Files
gitlore/src/core/db_tests.rs
Taylor Eernisse 7e0e6a91f2 refactor: extract unit tests into separate _tests.rs files
Move inline #[cfg(test)] mod tests { ... } blocks from 22 source files
into dedicated _tests.rs companion files, wired via:

    #[cfg(test)]
    #[path = "module_tests.rs"]
    mod tests;

This keeps implementation-focused source files leaner and more scannable
while preserving full access to private items through `use super::*;`.

Modules extracted:
  core:      db, note_parser, payloads, project, references, sync_run,
             timeline_collect, timeline_expand, timeline_seed
  cli:       list (55 tests), who (75 tests)
  documents: extractor (43 tests), regenerator
  embedding: change_detector, chunking
  gitlab:    graphql (wiremock async tests), transformers/issue
  ingestion: dirty_tracker, discussions, issues, mr_diffs

Also adds conflicts_with("explain_score") to the --detail flag in the
who command to prevent mutually exclusive flags from being combined.

All 629 unit tests pass. No behavior changes.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-13 10:54:02 -05:00

633 lines
19 KiB
Rust

use super::*;
fn setup_migrated_db() -> Connection {
let conn = create_connection(Path::new(":memory:")).unwrap();
run_migrations(&conn).unwrap();
conn
}
fn index_exists(conn: &Connection, index_name: &str) -> bool {
conn.query_row(
"SELECT COUNT(*) > 0 FROM sqlite_master WHERE type='index' AND name=?1",
[index_name],
|row| row.get(0),
)
.unwrap_or(false)
}
fn column_exists(conn: &Connection, table: &str, column: &str) -> bool {
let sql = format!("PRAGMA table_info({})", table);
let mut stmt = conn.prepare(&sql).unwrap();
let columns: Vec<String> = stmt
.query_map([], |row| row.get::<_, String>(1))
.unwrap()
.filter_map(|r| r.ok())
.collect();
columns.contains(&column.to_string())
}
#[test]
fn test_migration_022_indexes_exist() {
let conn = setup_migrated_db();
// New indexes from migration 022
assert!(
index_exists(&conn, "idx_notes_user_created"),
"idx_notes_user_created should exist"
);
assert!(
index_exists(&conn, "idx_notes_project_created"),
"idx_notes_project_created should exist"
);
assert!(
index_exists(&conn, "idx_notes_author_id"),
"idx_notes_author_id should exist"
);
// Discussion JOIN indexes (idx_discussions_issue_id is new;
// idx_discussions_mr_id already existed from migration 006 but
// IF NOT EXISTS makes it safe)
assert!(
index_exists(&conn, "idx_discussions_issue_id"),
"idx_discussions_issue_id should exist"
);
assert!(
index_exists(&conn, "idx_discussions_mr_id"),
"idx_discussions_mr_id should exist"
);
// author_id column on notes
assert!(
column_exists(&conn, "notes", "author_id"),
"notes.author_id column should exist"
);
}
// -- Helper: insert a minimal project for FK satisfaction --
fn insert_test_project(conn: &Connection) -> i64 {
conn.execute(
"INSERT INTO projects (gitlab_project_id, path_with_namespace, web_url) \
VALUES (1000, 'test/project', 'https://example.com/test/project')",
[],
)
.unwrap();
conn.last_insert_rowid()
}
// -- Helper: insert a minimal issue --
fn insert_test_issue(conn: &Connection, project_id: i64) -> i64 {
conn.execute(
"INSERT INTO issues (gitlab_id, project_id, iid, state, author_username, \
created_at, updated_at, last_seen_at) \
VALUES (100, ?1, 1, 'opened', 'alice', 1000, 1000, 1000)",
[project_id],
)
.unwrap();
conn.last_insert_rowid()
