Files
lancedb/nodejs/src/connection.rs
T
Jack Ye 21f11b4463 feat!: replace get_job/job_history with describe_job/query_job_events (#4130)
A 1M-row column refresh over 200 fragments produced no visible result,
and the client could only ever say `"running"`. Everything needed to
diagnose it already existed server-side — the job registry records a
`claim`/`claim_complete` pair per fragment carrying `rows_processed` —
but none of it was reachable.

## Before

Four ways to ask about a job, none of which told you much.

```python
job = table.refresh_column_async("embedding")
job.status()               # "running". That was the entire debug surface.
db.get_job(job_id)         # state, and a spec. No result, no progress.
db.job_history(job_id)     # raw record batches, no limit, no filter
db.job(job_id)             # a handle that knew nothing
```

## After

Open a job the way you open a table; the handle answers everything.

```python
job = db.open_job(job_id)      # raises JobNotFoundError if there is no such job
```

```python
>>> print(job)
Job(
    id='job-1',
    state='failed',
    job_type='refresh_column',
    creation_ms=1757000000000,
    spec={
        "column": "embedding",
        "num_workers": 4
    },
    failure=JobFailureInfo(phase='execute', message='worker died', retryable=True),
)
```

Individual fields are there too — `job.state`, `job.job_type`,
`job.creation_ms`, `job.spec`, `job.result`, `job.failure` — and
`job.result` carries `rows_assigned` / `rows_failed` as soon as the job
succeeds, with no `wait()` required.

Per-fragment progress *while it is still running*:

```python
done = job.events(filter="state = 'claim_complete'", limit=10_000)
done.column("rows_processed").to_pylist()      # [5000, 5000, ...]
```

The handle an async action returns is the same object, one `refresh()`
away:

```python
job = table.refresh_column_async("embedding")
job.refresh()
job.state, job.result
```

TypeScript is the same experience, down to `console.log`:

```ts
const job = await db.openJob(jobId);   // rejects if there is no such job
console.log(job);                      // same multi-line layout
job.state; job.jobType; job.spec; job.result; job.failure;
const done = await job.events({ filter: "state = 'claim_complete'", limit: 10_000 });
```

## Why each piece matters

- **A result without waiting.** `rows_assigned` / `rows_failed` used to
live only on the terminal result, so a job that never terminated
reported nothing at all.
- **`limit`.** The server caps event rows at 1000 and truncates without
saying so, which silently hid most of a 200-fragment job's history.
- **`filter`.** `claim_complete` rows carry per-claim `rows_processed` —
the only progress signal that exists mid-flight.
- **Events outlive the worker.** They live in the job registry, not in
pod logs that vanish with the pod.
- **One place to ask.** `open_job` replaces `describe_job`,
`query_job_events` and `job`, so a question about a job has one answer
instead of one per calling location.
- **A missing job is an error, not a `None`.** The common case is a job
id copied out of a log, where absence is the surprise worth raising —
and it matches `open_table`.
- **Printing is the debug surface.** Every field on its own line, JSON
payloads keeping their structure. An unrefreshed handle stays on one
line, because there is nothing to lay out.
- **In-process jobs say so.** A local refresh reports `state` and leaves
the rest null rather than inventing fields it has no record for.

`list_jobs` and `cancel_job` stay as they were: one lists, the other is
a one-shot action that should not need a describe first.

## Breaking

All shipped in 0.38.0. No deprecated aliases.

| Was | Now |
| --- | --- |
| `Connection.get_job` → `describe_job` | `Connection.open_job` returns
a populated `Job`, or raises |
| `Connection.job_history` → `query_job_events` | `job.events(...)` |
| `Connection.job` | `Connection.open_job` |
| Python events → `List[pa.RecordBatch]` | `pa.Table` |
| `JobDescription.spec_json` / `.result_json` | internal; use `job.spec`
/ `job.result` |

Node's `Job` is now a TypeScript class wrapping the native handle, so it
returns an Arrow table and parsed values like Python does. New
`Error::JobNotFound` / `JobNotFoundError`; the three job exceptions are
now in the Python API reference.
2026-09-05 16:24:23 -07:00

