docs(nodejs): regenerate Table.md for setLsmWriteSpec

JSDoc was updated in the previous commit; regenerate the markdown to match. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
fix: remove primary key constraint from MemWAL bucket sharding
2026-05-23 15:00:39 +00:00 · 2026-05-22 15:00:53 -07:00 · 2026-05-22 14:18:58 -07:00 · 2026-05-22 12:54:31 -07:00 · 2026-05-21 22:55:23 +00:00
11 changed files with 86 additions and 129 deletions
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -3284,8 +3284,8 @@ checksum = "42703706b716c37f96a77aea830392ad231f44c9e9a67872fa5548707e11b11c"

 [[package]]
 name = "fsst"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow-array",
 "rand 0.9.4",
@@ -4506,8 +4506,8 @@ checksum = "e037a2e1d8d5fdbd49b16a4ea09d5d6401c1f29eca5ff29d03d3824dba16256a"

 [[package]]
 name = "lance"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arc-swap",
 "arrow",
@@ -4525,6 +4525,7 @@ dependencies = [
 "async_cell",
 "aws-credential-types",
 "aws-sdk-dynamodb",
+ "bitpacking",
 "byteorder",
 "bytes",
 "chrono",
@@ -4554,6 +4555,7 @@ dependencies = [
 "lance-table",
 "lance-tokenizer",
 "log",
+ "moka",
 "object_store",
 "permutation",
 "pin-project",
@@ -4577,8 +4579,8 @@ dependencies = [

 [[package]]
 name = "lance-arrow"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow-array",
 "arrow-buffer",
@@ -4598,8 +4600,8 @@ dependencies = [

 [[package]]
 name = "lance-bitpacking"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrayref",
 "paste",
@@ -4608,8 +4610,8 @@ dependencies = [

 [[package]]
 name = "lance-core"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow-array",
 "arrow-buffer",
@@ -4644,8 +4646,8 @@ dependencies = [

 [[package]]
 name = "lance-datafusion"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow",
 "arrow-array",
@@ -4675,8 +4677,8 @@ dependencies = [

 [[package]]
 name = "lance-datagen"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow",
 "arrow-array",
@@ -4694,8 +4696,8 @@ dependencies = [

 [[package]]
 name = "lance-encoding"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow-arith",
 "arrow-array",
@@ -4730,8 +4732,8 @@ dependencies = [

 [[package]]
 name = "lance-file"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow-arith",
 "arrow-array",
@@ -4762,8 +4764,8 @@ dependencies = [

 [[package]]
 name = "lance-index"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arc-swap",
 "arrow",
@@ -4827,8 +4829,8 @@ dependencies = [

 [[package]]
 name = "lance-io"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow",
 "arrow-arith",
@@ -4870,8 +4872,8 @@ dependencies = [

 [[package]]
 name = "lance-linalg"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow-array",
 "arrow-buffer",
@@ -4887,8 +4889,8 @@ dependencies = [

 [[package]]
 name = "lance-namespace"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow",
 "async-trait",
@@ -4900,8 +4902,8 @@ dependencies = [

 [[package]]
 name = "lance-namespace-impls"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow",
 "arrow-ipc",
@@ -4936,9 +4938,9 @@ dependencies = [

 [[package]]
 name = "lance-namespace-reqwest-client"
-version = "0.7.6"
+version = "0.7.7"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f65e31bdaa13e01dab6e7cf566da31df243c34a542f0d915d3601ec0e01e61d2"
+checksum = "6369eee4682fb11edf538388b43c61ce288b8302fe89bb40944d7daa7faaae99"
 dependencies = [
 "reqwest 0.12.28",
 "serde",
@@ -4950,8 +4952,8 @@ dependencies = [

 [[package]]
 name = "lance-table"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow",
 "arrow-array",
@@ -4990,8 +4992,8 @@ dependencies = [

 [[package]]
 name = "lance-testing"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "arrow-array",
 "arrow-schema",
@@ -5002,8 +5004,8 @@ dependencies = [

 [[package]]
 name = "lance-tokenizer"
-version = "7.0.0-beta.13"
-source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-beta.13#929166e3ff51ed61b1fa42de2c63feaf51967ea1"
+version = "7.0.0-rc.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v7.0.0-rc.1#06f52e901dd9bd99d299263130fb2bda0a3f91af"
 dependencies = [
 "jieba-rs",
 "lindera",
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -13,20 +13,20 @@ categories = ["database-implementations"]
 rust-version = "1.91.0"

