relay-knowledge 1.1.10

Graph-database-based knowledge graph project.
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
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
use std::{thread, time::Duration};

use rusqlite::{Connection, params};

use crate::storage::StorageError;

struct DerivedTableSchema {
    table: &'static str,
    required_columns: &'static [&'static str],
}

#[derive(Debug)]
struct TableColumn {
    name: String,
    not_null: bool,
    default_value: Option<String>,
}

const INDEX_REFRESH_TASK_COLUMNS: &[&str] = &[
    "task_id",
    "kind",
    "source_scope",
    "modality",
    "target_graph_version",
    "state",
    "lease_owner",
    "lease_expires_at_ms",
    "attempt_count",
    "next_retry_at_ms",
    "input_fingerprint",
    "cursor_before",
    "cursor_after",
    "last_error_kind",
    "last_error_message",
    "created_at_ms",
    "updated_at_ms",
];
const LEGACY_NEXT_RETRY_AFTER_COLUMN: &str = "next_retry_after_ms";
const SCHEMA_COMPATIBILITY_RETRY_DELAYS_MS: [u64; 5] = [10, 30, 90, 270, 810];

const GRAPH_BM25_COLUMNS: &[&str] = &[
    "document_id",
    "document_kind",
    "evidence_id",
    "parent_evidence_id",
    "modality",
    "created_graph_version",
    "source_scope",
    "source_path",
    "entity_labels",
    "entity_aliases",
    "content",
];

const GRAPH_SEMANTIC_COLUMNS: &[&str] = &[
    "document_id",
    "document_kind",
    "evidence_id",
    "parent_evidence_id",
    "modality",
    "created_graph_version",
    "source_scope",
    "source_path",
    "entity_labels_json",
    "content",
    "token_signature_json",
    "model",
    "dimension",
    "source_hash",
    "tokenizer_version",
];

const GRAPH_VECTOR_COLUMNS: &[&str] = &[
    "document_id",
    "document_kind",
    "evidence_id",
    "parent_evidence_id",
    "modality",
    "created_graph_version",
    "source_scope",
    "source_path",
    "entity_labels_json",
    "content",
    "vector_json",
    "model",
    "dimension",
    "source_hash",
    "tokenizer_version",
];
const CODE_WORKSPACE_PACKAGE_MAPPING_COLUMNS: &[&str] = &[
    "set_id",
    "package_name",
    "ecosystem",
    "repository_id",
    "source_scope",
    "workspace_format",
    "created_at_ms",
];
const CODE_WORKSPACE_PACKAGE_MAPPING_UNIQUE: &[&str] = &["set_id", "package_name", "ecosystem"];

const CODE_GRAPH_SCHEMAS: &[DerivedTableSchema] = &[
    DerivedTableSchema {
        table: "code_files",
        required_columns: &[
            "source_scope",
            "path",
            "content_hash",
            "language_id",
            "parse_status",
            "diagnostic",
            "created_graph_version",
        ],
    },
    DerivedTableSchema {
        table: "code_symbols",
        required_columns: &[
            "source_scope",
            "path",
            "symbol_id",
            "name",
            "kind",
            "start_byte",
            "end_byte",
            "start_line",
            "end_line",
            "grammar_version",
            "query_name",
            "query_version",
            "node_kind",
            "capture_kind",
            "created_graph_version",
        ],
    },
    DerivedTableSchema {
        table: "code_references",
        required_columns: &[
            "source_scope",
            "path",
            "reference_id",
            "symbol_text",
            "kind",
            "start_byte",
            "end_byte",
            "start_line",
            "end_line",
            "resolution_state",
            "target_symbol_id",
            "grammar_version",
            "query_name",
            "query_version",
            "node_kind",
            "capture_kind",
            "created_graph_version",
        ],
    },
    DerivedTableSchema {
        table: "code_chunks",
        required_columns: &[
            "source_scope",
            "path",
            "chunk_id",
            "content",
            "start_byte",
            "end_byte",
            "start_line",
            "end_line",
            "grammar_version",
            "query_name",
            "query_version",
            "node_kind",
            "capture_kind",
            "created_graph_version",
        ],
    },
    DerivedTableSchema {
        table: "code_chunk_symbols",
        required_columns: &["source_scope", "path", "chunk_id", "symbol_id"],
    },
];