616 lines
20 KiB
Rust

// SPDX-License-Identifier: Apache-2.0
// SPDX-FileCopyrightText: Copyright The LanceDB Authors
use std::collections::HashMap;
use std::sync::Arc;
use lancedb::database::{CreateTableMode, Database};
use napi::bindgen_prelude::*;
use napi_derive::*;
use crate::ConnectNamespaceOptions;
use crate::ConnectionOptions;
use crate::error::NapiErrorExt;
use crate::header::JsHeaderProvider;
use crate::table::Table;
use lancedb::connection::{ConnectBuilder, Connection as LanceDBConnection, connect_namespace};
use lance_namespace::models::{
CreateNamespaceRequest, DescribeNamespaceRequest, DropNamespaceRequest, ListNamespacesRequest,
ListTablesRequest,
};
use lancedb::ipc::{ipc_file_to_batches, ipc_file_to_schema};
#[napi]
pub struct Connection {
inner: Option<LanceDBConnection>,
}
#[napi(object)]
pub struct DescribeNamespaceResponse {
pub properties: Option<HashMap<String, String>>,
}
#[napi(object)]
pub struct ListNamespacesResponse {
pub namespaces: Vec<String>,
pub page_token: Option<String>,
}
#[napi(object)]
pub struct ListTablesResponse {
pub tables: Vec<String>,
pub page_token: Option<String>,
}
#[napi(object)]
pub struct CreateNamespaceResponse {
pub properties: Option<HashMap<String, String>>,
pub transaction_id: Option<String>,
}
#[napi(object)]
pub struct DropNamespaceResponse {
pub properties: Option<HashMap<String, String>>,
pub transaction_id: Option<Vec<String>>,
}
impl Connection {
pub(crate) fn inner_new(inner: LanceDBConnection) -> Self {
Self { inner: Some(inner) }
}
fn get_inner(&self) -> napi::Result<&LanceDBConnection> {
self.inner
.as_ref()
.ok_or_else(|| napi::Error::from_reason("Connection is closed"))
}
}
impl Connection {
fn parse_create_mode_str(mode: &str) -> napi::Result<CreateTableMode> {
match mode {
"create" => Ok(CreateTableMode::Create),
"overwrite" => Ok(CreateTableMode::Overwrite),
"exist_ok" => Ok(CreateTableMode::exist_ok(|builder| builder)),
_ => Err(napi::Error::from_reason(format!("Invalid mode {}", mode))),
}
}
pub fn database(&self) -> napi::Result<Arc<dyn Database>> {
Ok(self.get_inner()?.database().clone())
}
}
#[napi]
impl Connection {
/// Create a new Connection instance from the given URI.
#[napi(factory)]
pub async fn new(
uri: String,
options: ConnectionOptions,
header_provider: Option<&JsHeaderProvider>,
) -> napi::Result<Self> {
let mut builder = ConnectBuilder::new(&uri);
if let Some(interval) = options.read_consistency_interval {
builder =
builder.read_consistency_interval(std::time::Duration::from_secs_f64(interval));
}
if let Some(storage_options) = options.storage_options {
for (key, value) in storage_options {
builder = builder.storage_option(key, value);
}
}
if let Some(manifest_enabled) = options.manifest_enabled {
builder = builder.manifest_enabled(manifest_enabled);
}
if let Some(namespace_client_properties) = options.namespace_client_properties {
builder = builder.namespace_client_properties(namespace_client_properties);
}
// Create client config, optionally with header provider
let client_config = options.client_config.unwrap_or_default();
let mut rust_config: lancedb::remote::ClientConfig = client_config.into();
if let Some(provider) = header_provider {
rust_config.header_provider =
Some(Arc::new(provider.clone()) as Arc<dyn lancedb::remote::HeaderProvider>);
}
builder = builder.client_config(rust_config);
if let Some(oauth_config) = options.oauth_config {
let config: lancedb::remote::oauth::OAuthConfig =
oauth_config.try_into().default_error()?;
builder = builder.oauth_config(config);
}
if let Some(api_key) = options.api_key {
builder = builder.api_key(&api_key);
}
if let Some(region) = options.region {
builder = builder.region(&region);
} else {
builder = builder.region("us-east-1");
}
if let Some(host_override) = options.host_override {
builder = builder.host_override(&host_override);
}
if let Some(session) = options.session {
builder = builder.session(session.inner.clone());
}
Ok(Self::inner_new(builder.execute().await.default_error()?))
}
/// Create a new Connection instance backed by a namespace implementation.
#[napi(factory)]
