fix: address release review (WAL-aware sync hash, empty MCP inputs)

- Sync hashed only the main database file, but local edits can sit in
  the -wal sidecar, so an edited database could match an earlier build
  and be replaced. Checkpoint the WAL before hashing; if changes remain,
  never take the rebuilt-snapshot path.
- MCP tools rejected nothing for empty symbols/sections/kinds and
  returned an empty success; they now fail with INVALIDINPUT, as the CLI
  requires input.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
hermes 2026-10-14 19:44:24 +00:00
commit 58d421028f
4 changed files with 126 additions and 23 deletions

View file

@ -93,7 +93,9 @@ pub struct OwnershipSyncResult {
struct LocalSnapshotState {
/// SHA-256 of the database file. Queries leave a synced file
/// byte-identical, so this identifies an unmodified published snapshot.
file_sha256: String,
/// `None` when committed changes could not be checkpointed out of the
/// WAL, so the file alone does not show the database's contents.
file_sha256: Option<String>,
latest_release: Option<OwnershipRelease>,
release_count: usize,
}
@ -338,11 +340,14 @@ fn inspect_local_db(db_path: &Path) -> Result<Option<LocalSnapshotState>, IdxErr
return Ok(None);
}
let file_sha256 = if checkpoint_wal(db_path) {
// Hash before opening: schema setup may write to an empty database.
let bytes = fs::read(db_path)
.map_err(|e| IdxError::Io(format!("failed to read {}: {e}", db_path.display())))?;
let file_sha256 = sha256_hex(&bytes);
drop(bytes);
Some(sha256_hex(&bytes))
} else {
None
};
let conn = Connection::open(db_path).map_err(|e| {
IdxError::DatabaseError(format!("failed to open {}: {e}", db_path.display()))
@ -357,6 +362,20 @@ fn inspect_local_db(db_path: &Path) -> Result<Option<LocalSnapshotState>, IdxErr
}))
}
/// Moves committed WAL frames into the database file, so its hash covers local
/// edits too. Returns false when changes remain in the WAL.
fn checkpoint_wal(db_path: &Path) -> bool {
let wal_path = sidecar_path(db_path, "-wal");
let wal_is_empty = || fs::metadata(&wal_path).map_or(true, |meta| meta.len() == 0);
if wal_is_empty() {
return true;
}
if let Ok(conn) = Connection::open(db_path) {
let _ = conn.query_row("PRAGMA wal_checkpoint(TRUNCATE)", [], |_| Ok(()));
}
wal_is_empty()
}
fn build_sync_decision(
local: Option<&LocalSnapshotState>,
snapshot: &OwnershipSnapshotArtifact,
@ -414,7 +433,8 @@ fn build_sync_decision(
};
}
if local.file_sha256 == snapshot.sqlite_sha256 {
let file_sha256 = local.file_sha256.as_deref();
if file_sha256 == Some(snapshot.sqlite_sha256.as_str()) {
return SyncDecision {
action: OwnershipSyncAction::NoChange,
should_download: false,
@ -429,7 +449,7 @@ fn build_sync_decision(
if snapshot
.replaces_sqlite_sha256
.iter()
.any(|sha256| sha256.eq_ignore_ascii_case(&local.file_sha256))
.any(|sha256| file_sha256.is_some_and(|local| sha256.eq_ignore_ascii_case(local)))
{
return SyncDecision {
action: OwnershipSyncAction::Updated,
@ -657,8 +677,7 @@ fn build_backup_path(db_path: &Path) -> PathBuf {
}
}
fn remove_sqlite_sidecars(db_path: &Path) -> Result<(), IdxError> {
for suffix in ["-wal", "-shm"] {
fn sidecar_path(db_path: &Path, suffix: &str) -> PathBuf {
let file_name = format!(
"{}{suffix}",
db_path
@ -666,10 +685,15 @@ fn remove_sqlite_sidecars(db_path: &Path) -> Result<(), IdxError> {
.and_then(|name| name.to_str())
.unwrap_or("ownership.db")
);
let sidecar = match db_path.parent() {
match db_path.parent() {
Some(parent) if !parent.as_os_str().is_empty() => parent.join(&file_name),
_ => PathBuf::from(&file_name),
};
}
}
fn remove_sqlite_sidecars(db_path: &Path) -> Result<(), IdxError> {
for suffix in ["-wal", "-shm"] {
let sidecar = sidecar_path(db_path, suffix);
if sidecar.exists() {
fs::remove_file(&sidecar).map_err(|e| {
@ -807,8 +831,7 @@ mod tests {
replaces_sqlite_sha256: Vec::new(),
};
let local = super::LocalSnapshotState {
file_sha256: "dddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd"
.to_string(),
file_sha256: Some("d".repeat(64)),
latest_release: Some(OwnershipRelease {
id: 1,
source_url: None,
@ -849,7 +872,7 @@ mod tests {
file_sha256: &str,
) -> super::LocalSnapshotState {
super::LocalSnapshotState {
file_sha256: file_sha256.to_string(),
file_sha256: Some(file_sha256.to_string()),
latest_release: Some(OwnershipRelease {
id: 3,
source_url: None,
@ -1036,4 +1059,49 @@ mod tests {
.unwrap();
assert_eq!(release_count, 1);
}
#[test]
fn local_hash_covers_committed_changes_still_in_the_wal() {
let dir = std::env::temp_dir().join(format!(
"idx-snapshot-wal-{}-{}",
std::process::id(),
fastrand::u64(..)
));
std::fs::create_dir_all(&dir).unwrap();
let db_path = dir.join("ownership.db");
let release = |sha256: &str| OwnershipRelease {
id: 0,
source_url: None,
sha256: sha256.to_string(),
as_of_date: NaiveDate::from_ymd_opt(2026, 9, 30).unwrap(),
row_count: 1,
imported_at: 1,
};
{
let conn = Connection::open(&db_path).unwrap();
ensure_schema(&conn).unwrap();
insert_release(&conn, &release("published")).unwrap();
}
let published = super::sha256_hex(&std::fs::read(&db_path).unwrap());
// A committed edit that stays in the WAL: the main file still hashes
// as the published build while the connection is open.
let editor = Connection::open(&db_path).unwrap();
ensure_schema(&editor).unwrap();
editor
.query_row("PRAGMA wal_autocheckpoint = 0", [], |_| Ok(()))
.unwrap();
insert_release(&editor, &release("local edit")).unwrap();
assert_eq!(
super::sha256_hex(&std::fs::read(&db_path).unwrap()),
published
);
let local = super::inspect_local_db(&db_path).unwrap().unwrap();
assert_ne!(local.file_sha256.as_deref(), Some(published.as_str()));
assert_eq!(local.release_count, 2);
drop(editor);
let _ = std::fs::remove_dir_all(&dir);
}
}