pub(super) fn prepare_existing_database(connection: &Connection) -> Result<(), StorageError> {
    for delay_ms in SCHEMA_COMPATIBILITY_RETRY_DELAYS_MS {
        match prepare_existing_database_once(connection) {
            Ok(()) => return Ok(()),
            Err(error) if schema_compatibility_error_is_retryable(&error) => {
                thread::sleep(Duration::from_millis(delay_ms));
            }
            Err(error) => return Err(error),
        }
    }

    prepare_existing_database_once(connection)
}

fn prepare_existing_database_once(connection: &Connection) -> Result<(), StorageError> {
    drop_incompatible_table(connection, "graph_bm25", GRAPH_BM25_COLUMNS)?;
    drop_incompatible_table(
        connection,
        "graph_semantic_documents",
        GRAPH_SEMANTIC_COLUMNS,
    )?;
    drop_incompatible_table(connection, "graph_vector_documents", GRAPH_VECTOR_COLUMNS)?;
    rebuild_incompatible_code_graph_tables(connection)?;
    rebuild_incompatible_index_refresh_tasks(connection)?;
    rebuild_incompatible_workspace_package_mappings(connection)?;

    Ok(())
}

fn schema_compatibility_error_is_retryable(error: &StorageError) -> bool {
    match error {
        StorageError::Sqlite(error) => {
            schema_compatibility_error_message_is_retryable(&error.to_string())
        }
        _ => false,
    }
}

fn schema_compatibility_error_message_is_retryable(message: &str) -> bool {
    message.contains("vtable constructor failed: graph_bm25")
        || message.contains("database schema is locked")
        || message.contains("database is locked")
}

fn rebuild_incompatible_index_refresh_tasks(connection: &Connection) -> Result<(), StorageError> {
    if !table_exists(connection, "index_refresh_tasks")? {
        return Ok(());
    }
    let columns = table_column_info(connection, "index_refresh_tasks")?;
    if !index_refresh_tasks_needs_rebuild(&columns) {
        return Ok(());
    }

    let select_expressions = index_refresh_task_select_expressions(&columns);
    let insert_columns = INDEX_REFRESH_TASK_COLUMNS.join(", ");
    let select_columns = select_expressions.join(", ");
    let migration = format!(
        "
        DROP TABLE IF EXISTS index_refresh_tasks_rebuild_legacy;
        ALTER TABLE index_refresh_tasks RENAME TO index_refresh_tasks_rebuild_legacy;
        CREATE TABLE index_refresh_tasks (
            task_id TEXT PRIMARY KEY,
            kind TEXT NOT NULL,
            source_scope TEXT NOT NULL,
            modality TEXT NOT NULL,
            target_graph_version INTEGER NOT NULL,
            state TEXT NOT NULL,
            lease_owner TEXT,
            lease_expires_at_ms INTEGER,
            attempt_count INTEGER NOT NULL,
            next_retry_at_ms INTEGER NOT NULL,
            input_fingerprint TEXT NOT NULL,
            cursor_before INTEGER NOT NULL,
            cursor_after INTEGER,
            last_error_kind TEXT,
            last_error_message TEXT,
            created_at_ms INTEGER NOT NULL,
            updated_at_ms INTEGER NOT NULL
        );
        INSERT INTO index_refresh_tasks ({insert_columns})
        SELECT {select_columns}
        FROM index_refresh_tasks_rebuild_legacy;
        DROP TABLE index_refresh_tasks_rebuild_legacy;
        ",
    );

    connection.execute_batch("BEGIN IMMEDIATE")?;
    let result = connection
        .execute_batch(&migration)
        .map_err(StorageError::from);
    match result {
        Ok(()) => connection
            .execute_batch("COMMIT")
            .map_err(StorageError::from),
        Err(error) => {
            let _ = connection.execute_batch("ROLLBACK");
            Err(error)
        }
    }
}

