunks_single_element_range() {
        assert_eq!(generate_chunks(42, 42, 100), vec![(42, 42)]);
        assert_eq!(generate_chunks(42, 42, 1), vec![(42, 42)]);
    }

    #[test]
    fn generate_chunks_chunk_size_larger_than_range() {
        let chunks = generate_chunks(10, 20, 10_000);
        assert_eq!(chunks, vec![(10, 20)]);
        assert_chunk_invariants(10, 20, 10_000, &chunks);
    }

    #[test]
    fn generate_chunks_boundary_exact_multiple() {
        let chunks = generate_chunks(1, 100, 25);
        assert_eq!(chunks, vec![(1, 25), (26, 50), (51, 75), (76, 100)]);
        assert_chunk_invariants(1, 100, 25, &chunks);
    }

    #[test]
    fn generate_chunks_boundary_off_by_one() {
        let chunks = generate_chunks(1, 101, 25);
        assert_eq!(chunks.last(), Some(&(101, 101)));
        assert_chunk_invariants(1, 101, 25, &chunks);
    }

    /// Regression for the i64::MAX overflow that crashed `start + chunk_size - 1`
    /// before the saturating fix. See QA backlog Task 3.2 / 4A.2.
    #[test]
    fn generate_chunks_does_not_overflow_on_near_i64_max() {
        let min = i64::MAX - 10;
        let max = i64::MAX;
        let chunks = generate_chunks(min, max, 3);
        assert_chunk_invariants(min, max, 3, &chunks);
    }

    /// Extreme boundary combinations: each triple must either succeed or
    /// terminate immediately (empty).  Never panic.  QA backlog Task 4A.2.
    #[test]
    fn generate_chunks_does_not_panic_on_extreme_boundaries() {
        // Every entry generates at most a small bounded number of chunks,
        // otherwise this test would hang.  "Wide range × tiny chunk_size" is
        // out of scope for a fuzz smoke — it needs a planner-level cap.
        let triples: &[(i64, i64, i64)] = &[
            (0, 0, 1),
            (i64::MIN, i64::MIN + 1, 1),
            (i64::MIN, i64::MIN, 1),
            (i64::MIN, i64::MAX, i64::MAX),
            (i64::MAX - 2, i64::MAX, 1),
            (i64::MAX - 2, i64::MAX, 10),
            (i64::MAX - 2, i64::MAX, i64::MAX),
            (-10, 10, 3),
            (0, 0, i64::MAX),
            (0, 0, i64::MIN),
            (5, 5, -1),
        ];
        for &(min, max, size) in triples {
            let _ = generate_chunks(min, max, size);
        }
    }

    // ── chunk_plan_fingerprint ────────────────────────────────────────────────

    fn fp(
        query: &str,
        col: &str,
        size: usize,
        count: Option<usize>,
        dense: bool,
        by_days: Option<u32>,
    ) -> String {
        chunk_plan_fingerprint(query, col, size, count, dense, by_days)
    }

    #[test]
    fn fingerprint_changes_when_chunk_count_changes() {
        let base = fp("SELECT * FROM t", "id", 10_000, None, false, None);
        let with_count = fp("SELECT * FROM t", "id", 10_000, Some(10), false, None);
        assert_ne!(base, with_count, "chunk_count=None vs Some(10) must differ");
    }

    #[test]
    fn fingerprint_changes_when_chunk_count_value_changes() {
        let c5 = fp("SELECT * FROM t", "id", 10_000, Some(5), false, None);
        let c10 = fp("SELECT * FROM t", "id", 10_000, Some(10), false, None);
        assert_ne!(c5, c10, "chunk_count=5 vs chunk_count=10 must differ");
    }

    #[test]
    fn fingerprint_chunk_count_ignores_chunk_size() {
        // When chunk_count is set, chunk_size is not encoded in the fingerprint
        // (it's computed dynamically from min/max). Two configs with different
        // chunk_size but same chunk_count must produce the same fingerprint.
        let a = fp("SELECT * FROM t", "id", 1_000, Some(5), false, None);
        let b = fp("SELECT * FROM t", "id", 50_000, Some(5), false, None);
        assert_eq!(
            a, b,
            "chunk_size is not part of the fingerprint when chunk_count is set"
        );
    }

    #[test]
    fn fingerprint_chunk_count_differs_from_dense_and_by_days() {
        let count = fp("SELECT * FROM t", "id", 100, Some(4), false, None);
        let dense = fp("SELECT * FROM t", "id", 100, None, true, None);
        let by_days = fp("SELECT * FROM t", "id", 100, None, false, Some(7));
        assert_ne!(count, dense);
        assert_ne!(count, by_days);
        assert_ne!(dense, by_days);
    }

    #[test]
    fn fingerprint_deterministic_for_same_inputs() {
        let a = fp("SELECT id FROM orders", "id", 5_000, Some(8), false, None);
        let b = fp("SELECT id FROM orders", "id", 5_000, Some(8), false, None);
        assert_eq!(a, b);
    }
}