pub async fn new_with_namespace(
impl_name: String,
properties: HashMap<String, String>,
options: ConnectNamespaceOptions,
) -> napi::Result<Self> {
if impl_name.is_empty() {
return Err(napi::Error::from_reason(
"implName must be a non-empty string",
));
}
let mut builder = connect_namespace(&impl_name, properties);
if let Some(interval) = options.read_consistency_interval {
builder =
builder.read_consistency_interval(std::time::Duration::from_secs_f64(interval));
}
if let Some(storage_options) = options.storage_options {
for (key, value) in storage_options {
builder = builder.storage_option(key, value);
}
}
if let Some(namespace_client_properties) = options.namespace_client_properties {
builder = builder.namespace_client_properties(namespace_client_properties);
}
if let Some(session) = options.session {
builder = builder.session(session.inner.clone());
}
Ok(Self::inner_new(builder.execute().await.default_error()?))
}
#[napi]
pub fn display(&self) -> napi::Result<String> {
Ok(self.get_inner()?.to_string())
}
#[napi]
pub fn is_open(&self) -> bool {
self.inner.is_some()
}
#[napi]
pub fn close(&mut self) {
self.inner.take();
}
/// List all tables in the dataset.
#[napi(catch_unwind)]
pub async fn table_names(
&self,
namespace_path: Option<Vec<String>>,
start_after: Option<String>,
limit: Option<u32>,
) -> napi::Result<Vec<String>> {
let mut op = self.get_inner()?.table_names();
op = op.namespace(namespace_path.unwrap_or_default());
if let Some(start_after) = start_after {
op = op.start_after(start_after);
}
if let Some(limit) = limit {
op = op.limit(limit);
}
op.execute().await.default_error()
}
/// List a page of tables in the database.
#[napi(catch_unwind)]
pub async fn list_tables(
&self,
namespace_path: Option<Vec<String>>,
page_token: Option<String>,
limit: Option<u32>,
) -> napi::Result<ListTablesResponse> {
let request = ListTablesRequest {
// The root namespace is an empty path, not an absent one: a namespace-backed
// database rejects a request that names no namespace.
id: Some(namespace_path.unwrap_or_default()),
page_token,
limit: limit.map(|limit| i32::try_from(limit).unwrap_or(i32::MAX)),
..Default::default()
};
let response = self
.get_inner()?
.list_tables(request)
.await
.default_error()?;
Ok(ListTablesResponse {
tables: response.tables,
page_token: response.page_token,
})
}
/// Create table from a Apache Arrow IPC (file) buffer.
///
/// Parameters:
/// - name: The name of the table.
/// - buf: The buffer containing the IPC file.
///
#[napi(catch_unwind)]
pub async fn create_table(
&self,
name: String,
buf: Buffer,
mode: String,
namespace_path: Option<Vec<String>>,
storage_options: Option<HashMap<String, String>>,
) -> napi::Result<Table> {
let batches = ipc_file_to_batches(buf.to_vec())
.map_err(|e| napi::Error::from_reason(format!("Failed to read IPC file: {}", e)))?;
let mode = Self::parse_create_mode_str(&mode)?;
let mut builder = self.get_inner()?.create_table(&name, batches).mode(mode);
builder = builder.namespace(namespace_path.unwrap_or_default());
if let Some(storage_options) = storage_options {
for (key, value) in storage_options {
builder = builder.storage_option(key, value);
}
}
let tbl = builder.execute().await.default_error()?;
Ok(Table::new(tbl))
}
#[napi(catch_unwind)]
pub async fn create_empty_table(
&self,
name: String,
schema_buf: Buffer,
mode: String,
namespace_path: Option<Vec<String>>,
storage_options: Option<HashMap<String, String>>,
) -> napi::Result<Table> {
let schema = ipc_file_to_schema(schema_buf.to_vec()).map_err(|e| {
napi::Error::from_reason(format!("Failed to marshal schema from JS to Rust: {}", e))
})?;
let mode = Self::parse_create_mode_str(&mode)?;
let mut builder = self
.get_inner()?
.create_empty_table(&name, schema)
.mode(mode);
builder = builder.namespace(namespace_path.unwrap_or_default());
if let Some(storage_options) = storage_options {
for (key, value) in storage_options {
builder = builder.storage_option(key, value);
}
}
let tbl = builder.execute().await.default_error()?;
Ok(Table::new(tbl))
}
#[napi(catch_unwind)]
pub async fn create_materialized_view(