View file

@ -99,6 +99,9 @@ for one or more IDX tickers."
)]
async fn get_quotes(&self, params: Parameters<SymbolsParams>) -> ToolResult {
let SymbolsParams { symbols } = params.0;
if let Some(err) = empty_input("symbols", &symbols) {
return err;
}
self.live(move |idx| {
let quotes = idx.market().quotes(&symbols, CachePolicy::Default)?;
Ok(json!({ "quotes": quotes }))
@ -155,6 +158,9 @@ earnings, analyst sentiment, insights, and news. Sections that fail are listed u
async fn get_company(&self, params: Parameters<CompanyParams>) -> ToolResult {
let CompanyParams { symbol, sections } = params.0;
let sections = sections.unwrap_or_else(|| CompanySection::ALL.to_vec());
if let Some(err) = empty_input("sections", &sections) {
return err;
}
self.live(move |idx| {
let financial = idx.financial();
let policy = CachePolicy::Default;
@ -195,6 +201,9 @@ are listed under `errors`; the call fails only if every analysis fails."
async fn analyze_stock(&self, params: Parameters<AnalyzeParams>) -> ToolResult {
let AnalyzeParams { symbol, kinds } = params.0;
let kinds = kinds.unwrap_or_else(|| AnalysisKind::ALL.to_vec());
if let Some(err) = empty_input("kinds", &kinds) {
return err;
}
self.live(move |idx| {
let analysis = idx.analysis();
let policy = CachePolicy::Default;
@ -224,6 +233,9 @@ tickers. Tickers that fail are listed under `failures`."
)]
async fn compare_stocks(&self, params: Parameters<SymbolsParams>) -> ToolResult {
let SymbolsParams { symbols } = params.0;
if let Some(err) = empty_input("symbols", &symbols) {
return err;
}
self.live(move |idx| {
let comparison = idx.analysis().compare(&symbols, CachePolicy::Default)?;
let failures: Vec<Value> = comparison
@ -377,6 +389,17 @@ where
})
}
/// An explicitly empty list is an input error, as in the CLI, rather than an
/// empty success an agent could mistake for data.
fn empty_input<T>(field: &str, values: &[T]) -> Option<ToolResult> {
values.is_empty().then(|| {
let err = IdxError::InvalidInput(format!("{field} must not be empty"));
Ok(CallToolResult::structured_error(Value::Object(error_body(
&err,
))))
})
}
fn to_json<T: Serialize>(value: T) -> Result<Value, IdxError> {
serde_json::to_value(value).map_err(|err| IdxError::ParseError(err.to_string()))
}

View file

@ -243,6 +243,17 @@ fn errors_carry_cli_codes() {
assert!(is_error);
assert_eq!(body["code"], "INVALIDINPUT");
for (tool, arguments) in [
("get_quotes", json!({"symbols": []})),
("compare_stocks", json!({"symbols": []})),
("get_company", json!({"symbol": "BBCA", "sections": []})),
("analyze_stock", json!({"symbol": "BBCA", "kinds": []})),
] {
let (is_error, body) = client.call(tool, arguments);
assert!(is_error, "{tool}: {body}");
assert_eq!(body["code"], "INVALIDINPUT", "{tool}");
}
// Empty ownership database: investor tools point at the sync path.
let status = client.ok("ownership_status", json!({}));
assert_eq!(status["populated"], false);

View file

@ -76,7 +76,8 @@ Without `--force`, sync behaves conservatively:
- local latest release older than snapshot: replace with snapshot
- local DB file is byte-identical to `sqlite_sha256`: no-op
- local DB file is byte-identical to a build in `replaces_sqlite_sha256`: replace
(queries never write to a synced file, so this means it was not modified)
(queries never write to a synced file, so this means it was not modified;
committed changes still in the `-wal` file are checkpointed before hashing)
- local latest release/date/count matches snapshot: no-op
- local latest release matches but has fewer releases than snapshot: replace to fill missing history
- local DB newer than snapshot: no-op and keep local data