refactor(python): use namespace-backed rust connection

## Summary
- pass `namespace_client` through the Python create-table path
- ensure schema-only namespace table creation uses the namespace-aware
empty-table flow
- fix reopening namespace tables created without initial data

.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.27.2"
+current_version = "0.28.0-beta.7"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

.github/ISSUE_TEMPLATE/documentation.yml

@@ -18,6 +18,6 @@ body:
       label: Link
       description: >
         Provide a link to the existing documentation, if applicable.
-      placeholder: ex. https://lancedb.com/docs/tables/...
+      placeholder: ex. https://docs.lancedb.com/tables/...
     validations:
       required: false

.github/workflows/nodejs.yml

@@ -8,6 +8,7 @@ on:
     paths:
       - Cargo.toml
       - Cargo.lock
+      - rust-toolchain.toml
       - nodejs/**
       - rust/**
       - docs/src/js/**

.github/workflows/python.yml

@@ -8,6 +8,7 @@ on:
     paths:
       - Cargo.toml
       - Cargo.lock
+      - rust-toolchain.toml
       - python/**
       - rust/**
       - .github/workflows/python.yml

.github/workflows/rust.yml

@@ -8,6 +8,7 @@ on:
     paths:
       - Cargo.toml
       - Cargo.lock
+      - rust-toolchain.toml
       - rust/**
       - .github/workflows/rust.yml
 

Cargo.lock

@@ -137,15 +137,6 @@ dependencies = [
  "object",
 ]
 
-[[package]]
-name = "arc-swap"
-version = "1.9.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a07d1f37ff60921c83bdfc7407723bdefe89b44b98a9b772f225c8f9d67141a6"
-dependencies = [
- "rustversion",
-]
-
 [[package]]
 name = "argminmax"
 version = "0.6.3"
@@ -302,7 +293,7 @@ dependencies = [
  "arrow-schema",
  "arrow-select",
  "flatbuffers",
- "lz4_flex 0.12.1",
+ "lz4_flex",
  "zstd",
 ]
 
@@ -1278,31 +1269,6 @@ dependencies = [
  "generic-array",
 ]
 
-[[package]]
-name = "bon"
-version = "3.9.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f47dbe92550676ee653353c310dfb9cf6ba17ee70396e1f7cf0a2020ad49b2fe"
-dependencies = [
- "bon-macros",
- "rustversion",
-]
-
-[[package]]
-name = "bon-macros"
-version = "3.9.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "519bd3116aeeb42d5372c29d982d16d0170d3d4a5ed85fc7dd91642ffff3c67c"
-dependencies = [
- "darling 0.20.11",
- "ident_case",
- "prettyplease",
- "proc-macro2",
- "quote",
- "rustversion",
- "syn 2.0.117",
-]
-
 [[package]]
 name = "brotli"
 version = "3.5.0"
@@ -1472,12 +1438,6 @@ dependencies = [
  "shlex",
 ]
 
-[[package]]
-name = "census"
-version = "0.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4f4c707c6a209cbe82d10abd08e1ea8995e9ea937d2550646e02798948992be0"
-
 [[package]]
 name = "cfg-if"
 version = "1.0.4"
@@ -1795,7 +1755,7 @@ dependencies = [
  "crossterm_winapi",
  "document-features",
  "parking_lot",
- "rustix 1.1.4",
+ "rustix",
  "winapi",
 ]
 
@@ -2714,12 +2674,6 @@ dependencies = [
  "litrs",
 ]
 
-[[package]]
-name = "downcast-rs"
-version = "2.0.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "117240f60069e65410b3ae1bb213295bd828f707b5bec6596a1afc8793ce0cbc"
-
 [[package]]
 name = "dtor"
 version = "0.1.1"
@@ -2955,12 +2909,6 @@ version = "0.2.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f8eb564c5c7423d25c886fb561d1e4ee69f72354d16918afa32c08811f6b6a55"
 
-[[package]]
-name = "fastdivide"
-version = "0.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9afc2bd4d5a73106dd53d10d73d3401c2f32730ba2c0b93ddb888a8983680471"
-
 [[package]]
 name = "fastrand"
 version = "2.3.0"
@@ -3054,16 +3002,6 @@ dependencies = [
  "percent-encoding",
 ]
 
-[[package]]
-name = "fs4"
-version = "0.8.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f7e180ac76c23b45e767bd7ae9579bc0bb458618c4bc71835926e098e61d15f8"
-dependencies = [
- "rustix 0.38.44",
- "windows-sys 0.52.0",
-]
-
 [[package]]
 name = "fs_extra"
 version = "1.3.0"
@@ -3072,8 +3010,8 @@ checksum = "42703706b716c37f96a77aea830392ad231f44c9e9a67872fa5548707e11b11c"
 
 [[package]]
 name = "fsst"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow-array",
  "rand 0.9.2",
@@ -3560,12 +3498,6 @@ dependencies = [
  "windows-sys 0.61.2",
 ]
 
-[[package]]
-name = "htmlescape"
-version = "0.3.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e9025058dae765dee5070ec375f591e2ba14638c63feff74f13805a72e523163"
-
 [[package]]
 name = "http"
 version = "0.2.12"
@@ -4134,13 +4066,14 @@ dependencies = [
 
 [[package]]
 name = "lance"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "arrow-arith",
  "arrow-array",
  "arrow-buffer",
+ "arrow-cast",
  "arrow-ipc",
  "arrow-ord",
  "arrow-row",
@@ -4177,12 +4110,14 @@ dependencies = [
  "lance-linalg",
  "lance-namespace",
  "lance-table",
+ "lance-tokenizer",
  "log",
  "moka",
  "object_store",
  "permutation",
  "pin-project",
  "prost",
+ "prost-build",
  "prost-types",
  "rand 0.9.2",
  "roaring",
@@ -4190,7 +4125,6 @@ dependencies = [
  "serde",
  "serde_json",
  "snafu 0.9.0",
- "tantivy",
  "tokio",
  "tokio-stream",
  "tokio-util",
@@ -4201,13 +4135,14 @@ dependencies = [
 
 [[package]]
 name = "lance-arrow"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
  "arrow-cast",
  "arrow-data",
+ "arrow-ipc",
  "arrow-ord",
  "arrow-schema",
  "arrow-select",
@@ -4222,8 +4157,8 @@ dependencies = [
 
 [[package]]
 name = "lance-bitpacking"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrayref",
  "paste",
@@ -4232,8 +4167,8 @@ dependencies = [
 
 [[package]]
 name = "lance-core"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4270,12 +4205,13 @@ dependencies = [
 
 [[package]]
 name = "lance-datafusion"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "arrow-array",
  "arrow-buffer",
+ "arrow-cast",
  "arrow-ord",
  "arrow-schema",
  "arrow-select",
@@ -4301,8 +4237,8 @@ dependencies = [
 
 [[package]]
 name = "lance-datagen"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4320,8 +4256,8 @@ dependencies = [
 
 [[package]]
 name = "lance-encoding"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -4358,8 +4294,8 @@ dependencies = [
 
 [[package]]
 name = "lance-file"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -4391,8 +4327,8 @@ dependencies = [
 
 [[package]]
 name = "lance-index"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "arrow-arith",
@@ -4429,6 +4365,7 @@ dependencies = [
  "lance-io",
  "lance-linalg",
  "lance-table",
+ "lance-tokenizer",
  "libm",
  "log",
  "ndarray",
@@ -4446,7 +4383,6 @@ dependencies = [
  "serde_json",
  "smallvec",
  "snafu 0.9.0",
- "tantivy",
  "tempfile",
  "tokio",
  "tracing",
@@ -4456,8 +4392,8 @@ dependencies = [
 
 [[package]]
 name = "lance-io"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "arrow-arith",
@@ -4501,8 +4437,8 @@ dependencies = [
 
 [[package]]
 name = "lance-linalg"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4518,8 +4454,8 @@ dependencies = [
 
 [[package]]
 name = "lance-namespace"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "async-trait",
@@ -4532,8 +4468,8 @@ dependencies = [
 
 [[package]]
 name = "lance-namespace-impls"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "arrow-ipc",
@@ -4578,8 +4514,8 @@ dependencies = [
 
 [[package]]
 name = "lance-table"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4618,8 +4554,8 @@ dependencies = [
 
 [[package]]
 name = "lance-testing"
-version = "5.0.0-beta.4"
+version = "6.0.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v5.0.0-beta.4#d9068e76a301df9e21d7282419f24f61a11375ac"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
 dependencies = [
  "arrow-array",
  "arrow-schema",
@@ -4628,9 +4564,19 @@ dependencies = [
  "rand 0.9.2",
 ]
 
+[[package]]
+name = "lance-tokenizer"
+version = "6.0.0-beta.1"
+source = "git+https://github.com/lance-format/lance.git?tag=v6.0.0-beta.1#c7a7d3a0e944646e793d297d4a2e2cf7e4fb28a3"
+dependencies = [
+ "rust-stemmers",
+ "serde",
+ "unicode-normalization",
+]
+
 [[package]]
 name = "lancedb"
-version = "0.27.2"
+version = "0.28.0-beta.7"
 dependencies = [
  "ahash",
  "anyhow",
@@ -4712,7 +4658,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb-nodejs"
-version = "0.27.2"
+version = "0.28.0-beta.7"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4734,7 +4680,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb-python"
-version = "0.30.2"
+version = "0.31.0-beta.7"
 dependencies = [
  "arrow",
  "async-trait",
@@ -4772,12 +4718,6 @@ version = "0.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "09edd9e8b54e49e587e4f6295a7d29c3ea94d469cb40ab8ca70b288248a81db2"
 
-[[package]]
-name = "levenshtein_automata"
-version = "0.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0c2cdeb66e45e9f36bfad5bbdb4d2384e70936afbee843c6f6543f0c551ebb25"
-
 [[package]]
 name = "lexical-core"
 version = "1.0.6"
@@ -4866,12 +4806,6 @@ dependencies = [
  "libc",
 ]
 
-[[package]]
-name = "linux-raw-sys"
-version = "0.4.15"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d26c52dbd32dccf2d10cac7725f8eae5296885fb5703b261f7d0a0739ec807ab"
-
 [[package]]
 name = "linux-raw-sys"
 version = "0.12.1"
@@ -4952,12 +4886,6 @@ dependencies = [
  "libc",
 ]
 
-[[package]]
-name = "lz4_flex"
-version = "0.11.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "373f5eceeeab7925e0c1098212f2fbc4d416adec9d35051a6ab251e824c1854a"
-
 [[package]]
 name = "lz4_flex"
 version = "0.12.1"
@@ -5032,15 +4960,6 @@ dependencies = [
  "digest",
 ]
 
-[[package]]
-name = "measure_time"
-version = "0.9.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "51c55d61e72fc3ab704396c5fa16f4c184db37978ae4e94ca8959693a235fc0e"
-dependencies = [
- "log",
-]
-
 [[package]]
 name = "memchr"
 version = "2.8.0"
@@ -5194,12 +5113,6 @@ dependencies = [
  "target-features",
 ]
 
-[[package]]
-name = "murmurhash32"
-version = "0.3.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2195bf6aa996a481483b29d62a7663eed3fe39600c460e323f8ff41e90bdd89b"
-
 [[package]]
 name = "napi"
 version = "3.8.3"
@@ -5488,12 +5401,6 @@ version = "1.70.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "384b8ab6d37215f3c5301a95a4accb5d64aa607f1fcb26a11b5303878451b4fe"
 
-[[package]]
-name = "oneshot"
-version = "0.1.13"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "269bca4c2591a28585d6bf10d9ed0332b7d76900a1b02bec41bdc3a2cdcda107"
-
 [[package]]
 name = "onig"
 version = "6.5.1"
@@ -5583,15 +5490,6 @@ version = "0.5.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1a80800c0488c3a21695ea981a54918fbb37abf04f4d0720c453632255e2ff0e"
 
-[[package]]
-name = "ownedbytes"
-version = "0.9.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2fbd56f7631767e61784dc43f8580f403f4475bd4aaa4da003e6295e1bab4a7e"
-dependencies = [
- "stable_deref_trait",
-]
-
 [[package]]
 name = "p256"
 version = "0.11.1"
@@ -7046,19 +6944,6 @@ dependencies = [
  "semver",
 ]
 