&self,
name: String,
source: String,
projections: Option<Vec<Vec<String>>>,
filter: Option<String>,
limit: Option<i64>,
) -> napi::Result<Table> {
let mut builder = self.get_inner()?.create_materialized_view(name, source);
if let Some(projections) = projections {
let mut pairs = Vec::with_capacity(projections.len());
for pair in projections {
let [output, expression]: [String; 2] = pair.try_into().map_err(|_| {
napi::Error::from_reason("each projection must be an [output, expression] pair")
})?;
pairs.push((output, expression));
}
builder = builder.select(pairs);
}
if let Some(filter) = filter {
builder = builder.only_if(filter);
}
if let Some(limit) = limit {
let limit = u64::try_from(limit)
.map_err(|_| napi::Error::from_reason("limit must be a non-negative integer"))?;
builder = builder.limit(limit);
}
let view = builder.execute().await.default_error()?;
Ok(Table::new(view.table().clone()))
}
#[napi(catch_unwind)]
pub async fn open_materialized_view(&self, name: String) -> napi::Result<Table> {
let view = self
.get_inner()?
.open_materialized_view(&name)
.await
.default_error()?;
Ok(Table::new(view.table().clone()))
}
#[napi(catch_unwind)]
pub async fn list_materialized_views(&self) -> napi::Result<Vec<String>> {
let views = self
.get_inner()?
.list_materialized_views()
.await
.default_error()?;
Ok(views.into_iter().map(|v| v.name).collect())
}
#[napi(catch_unwind)]
pub async fn open_table(
&self,
name: String,
namespace_path: Option<Vec<String>>,
storage_options: Option<HashMap<String, String>>,
index_cache_size: Option<u32>,
) -> napi::Result<Table> {
let mut builder = self.get_inner()?.open_table(&name);
builder = builder.namespace(namespace_path.unwrap_or_default());
if let Some(storage_options) = storage_options {
for (key, value) in storage_options {
builder = builder.storage_option(key, value);
}
}
if let Some(index_cache_size) = index_cache_size {
builder = builder.index_cache_size(index_cache_size);
}
let tbl = builder.execute().await.default_error()?;
Ok(Table::new(tbl))
}
#[napi(catch_unwind)]
pub async fn clone_table(
&self,
target_table_name: String,
source_uri: String,
target_namespace_path: Option<Vec<String>>,
source_version: Option<i64>,
source_tag: Option<String>,
is_shallow: bool,
) -> napi::Result<Table> {
let mut builder = self
.get_inner()?
.clone_table(&target_table_name, &source_uri);
builder = builder.target_namespace(target_namespace_path.unwrap_or_default());
if let Some(version) = source_version {
builder = builder.source_version(version as u64);
}
if let Some(tag) = source_tag {
builder = builder.source_tag(tag);
}
builder = builder.is_shallow(is_shallow);
let tbl = builder.execute().await.default_error()?;
Ok(Table::new(tbl))
}
/// Drop table with the name. Or raise an error if the table does not exist.
#[napi(catch_unwind)]
pub async fn drop_table(
&self,
name: String,
namespace_path: Option<Vec<String>>,
) -> napi::Result<()> {
let ns = namespace_path.unwrap_or_default();
self.get_inner()?
.drop_table(&name, &ns)
.await
.default_error()
}
/// Start dropping a table and return its cleanup job.
#[napi(catch_unwind)]
pub async fn drop_table_async(
&self,
name: String,
namespace_path: Option<Vec<String>>,
) -> napi::Result<crate::job::Job> {
let ns = namespace_path.unwrap_or_default();
let job = self
.get_inner()?
.drop_table_async(&name, &ns)
.await
.default_error()?;
Ok(crate::job::Job::new(job))
}
#[napi(catch_unwind)]
pub async fn drop_all_tables(&self, namespace_path: Option<Vec<String>>) -> napi::Result<()> {
let ns = namespace_path.unwrap_or_default();
self.get_inner()?.drop_all_tables(&ns).await.default_error()
}
/// Open a server-side job by id, returning a handle with its record
/// already populated. Rejects when the server has no such job.
///
/// The returned handle answers for its own state, specification, result,
/// failure and event history, so there is no separate connection-level
/// call for any of them.
#[napi(catch_unwind)]
pub async fn open_job(&self, job_id: String) -> napi::Result<crate::job::Job> {
let job = self.get_inner()?.open_job(&job_id).await.default_error()?;