}
// -- Helper: insert a minimal discussion --
fn insert_test_discussion(conn: &Connection, project_id: i64, issue_id: i64) -> i64 {
conn.execute(
"INSERT INTO discussions (gitlab_discussion_id, project_id, issue_id, \
noteable_type, last_seen_at) \
VALUES ('disc-001', ?1, ?2, 'Issue', 1000)",
rusqlite::params![project_id, issue_id],
)
.unwrap();
conn.last_insert_rowid()
}
// -- Helper: insert a minimal non-system note --
#[allow(clippy::too_many_arguments)]
fn insert_test_note(
conn: &Connection,
gitlab_id: i64,
discussion_id: i64,
project_id: i64,
is_system: bool,
) -> i64 {
conn.execute(
"INSERT INTO notes (gitlab_id, discussion_id, project_id, is_system, \
author_username, body, created_at, updated_at, last_seen_at) \
VALUES (?1, ?2, ?3, ?4, 'alice', 'note body', 1000, 1000, 1000)",
rusqlite::params![gitlab_id, discussion_id, project_id, is_system as i32],
)
.unwrap();
conn.last_insert_rowid()
}
// -- Helper: insert a document --
fn insert_test_document(
conn: &Connection,
source_type: &str,
source_id: i64,
project_id: i64,
) -> i64 {
conn.execute(
"INSERT INTO documents (source_type, source_id, project_id, content_text, content_hash) \
VALUES (?1, ?2, ?3, 'test content', 'hash123')",
rusqlite::params![source_type, source_id, project_id],
)
.unwrap();
conn.last_insert_rowid()
}
#[test]
fn test_migration_024_allows_note_source_type() {
let conn = setup_migrated_db();
let pid = insert_test_project(&conn);
// Should succeed -- 'note' is now allowed
conn.execute(
"INSERT INTO documents (source_type, source_id, project_id, content_text, content_hash) \
VALUES ('note', 1, ?1, 'note content', 'hash-note')",
[pid],
)
.expect("INSERT with source_type='note' into documents should succeed");
// dirty_sources should also accept 'note'
conn.execute(
"INSERT INTO dirty_sources (source_type, source_id, queued_at) \
VALUES ('note', 1, 1000)",
[],
)
.expect("INSERT with source_type='note' into dirty_sources should succeed");
}
#[test]
fn test_migration_024_preserves_existing_data() {
// Run migrations up to 023 only, insert data, then apply 024
// Migration 024 is at index 23 (0-based). Use hardcoded index so adding
// later migrations doesn't silently shift what this test exercises.
let conn = create_connection(Path::new(":memory:")).unwrap();
// Apply migrations 001-023 (indices 0..23)
run_migrations_up_to(&conn, 23);
let pid = insert_test_project(&conn);
// Insert a document with existing source_type
conn.execute(
"INSERT INTO documents (source_type, source_id, project_id, content_text, content_hash, title) \
VALUES ('issue', 1, ?1, 'issue content', 'hash-issue', 'Test Issue')",
[pid],
)
.unwrap();
let doc_id: i64 = conn.last_insert_rowid();
// Insert junction data
conn.execute(
"INSERT INTO document_labels (document_id, label_name) VALUES (?1, 'bug')",
[doc_id],
)
.unwrap();
conn.execute(
"INSERT INTO document_paths (document_id, path) VALUES (?1, 'src/main.rs')",
[doc_id],
)
.unwrap();
// Insert dirty_sources row
conn.execute(
"INSERT INTO dirty_sources (source_type, source_id, queued_at) VALUES ('issue', 1, 1000)",
[],
)
.unwrap();
// Now apply migration 024 (index 23) -- the table-rebuild migration
run_single_migration(&conn, 23);
// Verify document still exists with correct data
let (st, content, title): (String, String, String) = conn
.query_row(
"SELECT source_type, content_text, title FROM documents WHERE id = ?1",