-[[package]]
-name = "rustix"
-version = "0.38.44"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fdb5bc1ae2baa591800df16c9ca78619bf65c0488b41b96ccec5d11220d8c154"
-dependencies = [
- "bitflags",
- "errno",
- "libc",
- "linux-raw-sys 0.4.15",
- "windows-sys 0.59.0",
-]
-
 [[package]]
 name = "rustix"
 version = "1.1.4"
@@ -7068,7 +6953,7 @@ dependencies = [
  "bitflags",
  "errno",
  "libc",
- "linux-raw-sys 0.12.1",
+ "linux-raw-sys",
  "windows-sys 0.59.0",
 ]
 
@@ -7533,15 +7418,6 @@ version = "1.0.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b2aa850e253778c88a04c3d7323b043aeda9d3e30d5971937c1855769763678e"
 
-[[package]]
-name = "sketches-ddsketch"
-version = "0.3.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0c6f73aeb92d671e0cc4dca167e59b2deb6387c375391bc99ee743f326994a2b"
-dependencies = [
- "serde",
-]
-
 [[package]]
 name = "slab"
 version = "0.4.12"
@@ -7914,152 +7790,6 @@ version = "0.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "7b2093cf4c8eb1e67749a6762251bc9cd836b6fc171623bd0a9d324d37af2417"
 
-[[package]]
-name = "tantivy"
-version = "0.24.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "64a966cb0e76e311f09cf18507c9af192f15d34886ee43d7ba7c7e3803660c43"
-dependencies = [
- "aho-corasick",
- "arc-swap",
- "base64 0.22.1",
- "bitpacking",
- "bon",
- "byteorder",
- "census",
- "crc32fast",
- "crossbeam-channel",
- "downcast-rs",
- "fastdivide",
- "fnv",
- "fs4",
- "htmlescape",
- "hyperloglogplus",
- "itertools 0.14.0",
- "levenshtein_automata",
- "log",
- "lru",
- "lz4_flex 0.11.6",
- "measure_time",
- "memmap2 0.9.10",
- "once_cell",
- "oneshot",
- "rayon",
- "regex",
- "rust-stemmers",
- "rustc-hash",
- "serde",
- "serde_json",
- "sketches-ddsketch",
- "smallvec",
- "tantivy-bitpacker",
- "tantivy-columnar",
- "tantivy-common",
- "tantivy-fst",
- "tantivy-query-grammar",
- "tantivy-stacker",
- "tantivy-tokenizer-api",
- "tempfile",
- "thiserror 2.0.18",
- "time",
- "uuid",
- "winapi",
-]
-
-[[package]]
-name = "tantivy-bitpacker"
-version = "0.8.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1adc286a39e089ae9938935cd488d7d34f14502544a36607effd2239ff0e2494"
-dependencies = [
- "bitpacking",
-]
-
-[[package]]
-name = "tantivy-columnar"
-version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6300428e0c104c4f7db6f95b466a6f5c1b9aece094ec57cdd365337908dc7344"
-dependencies = [
- "downcast-rs",
- "fastdivide",
- "itertools 0.14.0",
- "serde",
- "tantivy-bitpacker",
- "tantivy-common",
- "tantivy-sstable",
- "tantivy-stacker",
-]
-
-[[package]]
-name = "tantivy-common"
-version = "0.9.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e91b6ea6090ce03dc72c27d0619e77185d26cc3b20775966c346c6d4f7e99d7f"
-dependencies = [
- "async-trait",
- "byteorder",
- "ownedbytes",
- "serde",
- "time",
-]
-
-[[package]]
-name = "tantivy-fst"
-version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d60769b80ad7953d8a7b2c70cdfe722bbcdcac6bccc8ac934c40c034d866fc18"
-dependencies = [
- "byteorder",
- "regex-syntax",
- "utf8-ranges",
-]
-
-[[package]]
-name = "tantivy-query-grammar"
-version = "0.24.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e810cdeeebca57fc3f7bfec5f85fdbea9031b2ac9b990eb5ff49b371d52bbe6a"
-dependencies = [
- "nom 7.1.3",
- "serde",
- "serde_json",
-]
-
-[[package]]
-name = "tantivy-sstable"
-version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "709f22c08a4c90e1b36711c1c6cad5ae21b20b093e535b69b18783dd2cb99416"
-dependencies = [
- "futures-util",
- "itertools 0.14.0",
- "tantivy-bitpacker",
- "tantivy-common",
- "tantivy-fst",
- "zstd",
-]
-
-[[package]]
-name = "tantivy-stacker"
-version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2bcdebb267671311d1e8891fd9d1301803fdb8ad21ba22e0a30d0cab49ba59c1"
-dependencies = [
- "murmurhash32",
- "rand_distr 0.4.3",
- "tantivy-common",
-]
-
-[[package]]
-name = "tantivy-tokenizer-api"
-version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "dfa942fcee81e213e09715bbce8734ae2180070b97b33839a795ba1de201547d"
-dependencies = [
- "serde",
-]
-
 [[package]]
 name = "tap"
 version = "1.0.1"
@@ -8087,7 +7817,7 @@ dependencies = [
  "fastrand",
  "getrandom 0.4.2",
  "once_cell",
- "rustix 1.1.4",
+ "rustix",
  "windows-sys 0.59.0",
 ]
 
@@ -8539,6 +8269,15 @@ version = "1.0.24"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
 
+[[package]]
+name = "unicode-normalization"
+version = "0.1.25"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5fd4f6878c9cb28d874b009da9e8d183b5abc80117c40bbd187a1fde336be6e8"
+dependencies = [
+ "tinyvec",
+]
+
 [[package]]
 name = "unicode-normalization-alignments"
 version = "0.1.12"

Cargo.toml

@@ -15,20 +15,20 @@ categories = ["database-implementations"]
 rust-version = "1.91.0"
 
 [workspace.dependencies]
-lance = { "version" = "=5.0.0-beta.4", default-features = false, "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance = { "version" = "=6.0.0-beta.1", default-features = false, "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-core = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-core = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-datagen = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-datagen = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-file = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-file = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-io = { "version" = "=5.0.0-beta.4", default-features = false, "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-io = { "version" = "=6.0.0-beta.1", default-features = false, "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-index = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-index = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-linalg = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-linalg = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-namespace = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-namespace = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-namespace-impls = { "version" = "=5.0.0-beta.4", default-features = false, "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-namespace-impls = { "version" = "=6.0.0-beta.1", default-features = false, "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-table = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-table = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-testing = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-testing = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-datafusion = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-datafusion = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-encoding = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-encoding = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-arrow = { "version" = "=5.0.0-beta.4", "tag" = "v5.0.0-beta.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-arrow = { "version" = "=6.0.0-beta.1", "tag" = "v6.0.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
 ahash = "0.8"
 # Note that this one does not include pyarrow
 arrow = { version = "57.2", optional = false }

README.md

@@ -15,7 +15,7 @@
 
 # **The Multimodal AI Lakehouse**
 
-[**How to Install** ](#how-to-install) ✦ [**Detailed Documentation**](https://lancedb.com/docs) ✦ [**Tutorials and Recipes**](https://github.com/lancedb/vectordb-recipes/tree/main) ✦  [**Contributors**](#contributors) 
+[**How to Install** ](#how-to-install) ✦ [**Detailed Documentation**](https://docs.lancedb.com) ✦ [**Tutorials and Recipes**](https://github.com/lancedb/vectordb-recipes/tree/main) ✦  [**Contributors**](#contributors) 
 
 **The ultimate multimodal data platform for AI/ML applications.** 
 
@@ -57,7 +57,7 @@ LanceDB is a central location where developers can build, train and analyze thei
 
 ## **How to Install**:
 
-Follow the [Quickstart](https://lancedb.com/docs/quickstart/) doc to set up LanceDB locally. 
+Follow the [Quickstart](https://docs.lancedb.com/quickstart) doc to set up LanceDB locally. 
 
 **API & SDK:** We also support Python, Typescript and Rust SDKs
 

docs/README.md

@@ -1,6 +1,6 @@
 # LanceDB Documentation
 
-LanceDB docs are available at [lancedb.com/docs](https://lancedb.com/docs).
+LanceDB docs are available at [docs.lancedb.com](https://docs.lancedb.com).
 
 The SDK docs are built and deployed automatically by [Github Actions](../.github/workflows/docs.yml)
 whenever a commit is pushed to the `main` branch. So it is possible for the docs to show

docs/src/java/java.md

@@ -14,7 +14,7 @@ Add the following dependency to your `pom.xml`:
 <dependency>
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-core</artifactId>
-    <version>0.27.2</version>
+    <version>0.28.0-beta.7</version>
 </dependency>
 ```
 

docs/src/js/README.md

@@ -34,7 +34,7 @@ const results = await table.vectorSearch([0.1, 0.3]).limit(20).toArray();
 console.log(results);
 ```
 
-The [quickstart](https://lancedb.com/docs/quickstart/basic-usage/) contains more complete examples.
+The [quickstart](https://docs.lancedb.com/quickstart/) contains more complete examples.
 
 ## Development
 

docs/src/js/interfaces/ClientConfig.md

@@ -53,3 +53,18 @@ optional tlsConfig: TlsConfig;
 ```ts
 optional userAgent: string;
 ```
+
+***
+
+### userId?
+
+```ts
+optional userId: string;
+```
+
+User identifier for tracking purposes.
+
+This is sent as the `x-lancedb-user-id` header in requests to LanceDB Cloud/Enterprise.
+It can be set directly, or via the `LANCEDB_USER_ID` environment variable.
+Alternatively, set `LANCEDB_USER_ID_ENV_KEY` to specify another environment
+variable that contains the user ID value.

docs/src/js/interfaces/ConnectionOptions.md

@@ -89,4 +89,4 @@ optional storageOptions: Record<string, string>;
 
 (For LanceDB OSS only): configuration for object storage.
 
-The available options are described at https://lancedb.com/docs/storage/
+The available options are described at https://docs.lancedb.com/storage/

docs/src/js/interfaces/CreateTableOptions.md

@@ -97,4 +97,4 @@ Configuration for object storage.
 Options already set on the connection will be inherited by the table,
 but can be overridden here.
 
-The available options are described at https://lancedb.com/docs/storage/
+The available options are described at https://docs.lancedb.com/storage/

docs/src/js/interfaces/OpenTableOptions.md

@@ -42,4 +42,4 @@ Configuration for object storage.
 Options already set on the connection will be inherited by the table,
 but can be overridden here.
 