 [workspace.dependencies]
-lance = { "version" = "=7.0.0-beta.13", default-features = false, "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-core = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-datagen = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-file = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-io = { "version" = "=7.0.0-beta.13", default-features = false, "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-index = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-linalg = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-namespace = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-namespace-impls = { "version" = "=7.0.0-beta.13", default-features = false, "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-table = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-testing = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-datafusion = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-encoding = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
-lance-arrow = { "version" = "=7.0.0-beta.13", "tag" = "v7.0.0-beta.13", "git" = "https://github.com/lance-format/lance.git" }
+lance = { "version" = "=7.0.0-rc.1", default-features = false, "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-core = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-datagen = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-file = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-io = { "version" = "=7.0.0-rc.1", default-features = false, "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-index = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-linalg = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-namespace = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-namespace-impls = { "version" = "=7.0.0-rc.1", default-features = false, "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-table = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-testing = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-datafusion = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-encoding = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
+lance-arrow = { "version" = "=7.0.0-rc.1", "tag" = "v7.0.0-rc.1", "git" = "https://github.com/lance-format/lance.git" }
 ahash = "0.8"
 # Note that this one does not include pyarrow
 arrow = { version = "58.0.0", optional = false }
--- a/docs/src/js/classes/Table.md
+++ b/docs/src/js/classes/Table.md
@@ -701,15 +701,11 @@ LSM-style write path for future `mergeInsert` calls.

 `LsmWriteSpec` chooses one of three sharding strategies via `specType`:

- `"bucket"` — hash-bucket writes by the single-column unenforced primary
-  key (`column` and `numBuckets` required).
+- `"bucket"` — hash-bucket writes by a scalar `column` (`column` and
+  `numBuckets` required).
 - `"identity"` — shard by the raw value of a scalar `column`.
 - `"unsharded"` — route every write to a single shard.

-All variants require the table to have an unenforced primary key
-([Table#setUnenforcedPrimaryKey](Table.md#setunenforcedprimarykey)); bucket sharding additionally
-requires it to be the single column being bucketed.
-
 #### Parameters

 * **spec**: [`LsmWriteSpec`](../interfaces/LsmWriteSpec.md)
@@ -722,7 +718,6 @@ requires it to be the single column being bucketed.
 #### Example

 ```ts
-await table.setUnenforcedPrimaryKey("id");
 await table.setLsmWriteSpec({
  specType: "bucket",
  column: "id",
--- a/java/pom.xml
+++ b/java/pom.xml
@@ -28,7 +28,7 @@
    <properties>
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
        <arrow.version>15.0.0</arrow.version>
-        <lance-core.version>7.0.0-beta.13</lance-core.version>
+        <lance-core.version>7.0.0-rc.1</lance-core.version>
        <spotless.skip>false</spotless.skip>
        <spotless.version>2.30.0</spotless.version>
        <spotless.java.googlejavaformat.version>1.7</spotless.java.googlejavaformat.version>
--- a/nodejs/test/connection.test.ts
+++ b/nodejs/test/connection.test.ts
@@ -7,6 +7,13 @@ import * as tmp from "tmp";
 import { Connection, Table, connect, connectNamespace } from "../lancedb";
 import { LocalTable } from "../lancedb/table";

+// The `_versions` directory may also contain `latest_version_hint.json`, which
+// Lance writes on non-lexically-ordered stores (e.g. the local filesystem) to
+// speed up latest-version lookup. Only assert on manifest files.
+function manifestFiles(versionsDir: string): string[] {
+  return readdirSync(versionsDir).filter((f) => f.endsWith(".manifest"));
+}
+
 describe("when connecting", () => {
  let tmpDir: tmp.DirResult;
  beforeEach(() => {
@@ -171,7 +178,7 @@ describe("given a connection", () => {

    let manifestDir =
      tmpDir.name + "/test_manifest_paths_v2_empty.lance/_versions";
-    readdirSync(manifestDir).forEach((file) => {
+    manifestFiles(manifestDir).forEach((file) => {
      expect(file).toMatch(/^\d{20}\.manifest$/);
    });

@@ -180,7 +187,7 @@ describe("given a connection", () => {
    })) as LocalTable;
    expect(await table.usesV2ManifestPaths()).toBe(true);
    manifestDir = tmpDir.name + "/test_manifest_paths_v2.lance/_versions";
-    readdirSync(manifestDir).forEach((file) => {
+    manifestFiles(manifestDir).forEach((file) => {
      expect(file).toMatch(/^\d{20}\.manifest$/);
    });
  });
@@ -199,14 +206,14 @@ describe("given a connection", () => {

    const manifestDir =
      tmpDir.name + "/test_manifest_path_migration.lance/_versions";
-    readdirSync(manifestDir).forEach((file) => {
+    manifestFiles(manifestDir).forEach((file) => {
      expect(file).toMatch(/^\d\.manifest$/);
    });

    await table.migrateManifestPathsV2();
    expect(await table.usesV2ManifestPaths()).toBe(true);

-    readdirSync(manifestDir).forEach((file) => {
+    manifestFiles(manifestDir).forEach((file) => {
      expect(file).toMatch(/^\d{20}\.manifest$/);
    });
  });
--- a/nodejs/lancedb/table.ts
+++ b/nodejs/lancedb/table.ts
@@ -537,19 +537,14 @@ export abstract class Table {
   *
   * `LsmWriteSpec` chooses one of three sharding strategies via `specType`:
   *
-   * - `"bucket"` — hash-bucket writes by the single-column unenforced primary
-   *   key (`column` and `numBuckets` required).
+   * - `"bucket"` — hash-bucket writes by a scalar `column` (`column` and
+   *   `numBuckets` required).
   * - `"identity"` — shard by the raw value of a scalar `column`.
   * - `"unsharded"` — route every write to a single shard.
-   *
-   * All variants require the table to have an unenforced primary key
-   * ({@link Table#setUnenforcedPrimaryKey}); bucket sharding additionally
-   * requires it to be the single column being bucketed.
   * @param {LsmWriteSpec} spec The sharding spec to install.
   * @returns {Promise<void>}
   * @example
   * ```ts
-   * await table.setUnenforcedPrimaryKey("id");
   * await table.setLsmWriteSpec({
   *   specType: "bucket",
   *   column: "id",
--- a/python/python/lancedb/table.py
+++ b/python/python/lancedb/table.py
@@ -3875,15 +3875,11 @@ class AsyncTable:
        strategies:

        - ``LsmWriteSpec.bucket(column, num_buckets)`` — hash-bucket writes by
-          the single-column unenforced primary key.
+          a scalar column.
        - ``LsmWriteSpec.identity(column)`` — shard by the raw value of a
          scalar column.
        - ``LsmWriteSpec.unsharded()`` — route every write to a single shard.

-        All variants require the table to have an unenforced primary key set
-        via [`set_unenforced_primary_key`]; bucket sharding additionally
-        requires it to be the single column being bucketed.
-
        Parameters
        ----------
        spec : LsmWriteSpec
@@ -3892,7 +3888,6 @@ class AsyncTable:
        Examples
        --------
        >>> from lancedb._lancedb import LsmWriteSpec
-        >>> # table.set_unenforced_primary_key("id")
        >>> # table.set_lsm_write_spec(LsmWriteSpec.bucket("id", 16))
        """
        await self._inner.set_lsm_write_spec(spec)
--- a/python/python/tests/test_db.py
+++ b/python/python/tests/test_db.py
@@ -450,6 +450,13 @@ async def test_create_exist_ok_async(tmp_db_async: lancedb.AsyncConnection):
    #     await db.create_table("test", schema=bad_schema, exist_ok=True)


+def _manifest_files(versions_dir):
+    # The `_versions` directory may also contain `latest_version_hint.json`,
+    # which Lance writes on non-lexically-ordered stores (e.g. the local
+    # filesystem) to speed up latest-version lookup. Only assert on manifests.
+    return [f for f in os.listdir(versions_dir) if f.endswith(".manifest")]
+
+
@pytest.mark.asyncio
 async def test_create_table_v2_manifest_paths_async(tmp_path):
    db_with_v2_paths = await lancedb.connect_async(
@@ -465,7 +472,7 @@ async def test_create_table_v2_manifest_paths_async(tmp_path):
    )
    assert await tbl.uses_v2_manifest_paths()
    manifests_dir = tmp_path / "test_v2_manifest_paths.lance" / "_versions"
-    for manifest in os.listdir(manifests_dir):
+    for manifest in _manifest_files(manifests_dir):
        assert re.match(r"\d{20}\.manifest", manifest)

    # Start a table in V1 mode then migrate
@@ -475,13 +482,13 @@ async def test_create_table_v2_manifest_paths_async(tmp_path):
    )
    assert not await tbl.uses_v2_manifest_paths()
    manifests_dir = tmp_path / "test_v2_migration.lance" / "_versions"
-    for manifest in os.listdir(manifests_dir):
+    for manifest in _manifest_files(manifests_dir):
        assert re.match(r"\d\.manifest", manifest)

    await tbl.migrate_manifest_paths_v2()
    assert await tbl.uses_v2_manifest_paths()