[doc_id],
|row| Ok((row.get(0)?, row.get(1)?, row.get(2)?)),
)
.unwrap();
assert_eq!(st, "issue");
assert_eq!(content, "issue content");
assert_eq!(title, "Test Issue");
// Verify junction data preserved
let label_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM document_labels WHERE document_id = ?1",
[doc_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(label_count, 1);
let path_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM document_paths WHERE document_id = ?1",
[doc_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(path_count, 1);
// Verify dirty_sources preserved
let dirty_count: i64 = conn
.query_row("SELECT COUNT(*) FROM dirty_sources", [], |row| row.get(0))
.unwrap();
assert_eq!(dirty_count, 1);
}
#[test]
fn test_migration_024_fts_triggers_intact() {
let conn = setup_migrated_db();
let pid = insert_test_project(&conn);
// Insert a document after migration -- FTS trigger should fire
let doc_id = insert_test_document(&conn, "note", 1, pid);
// Verify FTS entry exists
let fts_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents_fts WHERE documents_fts MATCH 'test'",
[],
|row| row.get(0),
)
.unwrap();
assert!(fts_count > 0, "FTS trigger should have created an entry");
// Verify update trigger works
conn.execute(
"UPDATE documents SET content_text = 'updated content' WHERE id = ?1",
[doc_id],
)
.unwrap();
let fts_updated: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents_fts WHERE documents_fts MATCH 'updated'",
[],
|row| row.get(0),
)
.unwrap();
assert!(
fts_updated > 0,
"FTS update trigger should reflect new content"
);
// Verify delete trigger works
conn.execute("DELETE FROM documents WHERE id = ?1", [doc_id])
.unwrap();
let fts_after_delete: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents_fts WHERE documents_fts MATCH 'updated'",
[],
|row| row.get(0),
)
.unwrap();
assert_eq!(
fts_after_delete, 0,
"FTS delete trigger should remove the entry"
);
}
#[test]
fn test_migration_024_row_counts_preserved() {
let conn = setup_migrated_db();
// After full migration, tables should exist and be queryable
let doc_count: i64 = conn
.query_row("SELECT COUNT(*) FROM documents", [], |row| row.get(0))
.unwrap();
assert_eq!(doc_count, 0, "Fresh DB should have 0 documents");
let dirty_count: i64 = conn
.query_row("SELECT COUNT(*) FROM dirty_sources", [], |row| row.get(0))
.unwrap();
assert_eq!(dirty_count, 0, "Fresh DB should have 0 dirty_sources");
}
#[test]
fn test_migration_024_integrity_checks_pass() {
let conn = setup_migrated_db();
// PRAGMA integrity_check
let integrity: String = conn
.query_row("PRAGMA integrity_check", [], |row| row.get(0))
.unwrap();
assert_eq!(integrity, "ok", "Database integrity check should pass");
// PRAGMA foreign_key_check (returns rows only if there are violations)
let fk_violations: i64 = conn
.query_row("SELECT COUNT(*) FROM pragma_foreign_key_check", [], |row| {
row.get(0)
})
.unwrap();
assert_eq!(fk_violations, 0, "No foreign key violations should exist");
}
#[test]
fn test_migration_024_note_delete_trigger_cleans_document() {
let conn = setup_migrated_db();
let pid = insert_test_project(&conn);
let issue_id = insert_test_issue(&conn, pid);
let disc_id = insert_test_discussion(&conn, pid, issue_id);
let note_id = insert_test_note(&conn, 200, disc_id, pid, false);
// Create a document for this note
insert_test_document(&conn, "note", note_id, pid);