-The available options are described at https://lancedb.com/docs/storage/
+The available options are described at https://docs.lancedb.com/storage/

java/lancedb-core/pom.xml

@@ -8,7 +8,7 @@
     <parent>
       <groupId>com.lancedb</groupId>
       <artifactId>lancedb-parent</artifactId>
-      <version>0.27.2-final.0</version>
+      <version>0.28.0-beta.7</version>
       <relativePath>../pom.xml</relativePath>
     </parent>
 

java/pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-parent</artifactId>
-    <version>0.27.2-final.0</version>
+    <version>0.28.0-beta.7</version>
     <packaging>pom</packaging>
     <name>${project.artifactId}</name>
     <description>LanceDB Java SDK Parent POM</description>
@@ -28,7 +28,7 @@
     <properties>
         <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
         <arrow.version>15.0.0</arrow.version>
-        <lance-core.version>5.0.0-beta.4</lance-core.version>
+        <lance-core.version>6.0.0-beta.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>

nodejs/Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "lancedb-nodejs"
 edition.workspace = true
-version = "0.27.2"
+version = "0.28.0-beta.7"
 license.workspace = true
 description.workspace = true
 repository.workspace = true

nodejs/README.md

@@ -30,7 +30,7 @@ const results = await table.vectorSearch([0.1, 0.3]).limit(20).toArray();
 console.log(results);
 ```
 
-The [quickstart](https://lancedb.com/docs/quickstart/basic-usage/) contains more complete examples.
+The [quickstart](https://docs.lancedb.com/quickstart/) contains more complete examples.
 
 ## Development
 

nodejs/lancedb/connection.ts

@@ -42,7 +42,7 @@ export interface CreateTableOptions {
    * Options already set on the connection will be inherited by the table,
    * but can be overridden here.
    *
-   * The available options are described at https://lancedb.com/docs/storage/
+   * The available options are described at https://docs.lancedb.com/storage/
    */
   storageOptions?: Record<string, string>;
 
@@ -78,7 +78,7 @@ export interface OpenTableOptions {
    * Options already set on the connection will be inherited by the table,
    * but can be overridden here.
    *
-   * The available options are described at https://lancedb.com/docs/storage/
+   * The available options are described at https://docs.lancedb.com/storage/
    */
   storageOptions?: Record<string, string>;
   /**

nodejs/npm/darwin-arm64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-arm64",
-	"version": "0.27.2",
+	"version": "0.28.0-beta.7",
 	"os": ["darwin"],
 	"cpu": ["arm64"],
 	"main": "lancedb.darwin-arm64.node",

nodejs/npm/linux-arm64-gnu/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-gnu",
-	"version": "0.27.2",
+	"version": "0.28.0-beta.7",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-gnu.node",

nodejs/npm/linux-arm64-musl/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-musl",
-	"version": "0.27.2",
+	"version": "0.28.0-beta.7",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-musl.node",

nodejs/npm/linux-x64-gnu/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-gnu",
-	"version": "0.27.2",
+	"version": "0.28.0-beta.7",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-gnu.node",

nodejs/npm/linux-x64-musl/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-musl",
-	"version": "0.27.2",
+	"version": "0.28.0-beta.7",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-musl.node",

nodejs/npm/win32-arm64-msvc/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lancedb/lancedb-win32-arm64-msvc",
-  "version": "0.27.2",
+  "version": "0.28.0-beta.7",
   "os": [
     "win32"
   ],

nodejs/npm/win32-x64-msvc/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-win32-x64-msvc",
-	"version": "0.27.2",
+	"version": "0.28.0-beta.7",
 	"os": ["win32"],
 	"cpu": ["x64"],
 	"main": "lancedb.win32-x64-msvc.node",

nodejs/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "@lancedb/lancedb",
-  "version": "0.27.2",
+  "version": "0.28.0-beta.7",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "@lancedb/lancedb",
-      "version": "0.27.2",
+      "version": "0.28.0-beta.7",
       "cpu": [
         "x64",
         "arm64"

nodejs/package.json

@@ -11,7 +11,7 @@
     "ann"
   ],
   "private": false,
-  "version": "0.27.2",
+  "version": "0.28.0-beta.7",
   "main": "dist/index.js",
   "exports": {
     ".": "./dist/index.js",

nodejs/src/lib.rs

@@ -35,7 +35,7 @@ pub struct ConnectionOptions {
     pub read_consistency_interval: Option<f64>,
     /// (For LanceDB OSS only): configuration for object storage.
     ///
-    /// The available options are described at https://lancedb.com/docs/storage/
+    /// The available options are described at https://docs.lancedb.com/storage/
     pub storage_options: Option<HashMap<String, String>>,
     /// (For LanceDB OSS only): the session to use for this connection. Holds
     /// shared caches and other session-specific state.

nodejs/src/remote.rs

@@ -92,6 +92,13 @@ pub struct ClientConfig {
     pub extra_headers: Option<HashMap<String, String>>,
     pub id_delimiter: Option<String>,
     pub tls_config: Option<TlsConfig>,
+    /// User identifier for tracking purposes.
+    ///
+    /// This is sent as the `x-lancedb-user-id` header in requests to LanceDB Cloud/Enterprise.
+    /// It can be set directly, or via the `LANCEDB_USER_ID` environment variable.
+    /// Alternatively, set `LANCEDB_USER_ID_ENV_KEY` to specify another environment
+    /// variable that contains the user ID value.
+    pub user_id: Option<String>,
 }
 
 impl From<TimeoutConfig> for lancedb::remote::TimeoutConfig {
@@ -145,6 +152,7 @@ impl From<ClientConfig> for lancedb::remote::ClientConfig {
             id_delimiter: config.id_delimiter,
             tls_config: config.tls_config.map(Into::into),
             header_provider: None, // the header provider is set separately later
+            user_id: config.user_id,
         }
     }
 }

python/.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.31.0-beta.0"
+current_version = "0.31.0-beta.7"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

python/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb-python"
-version = "0.31.0-beta.0"
+version = "0.31.0-beta.7"
 edition.workspace = true
 description = "Python bindings for LanceDB"
 license.workspace = true

python/pyproject.toml

@@ -45,7 +45,7 @@ repository = "https://github.com/lancedb/lancedb"
 
 [project.optional-dependencies]
 pylance = [
-    "pylance>=5.0.0b3",
+    "pylance>=5.0.0b5",
 ]
 tests = [
     "aiohttp>=3.9.0",
@@ -59,7 +59,7 @@ tests = [
     "polars>=0.19, <=1.3.0",
     "tantivy>=0.20.0",
     "pyarrow-stubs>=16.0",
-    "pylance>=5.0.0b3",
+    "pylance>=5.0.0b5",
     "requests>=2.31.0",
     "datafusion>=52,<53",
 ]
@@ -83,7 +83,7 @@ embeddings = [
     "colpali-engine>=0.3.10",
     "huggingface_hub>=0.19.0",
     "InstructorEmbedding>=1.0.1",
-    "google.generativeai>=0.3.0",
+    "google-genai>=1.0.0",
     "boto3>=1.28.57",
     "awscli>=1.44.38",
     "botocore>=1.31.57",

python/python/lancedb/__init__.py

@@ -110,7 +110,7 @@ def connect(
         default configuration is used.
     storage_options: dict, optional
         Additional options for the storage backend. See available options at
-        <https://lancedb.com/docs/storage/>
+        <https://docs.lancedb.com/storage/>
     session: Session, optional
         (For LanceDB OSS only)
         A session to use for this connection. Sessions allow you to configure
@@ -215,6 +215,85 @@ def connect(
     )
 
 
+WORKER_PROPERTY_PREFIX = "_lancedb_worker_"
+
+
+def _apply_worker_overrides(props: dict[str, str]) -> dict[str, str]:
+    """Apply worker property overrides.
+
+    Any key starting with ``_lancedb_worker_`` is extracted, the prefix
+    is stripped, and the resulting key-value pair is put back into the
+    map (overriding the existing value if present).  The original
+    prefixed key is removed.
+    """
+    worker_keys = [k for k in props if k.startswith(WORKER_PROPERTY_PREFIX)]
+    if not worker_keys:
+        return props
+    result = dict(props)
+    for key in worker_keys:
+        value = result.pop(key)
+        real_key = key[len(WORKER_PROPERTY_PREFIX) :]
+        result[real_key] = value
+    return result
+
+
+def deserialize_conn(
+    data: str,
+    *,
+    for_worker: bool = False,
+) -> DBConnection:
+    """Reconstruct a DBConnection from a serialized string.
+
+    The string must have been produced by
+    :meth:`DBConnection.serialize`.
+
+    Parameters
+    ----------
+    data : str
+        String produced by ``serialize()``.
+    for_worker : bool, default False
+        When ``True``, any namespace client property whose key starts
+        with ``_lancedb_worker_`` has that prefix stripped and the
+        value overrides the corresponding property.  For example,
+        ``_lancedb_worker_uri`` replaces ``uri``.
+
+    Returns
+    -------
+    DBConnection
+        A new connection matching the serialized state.
+    """
+    import json
+
+    parsed = json.loads(data)
+    connection_type = parsed.get("connection_type")
+
+    rci_secs = parsed.get("read_consistency_interval_seconds")
+    rci = timedelta(seconds=rci_secs) if rci_secs is not None else None
+    storage_options = parsed.get("storage_options")
+
+    if connection_type == "namespace":
+        props = dict(parsed.get("namespace_client_properties") or {})
+        if for_worker:
+            props = _apply_worker_overrides(props)
+        return connect_namespace(
+            namespace_client_impl=parsed["namespace_client_impl"],
+            namespace_client_properties=props,
+            read_consistency_interval=rci,
+            storage_options=storage_options,
+            namespace_client_pushdown_operations=parsed.get(
+                "namespace_client_pushdown_operations"
+            ),
+        )
+    elif connection_type == "local":
+        return LanceDBConnection(
+            parsed["uri"],
+            read_consistency_interval=rci,
+            storage_options=storage_options,
+        )
+    else:
+        raise ValueError(f"Unknown connection_type: {connection_type}")
+
+
 async def connect_async(
     uri: URI,
     *,
@@ -257,7 +336,7 @@ async def connect_async(
         default configuration is used.
     storage_options: dict, optional
         Additional options for the storage backend. See available options at
-        <https://lancedb.com/docs/storage/>
+        <https://docs.lancedb.com/storage/>
     session: Session, optional
         (For LanceDB OSS only)
         A session to use for this connection. Sessions allow you to configure

python/python/lancedb/_lancedb.pyi

@@ -151,6 +151,9 @@ class Connection(object):
     async def drop_all_tables(
         self, namespace_path: Optional[List[str]] = None
     ) -> None: ...
+    async def namespace_client_config(
+        self,
+    ) -> Dict[str, Any]: ...
 
 class Table:
     def name(self) -> str: ...

python/python/lancedb/common.py

@@ -96,7 +96,7 @@ def data_to_reader(
             f"Unknown data type {type(data)}. "
             "Supported types: list of dicts, pandas DataFrame, polars DataFrame, "
             "pyarrow Table/RecordBatch, or Pydantic models. "
-            "See https://lancedb.com/docs/tables/ for examples."
+            "See https://docs.lancedb.com/tables/ for examples."
         )
 
 

python/python/lancedb/db.py

@@ -23,11 +23,13 @@ from lancedb.embeddings.registry import EmbeddingFunctionRegistry
 from lancedb.common import data_to_reader, sanitize_uri, validate_schema
 from lancedb.background_loop import LOOP
 from lance_namespace import (
+    LanceNamespace,
     ListNamespacesResponse,
     CreateNamespaceResponse,
     DropNamespaceResponse,
     DescribeNamespaceResponse,
     ListTablesResponse,
+    connect as namespace_connect,
 )
 
 from . import __version__
@@ -280,7 +282,7 @@ class DBConnection(EnforceOverrides):
             Additional options for the storage backend. Options already set on the
             connection will be inherited by the table, but can be overridden here.
             See available options at
-            <https://lancedb.com/docs/storage/>
+            <https://docs.lancedb.com/storage/>
 
             To enable stable row IDs (row IDs remain stable after compaction,
             update, delete, and merges), set `new_table_enable_stable_row_ids`
@@ -431,7 +433,7 @@ class DBConnection(EnforceOverrides):
             Additional options for the storage backend. Options already set on the
             connection will be inherited by the table, but can be overridden here.
             See available options at
-            <https://lancedb.com/docs/storage/>
+            <https://docs.lancedb.com/storage/>
 
         Returns
         -------
@@ -507,6 +509,39 @@ class DBConnection(EnforceOverrides):
     def uri(self) -> str:
         return self._uri
 
+    def namespace_client(self) -> LanceNamespace:
+        """Get the equivalent namespace client for this connection.
+
+        For native storage connections, this returns a DirectoryNamespace
+        pointing to the same root with the same storage options.
+
+        For namespace connections, this returns the backing namespace client.
+
+        For enterprise (remote) connections, this returns a RestNamespace
+        with the same URI and authentication headers.
+
+        Returns
+        -------
+        LanceNamespace
+            The namespace client for this connection.
+        """
+        raise NotImplementedError(
+            "namespace_client is not supported for this connection type"
+        )
+
+    def serialize(self) -> str:
+        """Serialize this connection for reconstruction.
+
+        The returned string can be passed to :func:`lancedb.deserialize_conn`
+        to recreate an equivalent connection, e.g. in a remote worker.
+
+        Returns
+        -------
+        str
+            Serialized representation of this connection.