fn index_refresh_tasks_needs_rebuild(columns: &[TableColumn]) -> bool {
    columns.iter().any(|column| {
        column.name == LEGACY_NEXT_RETRY_AFTER_COLUMN
            || (!INDEX_REFRESH_TASK_COLUMNS.contains(&column.name.as_str())
                && column.not_null
                && column.default_value.is_none())
    })
}

fn index_refresh_task_select_expressions(columns: &[TableColumn]) -> Vec<String> {
    let now_ms = "CAST(strftime('%s', 'now') AS INTEGER) * 1000";
    let fingerprint =
        "kind || ':' || source_scope || ':' || modality || ':' || target_graph_version";
    let created_at_ms = timestamp_expression(columns, "created_at_ms", now_ms);

    vec![
        column_or(
            columns,
            "task_id",
            "'legacy-index-refresh:' || lower(hex(randomblob(8)))",
        ),
        column_or(columns, "kind", "'bm25'"),
        column_or(columns, "source_scope", "'graph'"),
        column_or(columns, "modality", "'text'"),
        column_or(columns, "target_graph_version", "0"),
        column_or(columns, "state", "'queued'"),
        column_or(columns, "lease_owner", "NULL"),
        column_or(columns, "lease_expires_at_ms", "NULL"),
        column_or(columns, "attempt_count", "0"),
        retry_at_expression(columns),
        if has_column(columns, "input_fingerprint") {
            format!("COALESCE(NULLIF(input_fingerprint, ''), {fingerprint})")
        } else {
            fingerprint.to_owned()
        },
        column_or(columns, "cursor_before", "0"),
        column_or(columns, "cursor_after", "NULL"),
        column_or(columns, "last_error_kind", "NULL"),
        column_or(columns, "last_error_message", "NULL"),
        created_at_ms.clone(),
        if has_column(columns, "updated_at_ms") {
            timestamp_expression(columns, "updated_at_ms", now_ms)
        } else {
            created_at_ms
        },
    ]
}

fn retry_at_expression(columns: &[TableColumn]) -> String {
    if has_column(columns, "next_retry_at_ms")
        && has_column(columns, LEGACY_NEXT_RETRY_AFTER_COLUMN)
    {
        format!(
            "CASE WHEN next_retry_at_ms IS NOT NULL AND next_retry_at_ms != 0 \
             THEN next_retry_at_ms \
             ELSE COALESCE({LEGACY_NEXT_RETRY_AFTER_COLUMN}, 0) END"
        )
    } else if has_column(columns, "next_retry_at_ms") {
        "COALESCE(next_retry_at_ms, 0)".to_owned()
    } else if has_column(columns, LEGACY_NEXT_RETRY_AFTER_COLUMN) {
        format!("COALESCE({LEGACY_NEXT_RETRY_AFTER_COLUMN}, 0)")
    } else {
        "0".to_owned()
    }
}

fn timestamp_expression(columns: &[TableColumn], column: &str, now_ms: &str) -> String {
    if has_column(columns, column) {
        format!("CASE WHEN {column} IS NULL OR {column} = 0 THEN {now_ms} ELSE {column} END")
    } else {
        now_ms.to_owned()
    }
}

fn column_or(columns: &[TableColumn], column: &str, fallback: &str) -> String {
    if has_column(columns, column) {
        column.to_owned()
    } else {
        fallback.to_owned()
    }
}

fn has_column(columns: &[TableColumn], expected: &str) -> bool {
    columns.iter().any(|column| column.name == expected)
}

fn drop_incompatible_table(
    connection: &Connection,
    table: &str,
    required_columns: &[&str],
) -> Result<(), StorageError> {
    if table_exists(connection, table)? && !table_has_columns(connection, table, required_columns)?
    {
        connection.execute(&format!("DROP TABLE {table}"), [])?;
    }