let doc_before: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents WHERE source_type = 'note' AND source_id = ?1",
[note_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(doc_before, 1);
// Delete the note -- trigger should remove the document
conn.execute("DELETE FROM notes WHERE id = ?1", [note_id])
.unwrap();
let doc_after: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents WHERE source_type = 'note' AND source_id = ?1",
[note_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(
doc_after, 0,
"notes_ad_cleanup trigger should delete the document"
);
}
#[test]
fn test_migration_024_note_system_flip_trigger_cleans_document() {
let conn = setup_migrated_db();
let pid = insert_test_project(&conn);
let issue_id = insert_test_issue(&conn, pid);
let disc_id = insert_test_discussion(&conn, pid, issue_id);
let note_id = insert_test_note(&conn, 201, disc_id, pid, false);
// Create a document for this note
insert_test_document(&conn, "note", note_id, pid);
let doc_before: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents WHERE source_type = 'note' AND source_id = ?1",
[note_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(doc_before, 1);
// Flip is_system from 0 to 1 -- trigger should remove the document
conn.execute("UPDATE notes SET is_system = 1 WHERE id = ?1", [note_id])
.unwrap();
let doc_after: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents WHERE source_type = 'note' AND source_id = ?1",
[note_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(
doc_after, 0,
"notes_au_system_cleanup trigger should delete the document"
);
}
#[test]
fn test_migration_024_system_note_delete_trigger_does_not_fire() {
let conn = setup_migrated_db();
let pid = insert_test_project(&conn);
let issue_id = insert_test_issue(&conn, pid);
let disc_id = insert_test_discussion(&conn, pid, issue_id);
// Insert a system note (is_system = true)
let note_id = insert_test_note(&conn, 202, disc_id, pid, true);
// Manually insert a document (shouldn't exist for system notes in practice,
// but we test the trigger guard)
insert_test_document(&conn, "note", note_id, pid);
let doc_before: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents WHERE source_type = 'note' AND source_id = ?1",
[note_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(doc_before, 1);
// Delete system note -- trigger has WHEN old.is_system = 0 so it should NOT fire
conn.execute("DELETE FROM notes WHERE id = ?1", [note_id])
.unwrap();
let doc_after: i64 = conn
.query_row(
"SELECT COUNT(*) FROM documents WHERE source_type = 'note' AND source_id = ?1",
[note_id],
|row| row.get(0),
)
.unwrap();
assert_eq!(
doc_after, 1,
"notes_ad_cleanup trigger should NOT fire for system notes"
);
}
/// Run migrations only up to version `up_to` (inclusive).
fn run_migrations_up_to(conn: &Connection, up_to: usize) {
conn.execute_batch(
"CREATE TABLE IF NOT EXISTS schema_version ( \
version INTEGER PRIMARY KEY, applied_at INTEGER NOT NULL, description TEXT);",
)
.unwrap();
for (version_str, sql) in &MIGRATIONS[..up_to] {
let version: i32 = version_str.parse().unwrap();
conn.execute_batch(sql).unwrap();
conn.execute(
"INSERT OR REPLACE INTO schema_version (version, applied_at, description) \
VALUES (?1, strftime('%s', 'now') * 1000, ?2)",
rusqlite::params![version, version_str],
)
.unwrap();
}
}
/// Run a single migration by index (0-based).