+        """
+        raise NotImplementedError("serialize is not supported for this connection type")
+
 
 class LanceDBConnection(DBConnection):
     """
@@ -559,6 +594,7 @@ class LanceDBConnection(DBConnection):
     ):
         if _inner is not None:
             self._conn = _inner
+            self._cached_namespace_client = None
             return
 
         if not isinstance(uri, Path):
@@ -606,6 +642,7 @@ class LanceDBConnection(DBConnection):
         # beyond _conn.
         self.storage_options = storage_options
         self._conn = AsyncConnection(LOOP.run(do_connect()))
+        self._cached_namespace_client: Optional[LanceNamespace] = None
 
     @property
     def read_consistency_interval(self) -> Optional[timedelta]:
@@ -630,6 +667,22 @@ class LanceDBConnection(DBConnection):
         val += ")"
         return val
 
+    @override
+    def serialize(self) -> str:
+        import json
+
+        rci = self.read_consistency_interval
+        return json.dumps(
+            {
+                "connection_type": "local",
+                "uri": self.uri,
+                "storage_options": self.storage_options,
+                "read_consistency_interval_seconds": (
+                    rci.total_seconds() if rci else None
+                ),
+            }
+        )
+
     async def _async_get_table_names(self, start_after: Optional[str], limit: int):
         conn = AsyncConnection(await lancedb_connect(self.uri))
         return await conn.table_names(start_after=start_after, limit=limit)
@@ -665,10 +718,10 @@ class LanceDBConnection(DBConnection):
         """
         if namespace_path is None:
             namespace_path = []
-        return LOOP.run(
+        return self._namespace_conn().list_namespaces(
-            self._conn.list_namespaces(
+            namespace_path=namespace_path,
-                namespace_path=namespace_path, page_token=page_token, limit=limit
+            page_token=page_token,
-            )
+            limit=limit,
         )
 
     @override
@@ -678,27 +731,10 @@ class LanceDBConnection(DBConnection):
         mode: Optional[str] = None,
         properties: Optional[Dict[str, str]] = None,
     ) -> CreateNamespaceResponse:
-        """Create a new namespace.
+        return self._namespace_conn().create_namespace(
-
+            namespace_path=namespace_path,
-        Parameters
+            mode=mode,
-        ----------
+            properties=properties,
-        namespace_path: List[str]
-            The namespace identifier to create.
-        mode: str, optional
-            Creation mode - "create" (fail if exists), "exist_ok" (skip if exists),
-            or "overwrite" (replace if exists). Case insensitive.
-        properties: Dict[str, str], optional
-            Properties to set on the namespace.
-
-        Returns
-        -------
-        CreateNamespaceResponse
-            Response containing the properties of the created namespace.
-        """
-        return LOOP.run(
-            self._conn.create_namespace(
-                namespace_path=namespace_path, mode=mode, properties=properties
-            )
         )
 
     @override
@@ -708,46 +744,19 @@ class LanceDBConnection(DBConnection):
         mode: Optional[str] = None,
         behavior: Optional[str] = None,
     ) -> DropNamespaceResponse:
-        """Drop a namespace.
+        return self._namespace_conn().drop_namespace(
-
+            namespace_path=namespace_path,
-        Parameters
+            mode=mode,
-        ----------
+            behavior=behavior,
-        namespace_path: List[str]
-            The namespace identifier to drop.
-        mode: str, optional
-            Whether to skip if not exists ("SKIP") or fail ("FAIL"). Case insensitive.
-        behavior: str, optional
-            Whether to restrict drop if not empty ("RESTRICT") or cascade ("CASCADE").
-            Case insensitive.
-
-        Returns
-        -------
-        DropNamespaceResponse
-            Response containing properties and transaction_id if applicable.
-        """
-        return LOOP.run(
-            self._conn.drop_namespace(
-                namespace_path=namespace_path, mode=mode, behavior=behavior
-            )
         )
 
     @override
     def describe_namespace(
         self, namespace_path: List[str]
     ) -> DescribeNamespaceResponse:
-        """Describe a namespace.
+        return self._namespace_conn().describe_namespace(
-
+            namespace_path=namespace_path,
-        Parameters
+        )
-        ----------
-        namespace_path: List[str]
-            The namespace identifier to describe.
-
-        Returns
-        -------
-        DescribeNamespaceResponse
-            Response containing the namespace properties.
-        """
-        return LOOP.run(self._conn.describe_namespace(namespace_path=namespace_path))
 
     @override
     def list_tables(
@@ -776,6 +785,12 @@ class LanceDBConnection(DBConnection):
         """
         if namespace_path is None:
             namespace_path = []
+        if namespace_path:
+            return self._namespace_conn().list_tables(
+                namespace_path=namespace_path,
+                page_token=page_token,
+                limit=limit,
+            )
         return LOOP.run(
             self._conn.list_tables(
                 namespace_path=namespace_path, page_token=page_token, limit=limit
@@ -864,6 +879,22 @@ class LanceDBConnection(DBConnection):
             raise ValueError("mode must be either 'create' or 'overwrite'")
         validate_table_name(name)
 
+        if namespace_path:
+            return self._namespace_conn().create_table(
+                name,
+                data=data,
+                schema=schema,
+                mode=mode,
+                exist_ok=exist_ok,
+                on_bad_vectors=on_bad_vectors,
+                fill_value=fill_value,
+                embedding_functions=embedding_functions,
+                namespace_path=namespace_path,
+                storage_options=storage_options,
+                data_storage_version=data_storage_version,
+                enable_v2_manifest_paths=enable_v2_manifest_paths,
+            )
+
         tbl = LanceTable.create(
             self,
             name,
@@ -879,6 +910,19 @@ class LanceDBConnection(DBConnection):
         )
         return tbl
 
+    def _namespace_conn(self) -> DBConnection:
+        """Return a LanceNamespaceDBConnection backed by this connection's
+        directory namespace.  Used to delegate child-namespace operations."""
+        from lancedb.namespace import LanceNamespaceDBConnection
+
+        return LanceNamespaceDBConnection(
+            self.namespace_client(),
+            read_consistency_interval=self.read_consistency_interval,
+            storage_options=self.storage_options,
+            namespace_client_impl=None,
+            namespace_client_properties=None,
+        )
+
     @override
     def open_table(
         self,
@@ -895,7 +939,8 @@ class LanceDBConnection(DBConnection):
         name: str
             The name of the table.
         namespace_path: List[str], optional
-            The namespace to open the table from.
+            The namespace to open the table from.  When non-empty, the
+            table is resolved through the directory namespace client.
 
         Returns
         -------
@@ -914,6 +959,14 @@ class LanceDBConnection(DBConnection):
                 stacklevel=2,
             )
 
+        if namespace_path:
+            return self._namespace_conn().open_table(
+                name,
+                namespace_path=namespace_path,
+                storage_options=storage_options,
+                index_cache_size=index_cache_size,
+            )
+
         return LanceTable.open(
             self,
             name,
@@ -998,6 +1051,9 @@ class LanceDBConnection(DBConnection):
         """
         if namespace_path is None:
             namespace_path = []
+        if namespace_path:
+            self._namespace_conn().drop_table(name, namespace_path=namespace_path)
+            return
         LOOP.run(
             self._conn.drop_table(
                 name, namespace_path=namespace_path, ignore_missing=ignore_missing
@@ -1044,6 +1100,23 @@ class LanceDBConnection(DBConnection):
             )
         )
 
+    @override
+    def namespace_client(self) -> LanceNamespace:
+        """Get the equivalent namespace client for this connection.
+
+        Returns a DirectoryNamespace pointing to the same root with the
+        same storage options.  The result is cached for the lifetime of
+        this connection.
+
+        Returns
+        -------
+        LanceNamespace
+            The namespace client for this connection.
+        """
+        if self._cached_namespace_client is None:
+            self._cached_namespace_client = LOOP.run(self._conn.namespace_client())
+        return self._cached_namespace_client
+
     @deprecation.deprecated(
         deprecated_in="0.15.1",
         removed_in="0.17",
@@ -1317,6 +1390,7 @@ class AsyncConnection(object):
         namespace_path: Optional[List[str]] = None,
         embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
         location: Optional[str] = None,
+        namespace_client: Optional[Any] = None,
     ) -> AsyncTable:
         """Create an [AsyncTable][lancedb.table.AsyncTable] in the database.
 
@@ -1361,7 +1435,7 @@ class AsyncConnection(object):
             Additional options for the storage backend. Options already set on the
             connection will be inherited by the table, but can be overridden here.
             See available options at
-            <https://lancedb.com/docs/storage/>
+            <https://docs.lancedb.com/storage/>
 
             To enable stable row IDs (row IDs remain stable after compaction,
             update, delete, and merges), set `new_table_enable_stable_row_ids`
@@ -1514,6 +1588,7 @@ class AsyncConnection(object):
                 namespace_path=namespace_path,
                 storage_options=storage_options,
                 location=location,
+                namespace_client=namespace_client,
             )
         else:
             data = data_to_reader(data, schema)
@@ -1524,6 +1599,7 @@ class AsyncConnection(object):
                 namespace_path=namespace_path,
                 storage_options=storage_options,
                 location=location,
+                namespace_client=namespace_client,
             )
 
         return AsyncTable(new_table)
@@ -1552,7 +1628,7 @@ class AsyncConnection(object):
             Additional options for the storage backend. Options already set on the
             connection will be inherited by the table, but can be overridden here.
             See available options at
-            <https://lancedb.com/docs/storage/>
+            <https://docs.lancedb.com/storage/>
         index_cache_size: int, default 256
             **Deprecated**: Use session-level cache configuration instead.
             Create a Session with custom cache sizes and pass it to lancedb.connect().
@@ -1716,6 +1792,25 @@ class AsyncConnection(object):
             namespace_path = []
         await self._inner.drop_all_tables(namespace_path=namespace_path)
 
+    async def namespace_client(self) -> LanceNamespace:
+        """Get the equivalent namespace client for this connection.
+
+        For native storage connections, this returns a DirectoryNamespace
+        pointing to the same root with the same storage options.
+
+        For namespace connections, this returns the backing namespace client.
+
+        For enterprise (remote) connections, this returns a RestNamespace
+        with the same URI and authentication headers.
+
+        Returns
+        -------
+        LanceNamespace
+            The namespace client for this connection.
+        """
+        config = await self._inner.namespace_client_config()
+        return namespace_connect(config["impl"], config["properties"])
+
     @deprecation.deprecated(
         deprecated_in="0.15.1",
         removed_in="0.17",

python/python/lancedb/embeddings/gemini_text.py

@@ -19,10 +19,10 @@ from .utils import TEXT, api_key_not_found_help
 @register("gemini-text")
 class GeminiText(TextEmbeddingFunction):
     """
-    An embedding function that uses the Google's Gemini API. Requires GOOGLE_API_KEY to
+    An embedding function that uses Google's Gemini API. Requires GOOGLE_API_KEY to
     be set.
 
-    https://ai.google.dev/docs/embeddings_guide
+    https://ai.google.dev/gemini-api/docs/embeddings
 
     Supports various tasks types:
     | Task Type               | Description                                            |
@@ -46,9 +46,12 @@ class GeminiText(TextEmbeddingFunction):
 
     Parameters
     ----------
-    name: str, default "models/embedding-001"
+    name: str, default "gemini-embedding-001"
-        The name of the model to use. See the Gemini documentation for a list of
+        The name of the model to use. Supported models include:
-        available models.
+        - "gemini-embedding-001" (768 dimensions)
+
+        Note: The legacy "models/embedding-001" format is also supported but
+        "gemini-embedding-001" is recommended.
 
     query_task_type: str, default "retrieval_query"
         Sets the task type for the queries.
@@ -77,7 +80,7 @@ class GeminiText(TextEmbeddingFunction):
 
     """
 
-    name: str = "models/embedding-001"
+    name: str = "gemini-embedding-001"
     query_task_type: str = "retrieval_query"
     source_task_type: str = "retrieval_document"
 
@@ -114,23 +117,48 @@ class GeminiText(TextEmbeddingFunction):
         texts: list[str] or np.ndarray (of str)
             The texts to embed
         """
-        if (
+        from google.genai import types
-            kwargs.get("task_type") == "retrieval_document"
-        ):  # Provide a title to use existing API design
-            title = "Embedding of a document"