-    for manifest in os.listdir(manifests_dir):
+    for manifest in _manifest_files(manifests_dir):
        assert re.match(r"\d{20}\.manifest", manifest)


--- a/python/python/tests/test_lsm_write_spec.py
+++ b/python/python/tests/test_lsm_write_spec.py
@@ -40,16 +40,6 @@ def _make_table(tmp_path):
 def test_set_lsm_write_spec_validates(tmp_path):
    _db, table = _make_table(tmp_path)

-    # No PK set yet.
-    with pytest.raises(Exception, match="primary key"):
-        table.set_lsm_write_spec(LsmWriteSpec.bucket("id", 4))
-
-    table.set_unenforced_primary_key("id")
-
-    # Column mismatch.
-    with pytest.raises(Exception, match="match"):
-        table.set_lsm_write_spec(LsmWriteSpec.bucket("v", 4))
-
    # Out-of-range num_buckets.
    with pytest.raises(Exception, match="num_buckets"):
        table.set_lsm_write_spec(LsmWriteSpec.bucket("id", 0))
@@ -70,7 +60,6 @@ def test_unset_lsm_write_spec(tmp_path):
        table.unset_lsm_write_spec()

    # Install a spec, then remove it; afterwards a fresh spec can be set.
-    table.set_unenforced_primary_key("id")
    table.set_lsm_write_spec(LsmWriteSpec.bucket("id", 4))
    table.unset_lsm_write_spec()
    # A second unset errors — there is no spec left to remove.
@@ -81,9 +70,7 @@ def test_unset_lsm_write_spec(tmp_path):

 def test_set_unsharded_spec(tmp_path):
    _db, table = _make_table(tmp_path)
-    # Lance MemWAL still requires a primary key on the dataset; Unsharded
-    # just skips per-row hashing.
-    table.set_unenforced_primary_key("id")
+    # Unsharded routes every write to a single shard, skipping per-row hashing.
    table.set_lsm_write_spec(LsmWriteSpec.unsharded())
    table.unset_lsm_write_spec()

@@ -120,7 +107,6 @@ async def test_async_set_unset_lsm_write_spec(tmp_path):
        pa.RecordBatchReader.from_batches(SCHEMA, [_batch(["seed"], [0])]),
    )

-    await table.set_unenforced_primary_key("id")
    await table.set_lsm_write_spec(LsmWriteSpec.bucket("id", 4))
    await table.unset_lsm_write_spec()
    # A second unset errors.
@@ -130,9 +116,7 @@ async def test_async_set_unset_lsm_write_spec(tmp_path):

 def test_set_identity_spec(tmp_path):
    _db, table = _make_table(tmp_path)
-    # Identity sharding still requires an unenforced primary key on the
-    # table; it shards by the raw value of the given column.
-    table.set_unenforced_primary_key("id")
+    # Identity sharding shards by the raw value of the given column.
    table.set_lsm_write_spec(LsmWriteSpec.identity("v"))
    table.unset_lsm_write_spec()

--- a/python/src/table.rs
+++ b/python/src/table.rs
@@ -185,7 +185,7 @@ pub struct LsmWriteSpec {

 #[pymethods]
 impl LsmWriteSpec {
-    /// Hash-bucket sharding by the unenforced primary key column.
+    /// Hash-bucket sharding by a scalar column.
    #[staticmethod]
    pub fn bucket(column: String, num_buckets: u32) -> Self {
        Self {
--- a/rust/lancedb/src/table.rs
+++ b/rust/lancedb/src/table.rs
@@ -282,17 +282,15 @@ pub use self::merge::MergeResult;
 /// date) and [`LsmWriteSpec::with_writer_config_defaults`] (default
 /// `ShardWriter` configuration recorded in the MemWAL index).
 ///
-/// All variants require the table to have an unenforced primary key.
-///
 /// Install a spec with [`Table::set_lsm_write_spec`] and remove it with
 /// [`Table::unset_lsm_write_spec`]. The actual `merge_insert` dispatch
 /// onto the MemWAL writer is a follow-up.
 #[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
 pub enum LsmWriteSpec {
-    /// Hash-bucket sharding by the unenforced primary key column.
+    /// Hash-bucket sharding by a scalar column.
    ///
-    /// `column` must equal the table's currently-set single-column
-    /// unenforced primary key. `num_buckets` must be in `[1, 1024]`.
+    /// `column` must be a non-nested column with a supported scalar type.
+    /// `num_buckets` must be in `[1, 1024]`.
    /// Iceberg-compatible Murmur3-x86-32 (seed 0) is used so each row's
    /// `bucket(column, num_buckets)` value is stable across processes.
    Bucket {
@@ -1298,21 +1296,15 @@ impl Table {
    ///
    /// [`LsmWriteSpec`] chooses one of three sharding strategies:
    ///
-    /// - [`LsmWriteSpec::bucket`] — hash-bucket writes by the single-column
-    ///   unenforced primary key.
+    /// - [`LsmWriteSpec::bucket`] — hash-bucket writes by a scalar column.
    /// - [`LsmWriteSpec::identity`] — shard by the raw value of a scalar column.
    /// - [`LsmWriteSpec::unsharded`] — route every write to a single shard.
    ///
-    /// All variants require the table to have an unenforced primary key
-    /// ([`Table::set_unenforced_primary_key`]); bucket sharding additionally
-    /// requires it to be the single column being bucketed.
-    ///
    /// # Example
    ///
    /// ```
    /// # use lancedb::table::{LsmWriteSpec, Table};
    /// # async fn example(table: &Table) -> Result<(), Box<dyn std::error::Error>> {
-    /// table.set_unenforced_primary_key(["id"]).await?;
    /// table
    ///     .set_lsm_write_spec(
    ///         LsmWriteSpec::bucket("id", 16).with_maintained_indexes(["id_idx"]),
@@ -4600,21 +4592,6 @@ mod tests {
            .unwrap();
        let table = conn.create_table("t", reader).execute().await.unwrap();