fn run_single_migration(conn: &Connection, index: usize) {
let (version_str, sql) = MIGRATIONS[index];
let version: i32 = version_str.parse().unwrap();
conn.execute_batch(sql).unwrap();
conn.execute(
"INSERT OR REPLACE INTO schema_version (version, applied_at, description) \
VALUES (?1, strftime('%s', 'now') * 1000, ?2)",
rusqlite::params![version, version_str],
)
.unwrap();
}
#[test]
fn test_migration_025_backfills_existing_notes() {
let conn = create_connection(Path::new(":memory:")).unwrap();
// Run all migrations through 024 (index 0..24)
run_migrations_up_to(&conn, 24);
let pid = insert_test_project(&conn);
let issue_id = insert_test_issue(&conn, pid);
let disc_id = insert_test_discussion(&conn, pid, issue_id);
// Insert 5 non-system notes
for i in 1..=5 {
insert_test_note(&conn, 300 + i, disc_id, pid, false);
}
// Insert 2 system notes
for i in 1..=2 {
insert_test_note(&conn, 400 + i, disc_id, pid, true);
}
// Run migration 025
run_single_migration(&conn, 24);
let dirty_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM dirty_sources WHERE source_type = 'note'",
[],
|row| row.get(0),
)
.unwrap();
assert_eq!(
dirty_count, 5,
"Migration 025 should backfill 5 non-system notes"
);
// Verify system notes were not backfilled
let system_note_ids: Vec<i64> = {
let mut stmt = conn
.prepare(
"SELECT source_id FROM dirty_sources WHERE source_type = 'note' ORDER BY source_id",
)
.unwrap();
stmt.query_map([], |row| row.get(0))
.unwrap()
.collect::<std::result::Result<Vec<_>, _>>()
.unwrap()
};
// System note ids should not appear
let all_system_note_ids: Vec<i64> = {
let mut stmt = conn
.prepare("SELECT id FROM notes WHERE is_system = 1 ORDER BY id")
.unwrap();
stmt.query_map([], |row| row.get(0))
.unwrap()
.collect::<std::result::Result<Vec<_>, _>>()
.unwrap()
};
for sys_id in &all_system_note_ids {
assert!(
!system_note_ids.contains(sys_id),
"System note id {} should not be in dirty_sources",
sys_id
);
}
}
#[test]
fn test_migration_025_idempotent_with_existing_documents() {
let conn = create_connection(Path::new(":memory:")).unwrap();
run_migrations_up_to(&conn, 24);
let pid = insert_test_project(&conn);
let issue_id = insert_test_issue(&conn, pid);
let disc_id = insert_test_discussion(&conn, pid, issue_id);
// Insert 3 non-system notes
let note_ids: Vec<i64> = (1..=3)
.map(|i| insert_test_note(&conn, 500 + i, disc_id, pid, false))
.collect();
// Create documents for 2 of 3 notes (simulating already-generated docs)
insert_test_document(&conn, "note", note_ids[0], pid);
insert_test_document(&conn, "note", note_ids[1], pid);
// Run migration 025
run_single_migration(&conn, 24);
let dirty_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM dirty_sources WHERE source_type = 'note'",
[],
|row| row.get(0),
)
.unwrap();
assert_eq!(
dirty_count, 1,
"Only the note without a document should be backfilled"
);
// Verify the correct note was queued
let queued_id: i64 = conn
.query_row(
"SELECT source_id FROM dirty_sources WHERE source_type = 'note'",
[],
|row| row.get(0),
)
.unwrap();
assert_eq!(queued_id, note_ids[2]);
}
#[test]
fn test_migration_025_skips_notes_already_in_dirty_queue() {
let conn = create_connection(Path::new(":memory:")).unwrap();
run_migrations_up_to(&conn, 24);
let pid = insert_test_project(&conn);
let issue_id = insert_test_issue(&conn, pid);
let disc_id = insert_test_discussion(&conn, pid, issue_id);
// Insert 3 non-system notes
let note_ids: Vec<i64> = (1..=3)
.map(|i| insert_test_note(&conn, 600 + i, disc_id, pid, false))
.collect();
// Pre-queue one note in dirty_sources
conn.execute(
"INSERT INTO dirty_sources (source_type, source_id, queued_at) VALUES ('note', ?1, 999)",
[note_ids[0]],
)
.unwrap();
// Run migration 025
run_single_migration(&conn, 24);
let dirty_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM dirty_sources WHERE source_type = 'note'",
[],
|row| row.get(0),
)
.unwrap();
assert_eq!(
dirty_count, 3,
"All 3 notes should be in dirty_sources (1 pre-existing + 2 new)"
);
// Verify the pre-existing entry preserved its original queued_at
let original_queued_at: i64 = conn
.query_row(
"SELECT queued_at FROM dirty_sources WHERE source_type = 'note' AND source_id = ?1",
[note_ids[0]],
|row| row.get(0),
)
.unwrap();
assert_eq!(
original_queued_at, 999,
"ON CONFLICT DO NOTHING should preserve the original queued_at"
);
}