fix: avoid manifest writes for read-only directory namespace opens (#3635)

Bumps Lance to v9.0.0-beta.19, which includes lance-format/lance#7687
for side-effect-free DirectoryNamespace read paths.

This fixes root-level read-only table opens that previously could
trigger `__manifest` creation through directory namespace construction,
including Hugging Face bucket reads with read-only tokens. A LanceDB
regression test now covers root listing operations without creating
`__manifest`.

Fixes #3633.

python/python/lancedb/__init__.py

@@ -149,8 +149,14 @@ def connect(
 
     For object storage, use a URI prefix:
 
-    >>> db = lancedb.connect("s3://my-bucket/lancedb",
-    ...                      storage_options={"aws_access_key_id": "***"})
+    >>> db = lancedb.connect(  # doctest: +SKIP
+    ...     "s3://my-bucket/lancedb",
+    ...     storage_options={
+    ...         "aws_access_key_id": "***",
+    ...         "aws_secret_access_key": "***",
+    ...         "aws_region": "us-east-1",
+    ...     },
+    ... )
 
     For tests and temporary data, use an in-memory database:
 

python/python/tests/test_s3_bucket_dots.py

@@ -7,7 +7,8 @@ Tests for S3 bucket names containing dots.
 Related issue: https://github.com/lancedb/lancedb/issues/1898
 
 These tests validate the early error checking for S3 bucket names with dots.
-No actual S3 connection is made - validation happens before connection.
+When validation succeeds, eager namespace initialization may still fail later
+because these tests intentionally do not provide real S3 credentials.
 """
 
 import pytest
@@ -20,6 +21,13 @@ BUCKET_WITH_DOTS_AND_AWS_REGION = ("s3://my.bucket.name", {"aws_region": "us-eas
 BUCKET_WITHOUT_DOTS = "s3://my-bucket/path"
 
 
+def assert_not_rejected_for_bucket_dots(connect):
+    try:
+        connect()
+    except ValueError as err:
+        assert "contains dots" not in str(err)
+
+
 class TestS3BucketWithDotsSync:
     """Tests for connect()."""
 
@@ -27,19 +35,22 @@ class TestS3BucketWithDotsSync:
         with pytest.raises(ValueError, match="contains dots"):
             lancedb.connect(BUCKET_WITH_DOTS)
 
-    def test_bucket_with_dots_and_region_passes(self):
+    def test_bucket_with_dots_and_region_is_not_rejected(self):
         uri, opts = BUCKET_WITH_DOTS_AND_REGION
-        db = lancedb.connect(uri, storage_options=opts)
-        assert db is not None
+        assert_not_rejected_for_bucket_dots(
+            lambda: lancedb.connect(uri, storage_options=opts)
+        )
 
-    def test_bucket_with_dots_and_aws_region_passes(self):
+    def test_bucket_with_dots_and_aws_region_is_not_rejected(self):
         uri, opts = BUCKET_WITH_DOTS_AND_AWS_REGION
-        db = lancedb.connect(uri, storage_options=opts)
-        assert db is not None
+        assert_not_rejected_for_bucket_dots(
+            lambda: lancedb.connect(uri, storage_options=opts)
+        )
 
-    def test_bucket_without_dots_passes(self):
-        db = lancedb.connect(BUCKET_WITHOUT_DOTS)
-        assert db is not None
+    def test_bucket_without_dots_is_not_rejected(self):
+        assert_not_rejected_for_bucket_dots(
+            lambda: lancedb.connect(BUCKET_WITHOUT_DOTS)
+        )
 
 
 class TestS3BucketWithDotsAsync:
@@ -51,18 +62,24 @@ class TestS3BucketWithDotsAsync:
             await lancedb.connect_async(BUCKET_WITH_DOTS)
 
     @pytest.mark.asyncio
-    async def test_bucket_with_dots_and_region_passes(self):
+    async def test_bucket_with_dots_and_region_is_not_rejected(self):
         uri, opts = BUCKET_WITH_DOTS_AND_REGION
-        db = await lancedb.connect_async(uri, storage_options=opts)
-        assert db is not None
+        try:
+            await lancedb.connect_async(uri, storage_options=opts)
+        except ValueError as err:
+            assert "contains dots" not in str(err)
 
     @pytest.mark.asyncio
-    async def test_bucket_with_dots_and_aws_region_passes(self):
+    async def test_bucket_with_dots_and_aws_region_is_not_rejected(self):
         uri, opts = BUCKET_WITH_DOTS_AND_AWS_REGION
-        db = await lancedb.connect_async(uri, storage_options=opts)
-        assert db is not None
+        try:
+            await lancedb.connect_async(uri, storage_options=opts)
+        except ValueError as err:
+            assert "contains dots" not in str(err)
 
     @pytest.mark.asyncio
-    async def test_bucket_without_dots_passes(self):
-        db = await lancedb.connect_async(BUCKET_WITHOUT_DOTS)
-        assert db is not None
+    async def test_bucket_without_dots_is_not_rejected(self):
+        try:
+            await lancedb.connect_async(BUCKET_WITHOUT_DOTS)
+        except ValueError as err:
+            assert "contains dots" not in str(err)