Ok(crate::job::Job::new(job))
}
/// List server-side jobs across the database's tables.
#[napi(catch_unwind)]
pub async fn list_jobs(&self) -> napi::Result<Vec<crate::job::JobInfo>> {
let jobs = self.get_inner()?.list_jobs().await.default_error()?;
Ok(jobs.into_iter().map(Into::into).collect())
}
/// Request cancellation of a server-side job by id. Returns true if the
/// server accepted the cancellation, false if no such job exists.
#[napi(catch_unwind)]
pub async fn cancel_job(&self, job_id: String) -> napi::Result<bool> {
self.get_inner()?.cancel_job(&job_id).await.default_error()
}
#[napi(catch_unwind)]
/// Describe a namespace and return its properties.
pub async fn describe_namespace(
&self,
namespace_path: Vec<String>,
) -> napi::Result<DescribeNamespaceResponse> {
let req = DescribeNamespaceRequest {
id: Some(namespace_path),
..Default::default()
};
let resp = self
.get_inner()?
.describe_namespace(req)
.await
.default_error()?;
Ok(DescribeNamespaceResponse {
properties: resp.properties,
})
}
#[napi(catch_unwind)]
/// List child namespaces under the given namespace path
pub async fn list_namespaces(
&self,
namespace_path: Option<Vec<String>>,
page_token: Option<String>,
limit: Option<u32>,
) -> napi::Result<ListNamespacesResponse> {
let req = ListNamespacesRequest {
id: namespace_path,
page_token,
limit: limit.map(|l| l as i32),
..Default::default()
};
let resp = self
.get_inner()?
.list_namespaces(req)
.await
.default_error()?;
Ok(ListNamespacesResponse {
namespaces: resp.namespaces,
page_token: resp.page_token,
})
}
#[napi(catch_unwind)]
/// Create a new namespace with optional properties.
pub async fn create_namespace(
&self,
namespace_path: Vec<String>,
mode: Option<String>,
properties: Option<HashMap<String, String>>,
) -> napi::Result<CreateNamespaceResponse> {
let mode_str = mode
.map(|m| match m.to_lowercase().as_str() {
"create" => Ok("Create".to_string()),
"exist_ok" => Ok("ExistOk".to_string()),
"overwrite" => Ok("Overwrite".to_string()),
_ => Err(napi::Error::from_reason(format!(
"Invalid mode '{}': expected one of 'create', 'exist_ok', 'overwrite'",
m
))),
})
.transpose()?;
let req = CreateNamespaceRequest {
id: Some(namespace_path),
mode: mode_str,
properties,
..Default::default()
};
let resp = self
.get_inner()?
.create_namespace(req)
.await
.default_error()?;
Ok(CreateNamespaceResponse {
properties: resp.properties,
transaction_id: resp.transaction_id,
})
}
#[napi(catch_unwind)]
/// Drop a namespace.
pub async fn drop_namespace(
&self,
namespace_path: Vec<String>,
mode: Option<String>,
behavior: Option<String>,
) -> napi::Result<DropNamespaceResponse> {
let mode_str = mode
.map(|m| match m.to_lowercase().as_str() {
"skip" => Ok("Skip".to_string()),
"fail" => Ok("Fail".to_string()),
_ => Err(napi::Error::from_reason(format!(
"Invalid mode '{}': expected one of 'skip', 'fail'",
m
))),
})
.transpose()?;
let behavior_str = behavior
.map(|b| match b.to_lowercase().as_str() {
"restrict" => Ok("Restrict".to_string()),
"cascade" => Ok("Cascade".to_string()),
_ => Err(napi::Error::from_reason(format!(
"Invalid behavior '{}': expected one of 'restrict', 'cascade'",
b
))),
})
.transpose()?;
let req = DropNamespaceRequest {
id: Some(namespace_path),
mode: mode_str,
behavior: behavior_str,
..Default::default()
};
let resp = self
.get_inner()?
.drop_namespace(req)
.await
.default_error()?;
Ok(DropNamespaceResponse {
properties: resp.properties,
transaction_id: resp.transaction_id,
})
}
/// Rename a table. `current_namespace_path` and `new_namespace_path` default to
/// the root namespace when omitted; the caller is expected to either pass both
/// or pass neither.
#[napi(catch_unwind)]
pub async fn rename_table(
&self,
current_name: String,
new_name: String,
current_namespace_path: Option<Vec<String>>,
new_namespace_path: Option<Vec<String>>,
) -> napi::Result<()> {
let cur_ns = current_namespace_path.unwrap_or_default();
let new_ns = new_namespace_path.unwrap_or_default();
self.get_inner()?
.rename_table(&current_name, &new_name, &cur_ns, &new_ns)
.await
.default_error()
}
}