-            kwargs["title"] = title
 
-        return [
+        task_type = kwargs.get("task_type")
-            self.client.embed_content(model=self.name, content=text, **kwargs)[
+
-                "embedding"
+        # Build content objects for embed_content
-            ]
+        contents = []
-            for text in texts
+        for text in texts:
-        ]
+            if task_type == "retrieval_document":
+                # Provide a title for retrieval_document task
+                contents.append(
+                    {"parts": [{"text": "Embedding of a document"}, {"text": text}]}
+                )
+            else:
+                contents.append({"parts": [{"text": text}]})
+
+        # Build config
+        config_kwargs = {}
+        if task_type:
+            config_kwargs["task_type"] = task_type.upper()  # API expects uppercase
+
+        # Call embed_content for each content
+        embeddings = []
+        for content in contents:
+            config = (
+                types.EmbedContentConfig(**config_kwargs) if config_kwargs else None
+            )
+            response = self.client.models.embed_content(
+                model=self.name,
+                contents=content,
+                config=config,
+            )
+            embeddings.append(response.embeddings[0].values)
+
+        return embeddings
 
     @cached_property
     def client(self):
-        genai = attempt_import_or_raise("google.generativeai", "google.generativeai")
+        attempt_import_or_raise("google.genai", "google-genai")
 
         if not os.environ.get("GOOGLE_API_KEY"):
             api_key_not_found_help("google")
-        return genai
+
+        from google import genai as genai_module
+
+        return genai_module.Client(api_key=os.environ.get("GOOGLE_API_KEY"))

python/python/lancedb/namespace.py

@@ -10,7 +10,6 @@ through a namespace abstraction.
 
 from __future__ import annotations
 
-import asyncio
 import sys
 from typing import TYPE_CHECKING, Any, Dict, Iterable, List, Optional, Union
 
@@ -25,7 +24,24 @@ if TYPE_CHECKING:
 from datetime import timedelta
 import pyarrow as pa
 
-from lancedb.db import DBConnection, LanceDBConnection
+from lance_namespace_urllib3_client.models.json_arrow_data_type import JsonArrowDataType
+from lance_namespace_urllib3_client.models.json_arrow_field import JsonArrowField
+from lance_namespace_urllib3_client.models.json_arrow_schema import JsonArrowSchema
+from lance_namespace_urllib3_client.models.query_table_request import QueryTableRequest
+from lance_namespace_urllib3_client.models.query_table_request_columns import (
+    QueryTableRequestColumns,
+)
+from lance_namespace_urllib3_client.models.query_table_request_full_text_query import (
+    QueryTableRequestFullTextQuery,
+)
+from lance_namespace_urllib3_client.models.query_table_request_vector import (
+    QueryTableRequestVector,
+)
+from lance_namespace_urllib3_client.models.string_fts_query import StringFtsQuery
+from lance_namespace.errors import TableNotFoundError
+from lancedb._lancedb import connect_namespace_client as _connect_namespace_client
+from lancedb.background_loop import LOOP
+from lancedb.db import AsyncConnection, DBConnection
 from lancedb.namespace_utils import (
     _normalize_create_namespace_mode,
     _normalize_drop_namespace_mode,
@@ -40,14 +56,11 @@ from lance_namespace import (
     ListNamespacesResponse,
     ListTablesResponse,
     ListTablesRequest,
-    DescribeTableRequest,
     DescribeNamespaceRequest,
     DropTableRequest,
     ListNamespacesRequest,
     CreateNamespaceRequest,
     DropNamespaceRequest,
-    DeclareTableRequest,
-    CreateTableRequest,
 )
 from lancedb.table import AsyncTable, LanceTable, Table
 from lancedb.util import validate_table_name
@@ -56,21 +69,6 @@ from lancedb.pydantic import LanceModel
 from lancedb.embeddings import EmbeddingFunctionConfig
 from ._lancedb import Session
 
-from lance_namespace_urllib3_client.models.json_arrow_schema import JsonArrowSchema
-from lance_namespace_urllib3_client.models.json_arrow_field import JsonArrowField
-from lance_namespace_urllib3_client.models.json_arrow_data_type import JsonArrowDataType
-from lance_namespace_urllib3_client.models.query_table_request import QueryTableRequest
-from lance_namespace_urllib3_client.models.query_table_request_vector import (
-    QueryTableRequestVector,
-)
-from lance_namespace_urllib3_client.models.query_table_request_columns import (
-    QueryTableRequestColumns,
-)
-from lance_namespace_urllib3_client.models.query_table_request_full_text_query import (
-    QueryTableRequestFullTextQuery,
-)
-from lance_namespace_urllib3_client.models.string_fts_query import StringFtsQuery
-
 
 def _query_to_namespace_request(
     table_id: List[str],
@@ -381,6 +379,8 @@ class LanceNamespaceDBConnection(DBConnection):
         storage_options: Optional[Dict[str, str]] = None,
         session: Optional[Session] = None,
         namespace_client_pushdown_operations: Optional[List[str]] = None,
+        namespace_client_impl: Optional[str] = None,
+        namespace_client_properties: Optional[Dict[str, str]] = None,
     ):
         """
         Initialize a namespace-based LanceDB connection.
@@ -406,12 +406,60 @@ class LanceNamespaceDBConnection(DBConnection):
               namespace.create_table() instead of using declare_table + local write.
 
             Default is None (no pushdown, all operations run locally).
+        namespace_client_impl : Optional[str]
+            The namespace implementation name used to create this connection.
+            Stored for serialization purposes.
+        namespace_client_properties : Optional[Dict[str, str]]
+            The namespace properties used to create this connection.
+            Stored for serialization purposes.
         """
         self._namespace_client = namespace_client
         self.read_consistency_interval = read_consistency_interval
         self.storage_options = storage_options or {}
         self.session = session
-        self._pushdown_operations = set(namespace_client_pushdown_operations or [])
+        self._namespace_client_pushdown_operations = set(
+            namespace_client_pushdown_operations or []
+        )
+        self._namespace_client_impl = namespace_client_impl
+        self._namespace_client_properties = namespace_client_properties
+        self._inner = AsyncConnection(
+            _connect_namespace_client(
+                namespace_client,
+                read_consistency_interval=(
+                    read_consistency_interval.total_seconds()
+                    if read_consistency_interval is not None
+                    else None
+                ),
+                storage_options=self.storage_options or None,
+                session=session,
+                namespace_client_pushdown_operations=(
+                    list(self._namespace_client_pushdown_operations)
+                ),
+                namespace_client_impl=namespace_client_impl,
+                namespace_client_properties=namespace_client_properties,
+            )
+        )
+
+    @override
+    def serialize(self) -> str:
+        import json
+
+        return json.dumps(
+            {
+                "connection_type": "namespace",
+                "namespace_client_impl": self._namespace_client_impl,
+                "namespace_client_properties": self._namespace_client_properties,
+                "namespace_client_pushdown_operations": sorted(
+                    self._namespace_client_pushdown_operations
+                ),
+                "storage_options": self.storage_options or None,
+                "read_consistency_interval_seconds": (
+                    self.read_consistency_interval.total_seconds()
+                    if self.read_consistency_interval
+                    else None
+                ),
+            }
+        )
 
     @override
     def table_names(
@@ -464,13 +512,10 @@ class LanceNamespaceDBConnection(DBConnection):
         if mode.lower() not in ["create", "overwrite"]:
             raise ValueError("mode must be either 'create' or 'overwrite'")
         validate_table_name(name)
-
+        async_table = LOOP.run(
-        table_id = namespace_path + [name]
+            self._inner.create_table(
-
+                name,
-        if "CreateTable" in self._pushdown_operations:
+                data,
-            return self._create_table_server_side(
-                name=name,
-                data=data,
                 schema=schema,
                 mode=mode,
                 exist_ok=exist_ok,
@@ -480,127 +525,15 @@ class LanceNamespaceDBConnection(DBConnection):
                 namespace_path=namespace_path,
                 storage_options=storage_options,
             )
-
-        # Local create path: declare_table + local write
-        # Step 1: Get the table location and storage options from namespace
-        # In overwrite mode, if table exists, use describe_table to get
-        # existing location. Otherwise, call create_empty_table to reserve
-        # a new location
-        location = None
-        namespace_storage_options = None
-        if mode.lower() == "overwrite":
-            # Try to describe the table first to see if it exists
-            try:
-                describe_request = DescribeTableRequest(id=table_id)
-                describe_response = self._namespace_client.describe_table(
-                    describe_request
-                )
-                location = describe_response.location
-                namespace_storage_options = describe_response.storage_options
-            except Exception:
-                # Table doesn't exist, will create a new one below
-                pass
-
-        if location is None:
-            # Table doesn't exist or mode is "create", reserve a new location
-            declare_request = DeclareTableRequest(
-                id=table_id,
-                location=None,
-                properties=self.storage_options if self.storage_options else None,
-            )
-            declare_response = self._namespace_client.declare_table(declare_request)
-
-            if not declare_response.location:
-                raise ValueError(
-                    "Table location is missing from declare_table response"
-                )
-
-            location = declare_response.location
-            namespace_storage_options = declare_response.storage_options
-
-        # Merge storage options: self.storage_options < user options < namespace options
-        merged_storage_options = dict(self.storage_options)
-        if storage_options:
-            merged_storage_options.update(storage_options)
-        if namespace_storage_options:
-            merged_storage_options.update(namespace_storage_options)
-
-        # Step 2: Create table using LanceTable.create with the location
-        # We need a temporary connection for the LanceTable.create method
-        temp_conn = LanceDBConnection(
-            location,  # Use the actual location as the connection URI
-            read_consistency_interval=self.read_consistency_interval,
-            storage_options=merged_storage_options,
-            session=self.session,
         )
 
-        # Note: storage_options_provider is auto-created in Rust from namespace_client
+        return LanceTable(
-        tbl = LanceTable.create(
+            self,
-            temp_conn,
             name,
-            data,
-            schema,
-            mode=mode,
-            exist_ok=exist_ok,
-            on_bad_vectors=on_bad_vectors,
-            fill_value=fill_value,
-            embedding_functions=embedding_functions,
             namespace_path=namespace_path,
-            storage_options=merged_storage_options,
-            location=location,
             namespace_client=self._namespace_client,
-            pushdown_operations=self._pushdown_operations,
+            pushdown_operations=self._namespace_client_pushdown_operations,
-        )
+            _async=async_table,
-
-        return tbl
-
-    def _create_table_server_side(
-        self,
-        name: str,
-        data: Optional[DATA],
-        schema: Optional[Union[pa.Schema, LanceModel]],
-        mode: str,
-        exist_ok: bool,
-        on_bad_vectors: str,
-        fill_value: float,
-        embedding_functions: Optional[List[EmbeddingFunctionConfig]],
-        namespace_path: Optional[List[str]],
-        storage_options: Optional[Dict[str, str]],
-    ) -> Table:
-        """Create a table using server-side namespace.create_table()."""
-        if namespace_path is None:
-            namespace_path = []
-        table_id = namespace_path + [name]
-
-        arrow_ipc_bytes = _data_to_arrow_ipc(
-            data=data,
-            schema=schema,
-            embedding_functions=embedding_functions,
-            on_bad_vectors=on_bad_vectors,
-            fill_value=fill_value,
-        )
-
-        request = CreateTableRequest(
-            id=table_id,
-            mode=_normalize_create_table_mode(mode),
-            properties=self.storage_options if self.storage_options else None,
-        )
-
-        try:
-            self._namespace_client.create_table(request, arrow_ipc_bytes)
-        except Exception as e:
-            if exist_ok and "already exists" in str(e).lower():
-                return self.open_table(
-                    name,
-                    namespace_path=namespace_path,
-                    storage_options=storage_options,
-                )
-            raise
-
-        return self.open_table(
-            name,
-            namespace_path=namespace_path,
-            storage_options=storage_options,
         )
 
     @override
@@ -614,30 +547,28 @@ class LanceNamespaceDBConnection(DBConnection):
     ) -> Table:
         if namespace_path is None:
             namespace_path = []
-        table_id = namespace_path + [name]
+        try:
-        request = DescribeTableRequest(id=table_id)