    Ok(())
}

fn rebuild_incompatible_code_graph_tables(connection: &Connection) -> Result<(), StorageError> {
    let any_code_graph_table_exists = CODE_GRAPH_SCHEMAS
        .iter()
        .map(|schema| table_exists(connection, schema.table))
        .collect::<Result<Vec<_>, _>>()?
        .into_iter()
        .any(|exists| exists);
    if !any_code_graph_table_exists {
        return Ok(());
    }

    let incompatible = CODE_GRAPH_SCHEMAS
        .iter()
        .map(|schema| {
            Ok(table_exists(connection, schema.table)?
                && !table_has_columns(connection, schema.table, schema.required_columns)?)
        })
        .collect::<Result<Vec<_>, StorageError>>()?
        .into_iter()
        .any(|value| value);
    if !incompatible {
        return Ok(());
    }

    for table in [
        "code_chunk_symbols",
        "code_chunks",
        "code_references",
        "code_symbols",
        "code_files",
    ] {
        if table_exists(connection, table)? {
            connection.execute(&format!("DROP TABLE {table}"), [])?;
        }
    }

    Ok(())
}

fn rebuild_incompatible_workspace_package_mappings(
    connection: &Connection,
) -> Result<(), StorageError> {
    if !table_exists(connection, "code_workspace_package_mappings")? {
        return Ok(());
    }
    if !table_has_columns(
        connection,
        "code_workspace_package_mappings",
        CODE_WORKSPACE_PACKAGE_MAPPING_COLUMNS,
    )? {
        connection.execute("DROP TABLE code_workspace_package_mappings", [])?;
        return Ok(());
    }
    if table_has_unique_columns(
        connection,
        "code_workspace_package_mappings",
        CODE_WORKSPACE_PACKAGE_MAPPING_UNIQUE,
    )? {
        return Ok(());
    }

    connection.execute_batch(
        "
        DROP TABLE IF EXISTS code_workspace_package_mappings_rebuild_legacy;
        ALTER TABLE code_workspace_package_mappings
            RENAME TO code_workspace_package_mappings_rebuild_legacy;
        CREATE TABLE code_workspace_package_mappings (
            id INTEGER PRIMARY KEY AUTOINCREMENT,
            set_id TEXT NOT NULL,
            package_name TEXT NOT NULL,
            ecosystem TEXT NOT NULL,
            repository_id TEXT NOT NULL,
            source_scope TEXT NOT NULL,
            workspace_format TEXT NOT NULL,
            created_at_ms INTEGER NOT NULL,
            UNIQUE (set_id, package_name, ecosystem)
        );
        INSERT OR IGNORE INTO code_workspace_package_mappings (
            id, set_id, package_name, ecosystem, repository_id, source_scope,
            workspace_format, created_at_ms
        )
        SELECT id, set_id, package_name, ecosystem, repository_id, source_scope,
               workspace_format, created_at_ms
        FROM code_workspace_package_mappings_rebuild_legacy
        ORDER BY created_at_ms DESC, id DESC;
        DROP TABLE code_workspace_package_mappings_rebuild_legacy;
        ",
    )?;
    Ok(())
}

fn table_has_columns(
    connection: &Connection,
    table: &str,
    required_columns: &[&str],
) -> Result<bool, StorageError> {
    let columns = table_columns(connection, table)?;

    Ok(required_columns
        .iter()
        .all(|required| columns.iter().any(|column| column == required)))
}

fn table_columns(connection: &Connection, table: &str) -> Result<Vec<String>, StorageError> {
    let mut statement = connection.prepare(&format!("PRAGMA table_info({table})"))?;
    let rows = statement.query_map([], |row| row.get::<_, String>(1))?;

    rows.collect::<Result<Vec<_>, _>>()
        .map_err(StorageError::from)
}

fn table_has_unique_columns(
    connection: &Connection,
    table: &str,
    expected_columns: &[&str],
) -> Result<bool, StorageError> {
    let mut statement = connection.prepare(&format!("PRAGMA index_list({table})"))?;
    let rows = statement.query_map([], |row| {
        Ok((row.get::<_, String>(1)?, row.get::<_, i64>(2)? != 0))
    })?;
    let indexes = rows.collect::<Result<Vec<_>, _>>()?;