-        // Reject when no PK is set.
-        let err = table
-            .set_lsm_write_spec(LsmWriteSpec::bucket("id", 4))
-            .await
-            .expect_err("should reject without PK");
-        assert!(matches!(err, Error::Lance { .. }), "got {:?}", err);
-
-        // Set PK, then a mismatched column on the spec must be rejected.
-        table.set_unenforced_primary_key(["id"]).await.unwrap();
-        let err = table
-            .set_lsm_write_spec(LsmWriteSpec::bucket("name", 4))
-            .await
-            .expect_err("should reject column != PK");
-        assert!(matches!(err, Error::Lance { .. }), "got {:?}", err);
-
        // Reject num_buckets out of range.
        for bad in [0u32, 1025] {
            let err = table
@@ -4680,9 +4657,6 @@ mod tests {
            .unwrap();
        let table = conn.create_table("t", reader).execute().await.unwrap();

-        // Lance's MemWAL still requires *some* unenforced primary key on
-        // the dataset; Unsharded just skips the per-row hashing step.
-        table.set_unenforced_primary_key(["id"]).await.unwrap();
        table
            .set_lsm_write_spec(LsmWriteSpec::unsharded())
            .await
@@ -4729,7 +4703,6 @@ mod tests {
            .unwrap();
        let table = conn.create_table("t", reader).execute().await.unwrap();

-        table.set_unenforced_primary_key(["id"]).await.unwrap();
        table
            .set_lsm_write_spec(
                LsmWriteSpec::identity("region")
@@ -4785,7 +4758,6 @@ mod tests {
        table.unset_lsm_write_spec().await.unwrap_err();

        // Install a spec, then unset it.
-        table.set_unenforced_primary_key(["id"]).await.unwrap();
        table
            .set_lsm_write_spec(LsmWriteSpec::bucket("id", 4))
            .await
Author	SHA1	Message	Date
Will Jones	e244d753e5	docs(nodejs): regenerate Table.md for setLsmWriteSpec JSDoc was updated in the previous commit; regenerate the markdown to match. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-22 15:00:53 -07:00
Will Jones	a69dd0f62b	fix: remove primary key constraint from MemWAL bucket sharding Lance v7.0.0-rc.1 intentionally removed the requirement for bucket_sharding to match (or even require) the unenforced primary key column. Update LanceDB to match: drop the PK-related doc comments and the test assertions that expected rejection when no PK is set or when the bucket column differs from the PK. The Rust changes are taken from #3435; this commit additionally applies the equivalent updates to the Python and TypeScript bindings. See https://github.com/lance-format/lance/issues/6917 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-22 14:18:58 -07:00
Will Jones	02b112e931	test: ignore latest_version_hint.json in manifest dir assertions Lance v7.0.0-rc.1 writes _versions/latest_version_hint.json on non-lexically-ordered stores (e.g. the local filesystem) to speed up latest-version lookup. The V2-manifest-path tests iterate the whole _versions directory and asserted every entry matches the manifest filename pattern; filter to *.manifest files so the hint file is ignored. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-22 12:54:31 -07:00
lancedb automation	05504f280a	chore: update lance dependency to v7.0.0-rc.1	2026-05-21 22:55:23 +00:00