+            async_table = LOOP.run(
-        response = self._namespace_client.describe_table(request)
+                self._inner.open_table(
+                    name,
+                    namespace_path=namespace_path,
+                    storage_options=storage_options,
+                    index_cache_size=index_cache_size,
+                )
+            )
+        except RuntimeError as e:
+            if "Table not found" in str(e):
+                table_id = namespace_path + [name]
+                raise TableNotFoundError(f"Table not found: {'$'.join(table_id)}")
+            raise
 
-        # Merge storage options: self.storage_options < user options < namespace options
+        return LanceTable(
-        merged_storage_options = dict(self.storage_options)
+            self,
-        if storage_options:
-            merged_storage_options.update(storage_options)
-        if response.storage_options:
-            merged_storage_options.update(response.storage_options)
-
-        # Pass managed_versioning to avoid redundant describe_table call in Rust.
-        # Convert None to False since we already have the answer from describe_table.
-        managed_versioning = response.managed_versioning is True
-
-        # Note: storage_options_provider is auto-created in Rust from namespace_client
-        return self._lance_table_from_uri(
             name,
-            response.location,
             namespace_path=namespace_path,
-            storage_options=merged_storage_options,
-            index_cache_size=index_cache_size,
             namespace_client=self._namespace_client,
-            managed_versioning=managed_versioning,
+            pushdown_operations=self._namespace_client_pushdown_operations,
+            _async=async_table,
         )
 
     @override
@@ -861,35 +792,50 @@ class LanceNamespaceDBConnection(DBConnection):
         namespace_client: Optional[Any] = None,
         managed_versioning: Optional[bool] = None,
     ) -> LanceTable:
-        # Open a table directly from a URI using the location parameter
+        # Open a table directly from the namespace-resolved physical location.
-        # Note: storage_options should already be merged by the caller
+        #
-        # Note: storage_options_provider is auto-created in Rust from namespace_client
+        # Open the table through the Rust namespace-backed connection.  The Rust
+        # layer keeps the logical namespace path and namespace client intact.
         if namespace_path is None:
             namespace_path = []
-        temp_conn = LanceDBConnection(
+
-            table_uri,  # Use the table location as the connection URI
+        async_table = LOOP.run(
-            read_consistency_interval=self.read_consistency_interval,
+            self._inner.open_table(
-            storage_options=storage_options if storage_options is not None else {},
+                name,
-            session=self.session,
+                namespace_path=namespace_path,
+                storage_options=storage_options,
+                index_cache_size=index_cache_size,
+                location=None,
+                namespace_client=namespace_client,
+                managed_versioning=managed_versioning,
+            )
         )
 
-        # Open the table using the temporary connection with the location parameter
+        return LanceTable(
-        # Pass namespace_client to enable managed versioning support and auto-create
+            self,
-        # storage options provider
-        # Pass managed_versioning to avoid redundant describe_table call
-        # Pass pushdown_operations if configured on this connection
-        return LanceTable.open(
-            temp_conn,
             name,
             namespace_path=namespace_path,
-            storage_options=storage_options,
-            index_cache_size=index_cache_size,
             location=table_uri,
             namespace_client=namespace_client,
             managed_versioning=managed_versioning,
-            pushdown_operations=self._pushdown_operations,
+            pushdown_operations=self._namespace_client_pushdown_operations,
+            _async=async_table,
         )
 
+    @override
+    def namespace_client(self) -> LanceNamespace:
+        """Get the namespace client for this connection.
+
+        For namespace connections, this returns the backing namespace client
+        that was provided during construction.
+
+        Returns
+        -------
+        LanceNamespace
+            The namespace client for this connection.
+        """
+        return self._namespace_client
+
 
 class AsyncLanceNamespaceDBConnection:
     """
@@ -937,7 +883,26 @@ class AsyncLanceNamespaceDBConnection:
         self.read_consistency_interval = read_consistency_interval
         self.storage_options = storage_options or {}
         self.session = session
-        self._pushdown_operations = set(namespace_client_pushdown_operations or [])
+        self._namespace_client_pushdown_operations = set(
+            namespace_client_pushdown_operations or []
+        )
+        self._inner = AsyncConnection(
+            _connect_namespace_client(
+                namespace_client,
+                read_consistency_interval=(
+                    read_consistency_interval.total_seconds()
+                    if read_consistency_interval is not None
+                    else None
+                ),
+                storage_options=self.storage_options or None,
+                session=session,
+                namespace_client_pushdown_operations=(
+                    list(self._namespace_client_pushdown_operations)
+                ),
+                namespace_client_impl=None,
+                namespace_client_properties=None,
+            )
+        )
 
     async def table_names(
         self,
@@ -989,145 +954,16 @@ class AsyncLanceNamespaceDBConnection:
         if mode.lower() not in ["create", "overwrite"]:
             raise ValueError("mode must be either 'create' or 'overwrite'")
         validate_table_name(name)
-
+        return await self._inner.create_table(
-        table_id = namespace_path + [name]
-
-        if "CreateTable" in self._pushdown_operations:
-            return await self._create_table_server_side(
-                name=name,
-                data=data,
-                schema=schema,
-                mode=mode,
-                exist_ok=exist_ok,
-                on_bad_vectors=on_bad_vectors,
-                fill_value=fill_value,
-                embedding_functions=embedding_functions,
-                namespace_path=namespace_path,
-                storage_options=storage_options,
-            )
-
-        # Local create path: declare_table + local write
-        # Step 1: Get the table location and storage options from namespace
-        location = None
-        namespace_storage_options = None
-        if mode.lower() == "overwrite":
-            # Try to describe the table first to see if it exists
-            try:
-                describe_request = DescribeTableRequest(id=table_id)
-                describe_response = self._namespace_client.describe_table(
-                    describe_request
-                )
-                location = describe_response.location
-                namespace_storage_options = describe_response.storage_options
-            except Exception:
-                # Table doesn't exist, will create a new one below
-                pass
-
-        if location is None:
-            # Table doesn't exist or mode is "create", reserve a new location
-            declare_request = DeclareTableRequest(
-                id=table_id,
-                location=None,
-                properties=self.storage_options if self.storage_options else None,
-            )
-            declare_response = self._namespace_client.declare_table(declare_request)
-
-            if not declare_response.location:
-                raise ValueError(
-                    "Table location is missing from declare_table response"
-                )
-
-            location = declare_response.location
-            namespace_storage_options = declare_response.storage_options
-
-        # Merge storage options: self.storage_options < user options < namespace options
-        merged_storage_options = dict(self.storage_options)
-        if storage_options:
-            merged_storage_options.update(storage_options)
-        if namespace_storage_options:
-            merged_storage_options.update(namespace_storage_options)
-
-        # Step 2: Create table using LanceTable.create with the location
-        # Run the sync operation in a thread
-        def _create_table():
-            temp_conn = LanceDBConnection(
-                location,
-                read_consistency_interval=self.read_consistency_interval,
-                storage_options=merged_storage_options,
-                session=self.session,
-            )
-
-            # storage_options_provider is auto-created in Rust from namespace_client
-            return LanceTable.create(
-                temp_conn,
-                name,
-                data,
-                schema,
-                mode=mode,
-                exist_ok=exist_ok,
-                on_bad_vectors=on_bad_vectors,
-                fill_value=fill_value,
-                embedding_functions=embedding_functions,
-                namespace_path=namespace_path,
-                storage_options=merged_storage_options,
-                location=location,
-                namespace_client=self._namespace_client,
-                pushdown_operations=self._pushdown_operations,
-            )
-
-        lance_table = await asyncio.to_thread(_create_table)
-        # Get the underlying async table from LanceTable
-        return lance_table._table
-
-    async def _create_table_server_side(
-        self,
-        name: str,
-        data: Optional[DATA],
-        schema: Optional[Union[pa.Schema, LanceModel]],
-        mode: str,
-        exist_ok: bool,
-        on_bad_vectors: str,
-        fill_value: float,
-        embedding_functions: Optional[List[EmbeddingFunctionConfig]],
-        namespace_path: Optional[List[str]],
-        storage_options: Optional[Dict[str, str]],
-    ) -> AsyncTable:
-        """Create a table using server-side namespace.create_table()."""
-        if namespace_path is None:
-            namespace_path = []
-        table_id = namespace_path + [name]
-
-        def _prepare_and_create():
-            arrow_ipc_bytes = _data_to_arrow_ipc(
-                data=data,
-                schema=schema,
-                embedding_functions=embedding_functions,
-                on_bad_vectors=on_bad_vectors,
-                fill_value=fill_value,
-            )
-
-            request = CreateTableRequest(
-                id=table_id,
-                mode=_normalize_create_table_mode(mode),
-                properties=self.storage_options if self.storage_options else None,
-            )
-
-            self._namespace_client.create_table(request, arrow_ipc_bytes)
-
-        try:
-            await asyncio.to_thread(_prepare_and_create)
-        except Exception as e:
-            if exist_ok and "already exists" in str(e).lower():
-                return await self.open_table(
-                    name,
-                    namespace_path=namespace_path,
-                    storage_options=storage_options,
-                )
-            raise
-
-        return await self.open_table(
             name,
+            data,
+            schema=schema,
+            mode=mode,
+            exist_ok=exist_ok,
+            on_bad_vectors=on_bad_vectors,
+            fill_value=fill_value,
             namespace_path=namespace_path,
+            embedding_functions=embedding_functions,
             storage_options=storage_options,
         )
 
@@ -1142,45 +978,18 @@ class AsyncLanceNamespaceDBConnection:
         """Open an existing table from the namespace."""
         if namespace_path is None:
             namespace_path = []
-        table_id = namespace_path + [name]
+        try:
-        request = DescribeTableRequest(id=table_id)
+            return await self._inner.open_table(
-        response = self._namespace_client.describe_table(request)
-
-        # Merge storage options: self.storage_options < user options < namespace options
-        merged_storage_options = dict(self.storage_options)
-        if storage_options:
-            merged_storage_options.update(storage_options)
-        if response.storage_options:
-            merged_storage_options.update(response.storage_options)
-
-        # Capture managed_versioning from describe response.
-        # Convert None to False since we already have the answer from describe_table.
-        managed_versioning = response.managed_versioning is True
-
-        # Open table in a thread
-        # Note: storage_options_provider is auto-created in Rust from namespace_client
-        def _open_table():
-            temp_conn = LanceDBConnection(
-                response.location,
-                read_consistency_interval=self.read_consistency_interval,
-                storage_options=merged_storage_options,
-                session=self.session,
-            )
-
-            return LanceTable.open(
-                temp_conn,
                 name,
                 namespace_path=namespace_path,
-                storage_options=merged_storage_options,
+                storage_options=storage_options,
                 index_cache_size=index_cache_size,
-                location=response.location,
-                namespace_client=self._namespace_client,
-                managed_versioning=managed_versioning,
-                pushdown_operations=self._pushdown_operations,
             )
-
+        except RuntimeError as e:
-        lance_table = await asyncio.to_thread(_open_table)
+            if "Table not found" in str(e):
-        return lance_table._table
+                table_id = namespace_path + [name]