    for (index_name, unique) in indexes {
        if !unique {
            continue;
        }
        let mut statement = connection.prepare(&format!("PRAGMA index_info({index_name})"))?;
        let rows = statement.query_map([], |row| row.get::<_, String>(2))?;
        let columns = rows.collect::<Result<Vec<_>, _>>()?;
        if columns
            .iter()
            .map(String::as_str)
            .eq(expected_columns.iter().copied())
        {
            return Ok(true);
        }
    }

    Ok(false)
}

fn table_column_info(
    connection: &Connection,
    table: &str,
) -> Result<Vec<TableColumn>, StorageError> {
    let mut statement = connection.prepare(&format!("PRAGMA table_info({table})"))?;
    let rows = statement.query_map([], |row| {
        Ok(TableColumn {
            name: row.get(1)?,
            not_null: row.get(3)?,
            default_value: row.get(4)?,
        })
    })?;

    rows.collect::<Result<Vec<_>, _>>()
        .map_err(StorageError::from)
}

fn table_exists(connection: &Connection, table: &str) -> Result<bool, StorageError> {
    connection
        .query_row(
            "
            SELECT EXISTS (
                SELECT 1 FROM sqlite_master
                WHERE type = 'table' AND name = ?1
            )
            ",
            params![table],
            |row| row.get::<_, bool>(0),
        )
        .map_err(StorageError::from)
}

#[cfg(test)]
mod tests {
    use super::{
        CODE_WORKSPACE_PACKAGE_MAPPING_UNIQUE, prepare_existing_database_once,
        schema_compatibility_error_is_retryable, schema_compatibility_error_message_is_retryable,
        table_has_unique_columns,
    };
    use crate::storage::StorageError;
    use rusqlite::Connection;

    #[test]
    fn schema_compatibility_retry_is_limited_to_transient_open_errors() {
        assert!(schema_compatibility_error_message_is_retryable(
            "vtable constructor failed: graph_bm25"
        ));
        assert!(schema_compatibility_error_message_is_retryable(
            "database schema is locked"
        ));
        assert!(!schema_compatibility_error_message_is_retryable(
            "no such table: graph_bm25"
        ));
        assert!(!schema_compatibility_error_is_retryable(
            &StorageError::InvalidInput("database is locked".to_owned())
        ));
    }

    #[test]
    fn migration_rebuilds_legacy_workspace_package_mapping_uniqueness() {
        let connection = Connection::open_in_memory().expect("connection should open");
        connection
            .execute_batch(
                "
                CREATE TABLE code_workspace_package_mappings (
                    id INTEGER PRIMARY KEY AUTOINCREMENT,
                    set_id TEXT NOT NULL,
                    package_name TEXT NOT NULL,
                    ecosystem TEXT NOT NULL,
                    repository_id TEXT NOT NULL,
                    source_scope TEXT NOT NULL,
                    workspace_format TEXT NOT NULL,
                    created_at_ms INTEGER NOT NULL,
                    UNIQUE (set_id, package_name)
                );
                INSERT INTO code_workspace_package_mappings (
                    set_id, package_name, ecosystem, repository_id, source_scope,
                    workspace_format, created_at_ms
                )
                VALUES ('set-1', 'core', 'npm', 'repo', 'scope', 'pnpm', 1);
                ",
            )
            .expect("legacy mapping table should create");

        prepare_existing_database_once(&connection).expect("migration should run");

        assert!(
            table_has_unique_columns(
                &connection,
                "code_workspace_package_mappings",
                CODE_WORKSPACE_PACKAGE_MAPPING_UNIQUE
            )
            .expect("unique key should inspect")
        );
        let row_count: u32 = connection
            .query_row(
                "SELECT COUNT(*) FROM code_workspace_package_mappings",
                [],
                |row| row.get(0),
            )
            .expect("row count should load");
        assert_eq!(row_count, 1);
    }
}