+                raise TableNotFoundError(f"Table not found: {'$'.join(table_id)}")
+            raise
 
     async def drop_table(self, name: str, namespace_path: Optional[List[str]] = None):
         """Drop a table from the namespace."""
@@ -1387,6 +1196,19 @@ class AsyncLanceNamespaceDBConnection:
             page_token=response.page_token,
         )
 
+    async def namespace_client(self) -> LanceNamespace:
+        """Get the namespace client for this connection.
+
+        For namespace connections, this returns the backing namespace client
+        that was provided during construction.
+
+        Returns
+        -------
+        LanceNamespace
+            The namespace client for this connection.
+        """
+        return self._namespace_client
+
 
 def connect_namespace(
     namespace_client_impl: str,
@@ -1445,6 +1267,8 @@ def connect_namespace(
         storage_options=storage_options,
         session=session,
         namespace_client_pushdown_operations=namespace_client_pushdown_operations,
+        namespace_client_impl=namespace_client_impl,
+        namespace_client_properties=namespace_client_properties,
     )
 
 

python/python/lancedb/permutation.py

@@ -284,9 +284,8 @@ class Permutations:
         self.permutation_table = permutation_table
 
         if permutation_table.schema.metadata is not None:
-            split_names = permutation_table.schema.metadata.get(
+            raw = permutation_table.schema.metadata.get(b"split_names")
-                b"split_names", None
+            split_names = raw.decode("utf-8") if raw is not None else None
-            ).decode("utf-8")
             if split_names is not None:
                 self.split_names = json.loads(split_names)
                 self.split_dict = {
@@ -460,9 +459,8 @@ class Permutation:
                         f"Cannot create a permutation on split `{split}`"
                         " because no split names are defined in the permutation table"
                     )
-                split_names = permutation_table.schema.metadata.get(
+                raw = permutation_table.schema.metadata.get(b"split_names")
-                    b"split_names", None
+                split_names = raw.decode("utf-8") if raw is not None else None
-                ).decode("utf-8")
                 if split_names is None:
                     raise ValueError(
                         f"Cannot create a permutation on split `{split}`"

python/python/lancedb/pydantic.py

@@ -10,6 +10,7 @@ import sys
 import types
 from abc import ABC, abstractmethod
 from datetime import date, datetime
+from enum import Enum
 from typing import (
     TYPE_CHECKING,
     Any,
@@ -314,6 +315,19 @@ def _pydantic_type_to_arrow_type(tp: Any, field: FieldInfo) -> pa.DataType:
                 return pa.list_(pa.list_(tp.value_arrow_type(), tp.dim()))
             # For regular Vector
             return pa.list_(tp.value_arrow_type(), tp.dim())
+        if _safe_issubclass(tp, Enum):
+            # Map Enum to the Arrow type of its value.
+            # For string-valued enums, use dictionary encoding for efficiency.
+            # For integer enums, use the native type.
+            # Fall back to utf8 for mixed-type or empty enums.
+            value_types = {type(m.value) for m in tp}
+            if len(value_types) == 1:
+                value_type = value_types.pop()
+                if value_type is str:
+                    # Use dictionary encoding for string enums
+                    return pa.dictionary(pa.int32(), pa.utf8())
+                return _py_type_to_arrow_type(value_type, field)
+            return pa.utf8()
     return _py_type_to_arrow_type(tp, field)
 
 

python/python/lancedb/remote/__init__.py

@@ -145,6 +145,33 @@ class TlsConfig:
 
 @dataclass
 class ClientConfig:
+    """Configuration for the LanceDB Cloud HTTP client.
+
+    Attributes
+    ----------
+    user_agent: str
+        User agent string sent with requests.
+    retry_config: RetryConfig
+        Configuration for retrying failed requests.
+    timeout_config: Optional[TimeoutConfig]
+        Configuration for request timeouts.
+    extra_headers: Optional[dict]
+        Additional headers to include in requests.
+    id_delimiter: Optional[str]
+        The delimiter to use when constructing object identifiers.
+    tls_config: Optional[TlsConfig]
+        TLS/mTLS configuration for secure connections.
+    header_provider: Optional[HeaderProvider]
+        Provider for dynamic headers to be added to each request.
+    user_id: Optional[str]
+        User identifier for tracking purposes. This is sent as the
+        `x-lancedb-user-id` header in requests to LanceDB Cloud/Enterprise.
+
+        This can also be set via the `LANCEDB_USER_ID` environment variable.
+        Alternatively, set `LANCEDB_USER_ID_ENV_KEY` to specify another
+        environment variable that contains the user ID value.
+    """
+
     user_agent: str = f"LanceDB-Python-Client/{__version__}"
     retry_config: RetryConfig = field(default_factory=RetryConfig)
     timeout_config: Optional[TimeoutConfig] = field(default_factory=TimeoutConfig)
@@ -152,6 +179,7 @@ class ClientConfig:
     id_delimiter: Optional[str] = None
     tls_config: Optional[TlsConfig] = None
     header_provider: Optional["HeaderProvider"] = None
+    user_id: Optional[str] = None
 
     def __post_init__(self):
         if isinstance(self.retry_config, dict):

python/python/lancedb/remote/db.py

@@ -24,6 +24,7 @@ from ..common import DATA
 from ..db import DBConnection, LOOP
 from ..embeddings import EmbeddingFunctionConfig
 from lance_namespace import (
+    LanceNamespace,
     CreateNamespaceResponse,
     DescribeNamespaceResponse,
     DropNamespaceResponse,
@@ -570,6 +571,19 @@ class RemoteDBConnection(DBConnection):
             )
         )
 
+    @override
+    def namespace_client(self) -> LanceNamespace:
+        """Get the equivalent namespace client for this connection.
+
+        Returns a RestNamespace with the same URI and authentication headers.
+
+        Returns
+        -------
+        LanceNamespace
+            The namespace client for this connection.
+        """
+        return LOOP.run(self._conn.namespace_client())
+
     async def close(self):
         """Close the connection to the database."""
         self._conn.close()

python/python/lancedb/table.py

@@ -191,7 +191,7 @@ def _into_pyarrow_reader(
             f"Unknown data type {type(data)}. "
             "Supported types: list of dicts, pandas DataFrame, polars DataFrame, "
             "pyarrow Table/RecordBatch, or Pydantic models. "
-            "See https://lancedb.com/docs/tables/ for examples."
+            "See https://docs.lancedb.com/tables/ for examples."
         )
 
 
@@ -270,15 +270,17 @@ def _sanitize_data(
         reader,
         on_bad_vectors=on_bad_vectors,
         fill_value=fill_value,
+        target_schema=target_schema,
+        metadata=metadata,
     )
 
     if target_schema is None:
         target_schema, reader = _infer_target_schema(reader)
 
     if metadata:
-        new_metadata = target_schema.metadata or {}
+        target_schema = target_schema.with_metadata(
-        new_metadata.update(metadata)
+            _merge_metadata(target_schema.metadata, metadata)
-        target_schema = target_schema.with_metadata(new_metadata)
+        )
 
     _validate_schema(target_schema)
     reader = _cast_to_target_schema(reader, target_schema, allow_subschema)
@@ -294,7 +296,7 @@ def _cast_to_target_schema(
     # pa.Table.cast expects field order not to be changed.
     # Lance doesn't care about field order, so we don't need to rearrange fields
     # to match the target schema. We just need to correctly cast the fields.
-    if reader.schema == target_schema:
+    if reader.schema.equals(target_schema, check_metadata=True):
         # Fast path when the schemas are already the same
         return reader
 
@@ -314,7 +316,13 @@ def _cast_to_target_schema(
     def gen():
         for batch in reader:
             # Table but not RecordBatch has cast.
-            yield pa.Table.from_batches([batch]).cast(reordered_schema).to_batches()[0]
+            cast_batches = (
+                pa.Table.from_batches([batch]).cast(reordered_schema).to_batches()
+            )
+            if cast_batches:
+                yield pa.RecordBatch.from_arrays(
+                    cast_batches[0].columns, schema=reordered_schema
+                )
 
     return pa.RecordBatchReader.from_batches(reordered_schema, gen())
 
@@ -332,37 +340,51 @@ def _align_field_types(
         if target_field is None:
             raise ValueError(f"Field '{field.name}' not found in target schema")
         if pa.types.is_struct(target_field.type):
-            new_type = pa.struct(
+            if pa.types.is_struct(field.type):
-                _align_field_types(
+                new_type = pa.struct(
-                    field.type.fields,
+                    _align_field_types(
-                    target_field.type.fields,
+                        field.type.fields,
+                        target_field.type.fields,
+                    )
                 )
-            )
+            else:
+                new_type = target_field.type
         elif pa.types.is_list(target_field.type):
-            new_type = pa.list_(
+            if _is_list_like(field.type):
-                _align_field_types(
+                new_type = pa.list_(
-                    [field.type.value_field],
+                    _align_field_types(
-                    [target_field.type.value_field],
+                        [field.type.value_field],
-                )[0]
+                        [target_field.type.value_field],
-            )
+                    )[0]
+                )
+            else:
+                new_type = target_field.type
         elif pa.types.is_large_list(target_field.type):
-            new_type = pa.large_list(
+            if _is_list_like(field.type):
-                _align_field_types(
+                new_type = pa.large_list(
-                    [field.type.value_field],
+                    _align_field_types(
-                    [target_field.type.value_field],
+                        [field.type.value_field],
-                )[0]
+                        [target_field.type.value_field],
-            )
+                    )[0]
+                )
+            else:
+                new_type = target_field.type
         elif pa.types.is_fixed_size_list(target_field.type):
-            new_type = pa.list_(
+            if _is_list_like(field.type):
-                _align_field_types(
+                new_type = pa.list_(
-                    [field.type.value_field],
+                    _align_field_types(
-                    [target_field.type.value_field],
+                        [field.type.value_field],
-                )[0],
+                        [target_field.type.value_field],
-                target_field.type.list_size,
+                    )[0],
-            )
+                    target_field.type.list_size,
+                )
+            else:
+                new_type = target_field.type
         else:
             new_type = target_field.type
-        new_fields.append(pa.field(field.name, new_type, field.nullable))
+        new_fields.append(
+            pa.field(field.name, new_type, field.nullable, target_field.metadata)
+        )
     return new_fields
 
 
@@ -440,6 +462,7 @@ def sanitize_create_table(
             schema = data.schema
 
     if metadata:
+        metadata = _merge_metadata(schema.metadata, metadata)
         schema = schema.with_metadata(metadata)
         # Need to apply metadata to the data as well
         if isinstance(data, pa.Table):
@@ -492,9 +515,9 @@ def _append_vector_columns(
     vector columns to the table.
     """
     if schema is None:
-        metadata = metadata or {}
+        metadata = _merge_metadata(metadata)
     else:
-        metadata = schema.metadata or metadata or {}
+        metadata = _merge_metadata(schema.metadata, metadata)
     functions = EmbeddingFunctionRegistry.get_instance().parse_functions(metadata)
 
     if not functions:
@@ -2906,6 +2929,7 @@ class LanceTable(Table):
                 namespace_path=namespace_path,
                 storage_options=storage_options,
                 location=location,
+                namespace_client=namespace_client,
             )
         )
         return self
@@ -3211,43 +3235,157 @@ def _handle_bad_vectors(
     reader: pa.RecordBatchReader,
     on_bad_vectors: Literal["error", "drop", "fill", "null"] = "error",
     fill_value: float = 0.0,
+    target_schema: Optional[pa.Schema] = None,
+    metadata: Optional[dict] = None,
 ) -> pa.RecordBatchReader:
-    vector_columns = []
+    vector_columns = _find_vector_columns(reader.schema, target_schema, metadata)
+    if not vector_columns:
+        return reader
 
-    for field in reader.schema:
+    output_schema = _vector_output_schema(reader.schema, vector_columns)
-        # They can provide a 'vector' column that isn't yet a FSL
-        named_vector_col = (
-            (
-                pa.types.is_list(field.type)
-                or pa.types.is_large_list(field.type)
-                or pa.types.is_fixed_size_list(field.type)
-            )
-            and pa.types.is_floating(field.type.value_type)
-            and field.name == VECTOR_COLUMN_NAME
-        )
-        # TODO: we're making an assumption that fixed size list of 10 or more
-        # is a vector column. This is definitely a bit hacky.
-        likely_vector_col = (
-            pa.types.is_fixed_size_list(field.type)
-            and pa.types.is_floating(field.type.value_type)
-            and (field.type.list_size >= 10)
-        )
-
-        if named_vector_col or likely_vector_col:
-            vector_columns.append(field.name)
 
     def gen():
         for batch in reader:
-            for name in vector_columns:
+            pending_dims = []
+            for vector_column in vector_columns:
+                dim = vector_column["expected_dim"]
+                if target_schema is not None and dim is None:
+                    dim = _infer_vector_dim(batch[vector_column["name"]])
+                    pending_dims.append(vector_column)
                 batch = _handle_bad_vector_column(
                     batch,
-                    vector_column_name=name,
+                    vector_column_name=vector_column["name"],
                     on_bad_vectors=on_bad_vectors,
                     fill_value=fill_value,
+                    expected_dim=dim,
+                    expected_value_type=vector_column["expected_value_type"],
                 )
-            yield batch
+            for vector_column in pending_dims:
+                if vector_column["expected_dim"] is None:
+                    vector_column["expected_dim"] = _infer_vector_dim(
+                        batch[vector_column["name"]]
+                    )
+            if batch.schema.equals(output_schema, check_metadata=True):
+                yield batch
+                continue
 
-    return pa.RecordBatchReader.from_batches(reader.schema, gen())
+            cast_batches = (
+                pa.Table.from_batches([batch]).cast(output_schema).to_batches()
+            )
+            if cast_batches:
+                yield pa.RecordBatch.from_arrays(
+                    cast_batches[0].columns,
+                    schema=output_schema,
+                )
+
+    return pa.RecordBatchReader.from_batches(output_schema, gen())
+
+
+def _find_vector_columns(
+    reader_schema: pa.Schema,
+    target_schema: Optional[pa.Schema],
+    metadata: Optional[dict],
+) -> List[dict]:
+    if target_schema is None:
+        vector_columns = []
+        for field in reader_schema:
+            named_vector_col = (
+                _is_list_like(field.type)
+                and pa.types.is_floating(field.type.value_type)
+                and field.name == VECTOR_COLUMN_NAME
+            )
+            likely_vector_col = (
+                pa.types.is_fixed_size_list(field.type)
+                and pa.types.is_floating(field.type.value_type)
+                and (field.type.list_size >= 10)
+            )
+            if named_vector_col or likely_vector_col:
+                vector_columns.append(
+                    {
+                        "name": field.name,
+                        "expected_dim": None,
+                        "expected_value_type": None,
+                    }
+                )
+        return vector_columns
+
+    reader_column_names = set(reader_schema.names)
+    active_metadata = _merge_metadata(target_schema.metadata, metadata)
+    embedding_function_columns = set(
+        EmbeddingFunctionRegistry.get_instance().parse_functions(active_metadata).keys()
+    )
+    vector_columns = []
+    for field in target_schema:
+        if field.name not in reader_column_names:
+            continue
+        if not _is_list_like(field.type) or not pa.types.is_floating(
+            field.type.value_type
+        ):
+            continue
+
+        reader_field = reader_schema.field(field.name)
+        named_vector_col = (
+            field.name in embedding_function_columns
+            or field.name == VECTOR_COLUMN_NAME
+            or (field.name == "embedding" and pa.types.is_fixed_size_list(field.type))
+        )
+        typed_fixed_vector_col = (
+            pa.types.is_fixed_size_list(reader_field.type)
+            and pa.types.is_floating(reader_field.type.value_type)
+            and reader_field.type.list_size >= 10
+        )
+
+        if named_vector_col or typed_fixed_vector_col:
+            vector_columns.append(
+                {
+                    "name": field.name,
+                    "expected_dim": (
+                        field.type.list_size
+                        if pa.types.is_fixed_size_list(field.type)
+                        else None
+                    ),
+                    "expected_value_type": field.type.value_type,
+                }
+            )
+
+    return vector_columns
+
+
+def _vector_output_schema(
+    reader_schema: pa.Schema,
+    vector_columns: List[dict],
+) -> pa.Schema:
+    columns_by_name = {column["name"]: column for column in vector_columns}
+    fields = []
+    for field in reader_schema:
+        column = columns_by_name.get(field.name)
+        if column is None:
+            output_type = field.type
+        else:
+            output_type = _vector_output_type(field, column)
+        fields.append(pa.field(field.name, output_type, field.nullable, field.metadata))
+    return pa.schema(fields, metadata=reader_schema.metadata)
+
+
+def _vector_output_type(field: pa.Field, vector_column: dict) -> pa.DataType:
+    if not _is_list_like(field.type):
+        return field.type
+
+    if vector_column["expected_value_type"] is not None and (
+        pa.types.is_null(field.type.value_type)
+        or pa.types.is_integer(field.type.value_type)
+        or pa.types.is_unsigned_integer(field.type.value_type)
+    ):
+        return pa.list_(vector_column["expected_value_type"])
+
+    if (
+        vector_column["expected_dim"] is not None
+        and pa.types.is_fixed_size_list(field.type)
+        and field.type.list_size != vector_column["expected_dim"]
+    ):
+        return pa.list_(field.type.value_type)
+
+    return field.type
 
 
 def _handle_bad_vector_column(
@@ -3255,6 +3393,8 @@ def _handle_bad_vector_column(
     vector_column_name: str,
     on_bad_vectors: str = "error",
     fill_value: float = 0.0,
+    expected_dim: Optional[int] = None,
+    expected_value_type: Optional[pa.DataType] = None,
 ) -> pa.RecordBatch:
     """
     Ensure that the vector column exists and has type fixed_size_list(float)
@@ -3271,14 +3411,39 @@ def _handle_bad_vector_column(
     fill_value: float, default 0.0
         The value to use when filling vectors. Only used if on_bad_vectors="fill".
     """
+    position = data.column_names.index(vector_column_name)
     vec_arr = data[vector_column_name]
+    if not _is_list_like(vec_arr.type):
+        return data
 
-    has_nan = has_nan_values(vec_arr)
+    if (
+        expected_dim is not None
+        and pa.types.is_fixed_size_list(vec_arr.type)
+        and vec_arr.type.list_size != expected_dim
+    ):
+        vec_arr = pa.array(vec_arr.to_pylist(), type=pa.list_(vec_arr.type.value_type))
+        data = data.set_column(position, vector_column_name, vec_arr)
 
-    if pa.types.is_fixed_size_list(vec_arr.type):
+    if expected_value_type is not None and (
+        pa.types.is_integer(vec_arr.type.value_type)
+        or pa.types.is_unsigned_integer(vec_arr.type.value_type)
+    ):
+        vec_arr = pa.array(vec_arr.to_pylist(), type=pa.list_(expected_value_type))
+        data = data.set_column(position, vector_column_name, vec_arr)
+
+    if pa.types.is_floating(vec_arr.type.value_type):
+        has_nan = has_nan_values(vec_arr)
+    else:
+        has_nan = pa.array([False] * len(vec_arr))
+
+    if expected_dim is not None:
+        dim = expected_dim
+    elif pa.types.is_fixed_size_list(vec_arr.type):
         dim = vec_arr.type.list_size
     else:
-        dim = _modal_list_size(vec_arr)
+        dim = _infer_vector_dim(vec_arr)
+        if dim is None:
+            return data
     has_wrong_dim = pc.not_equal(pc.list_value_length(vec_arr), dim)
 
     has_bad_vectors = pc.any(has_nan).as_py() or pc.any(has_wrong_dim).as_py()
@@ -3316,13 +3481,12 @@ def _handle_bad_vector_column(
                 )
             vec_arr = pc.if_else(
                 is_bad,
-                pa.scalar([fill_value] * dim),
+                pa.scalar([fill_value] * dim, type=vec_arr.type),
                 vec_arr,
             )
         else:
             raise ValueError(f"Invalid value for on_bad_vectors: {on_bad_vectors}")
 
-    position = data.column_names.index(vector_column_name)
     return data.set_column(position, vector_column_name, vec_arr)
 
 
@@ -3343,6 +3507,28 @@ def has_nan_values(arr: Union[pa.ListArray, pa.ChunkedArray]) -> pa.BooleanArray
     return pc.is_in(indices, has_nan_indices)
 
 
+def _is_list_like(data_type: pa.DataType) -> bool:
+    return (
+        pa.types.is_list(data_type)
+        or pa.types.is_large_list(data_type)
+        or pa.types.is_fixed_size_list(data_type)
+    )
+
+
+def _merge_metadata(*metadata_dicts: Optional[dict]) -> dict:
+    merged = {}
+    for metadata in metadata_dicts:
+        if metadata is None:
+            continue
+        for key, value in metadata.items():
+            if isinstance(key, str):
+                key = key.encode("utf-8")
+            if isinstance(value, str):
+                value = value.encode("utf-8")
+            merged[key] = value
+    return merged
+
+
 def _name_suggests_vector_column(field_name: str) -> bool:
     """Check if a field name indicates a vector column."""
     name_lower = field_name.lower()
@@ -3410,6 +3596,16 @@ def _modal_list_size(arr: Union[pa.ListArray, pa.ChunkedArray]) -> int:
     return pc.mode(pc.list_value_length(arr))[0].as_py()["mode"]
 
 
+def _infer_vector_dim(arr: Union[pa.Array, pa.ChunkedArray]) -> Optional[int]:
+    if not _is_list_like(arr.type):
+        return None
+    lengths = pc.list_value_length(arr)
+    lengths = pc.filter(lengths, pc.greater(lengths, 0))
+    if len(lengths) == 0:
+        return None
+    return pc.mode(lengths)[0].as_py()["mode"]
+
+
 def _validate_schema(schema: pa.Schema):
     """
     Make sure the metadata is valid utf8

python/python/tests/test_db.py

@@ -3,6 +3,7 @@
 
 
 import re
+import sys
 from datetime import timedelta
 import os
 
@@ -896,42 +897,22 @@ def test_bypass_vector_index_sync(tmp_db: lancedb.DBConnection):
 
 
 def test_local_namespace_operations(tmp_path):
-    """Test that local mode namespace operations behave as expected."""
+    """Test that local mode namespace operations work via directory namespace."""
-    # Create a local database connection
     db = lancedb.connect(tmp_path)
 
-    # Test list_namespaces returns empty list for root namespace
+    # Root namespace starts empty
-    namespaces = db.list_namespaces().namespaces
+    assert db.list_namespaces().namespaces == []
-    assert namespaces == []
 
-    # Test list_namespaces with non-empty namespace raises NotImplementedError
+    # Create and list child namespace
-    with pytest.raises(
+    db.create_namespace(["child"])
-        NotImplementedError,
+    assert "child" in db.list_namespaces().namespaces
-        match="Namespace operations are not supported for listing database",
-    ):
-        db.list_namespaces(namespace_path=["test"])
 
+    # List namespaces under child
+    assert db.list_namespaces(namespace_path=["child"]).namespaces == []
 
-def test_local_create_namespace_not_supported(tmp_path):
+    # Drop namespace
-    """Test that create_namespace is not supported in local mode."""
+    db.drop_namespace(["child"])
-    db = lancedb.connect(tmp_path)
+    assert db.list_namespaces().namespaces == []
-
-    with pytest.raises(
-        NotImplementedError,
-        match="Namespace operations are not supported for listing database",
-    ):
-        db.create_namespace(["test_namespace"])
-
-
-def test_local_drop_namespace_not_supported(tmp_path):
-    """Test that drop_namespace is not supported in local mode."""
-    db = lancedb.connect(tmp_path)
-
-    with pytest.raises(
-        NotImplementedError,
-        match="Namespace operations are not supported for listing database",
-    ):
-        db.drop_namespace(["test_namespace"])
 
 
 def test_clone_table_latest_version(tmp_path):
@@ -1048,3 +1029,59 @@ def test_clone_table_deep_clone_fails(tmp_path):
     source_uri = os.path.join(tmp_path, "source.lance")
     with pytest.raises(Exception, match="Deep clone is not yet implemented"):
         db.clone_table("cloned", source_uri, is_shallow=False)
+
+
+@pytest.mark.skipif(sys.platform == "win32", reason="Namespace client issues")
+def test_namespace_client_native_storage(tmp_path):
+    """Test namespace_client() returns DirectoryNamespace for native storage."""
+    from lance.namespace import DirectoryNamespace
+
+    db = lancedb.connect(tmp_path)
+    ns_client = db.namespace_client()
+
+    assert isinstance(ns_client, DirectoryNamespace)
+    assert str(tmp_path) in ns_client.namespace_id()
+
+
+@pytest.mark.skipif(sys.platform == "win32", reason="Namespace client issues")
+def test_namespace_client_with_storage_options(tmp_path):
+    """Test namespace_client() preserves storage options."""
+    from lance.namespace import DirectoryNamespace
+
+    storage_options = {"timeout": "10s"}
+    db = lancedb.connect(tmp_path, storage_options=storage_options)
+    ns_client = db.namespace_client()
+
+    assert isinstance(ns_client, DirectoryNamespace)
+
+
+@pytest.mark.skipif(sys.platform == "win32", reason="Namespace client issues")
+def test_namespace_client_operations(tmp_path):
+    """Test that namespace_client() returns a functional namespace client."""
+    db = lancedb.connect(tmp_path)
+    ns_client = db.namespace_client()
+
+    # Create a table through the main db connection
+    data = [{"id": 1, "text": "hello", "vector": [1.0, 2.0]}]
+    db.create_table("test_table", data=data)
+
+    # Verify the namespace client can see the table
+    from lance_namespace import ListTablesRequest
+
+    # id=[] means root namespace
+    response = ns_client.list_tables(ListTablesRequest(id=[]))
+    # Tables can be strings or objects with name attribute
+    table_names = [t.name if hasattr(t, "name") else t for t in response.tables]
+    assert "test_table" in table_names
+
+
+@pytest.mark.skipif(sys.platform == "win32", reason="Namespace client issues")
+def test_namespace_client_namespace_connection(tmp_path):
+    """Test namespace_client() returns the backing client for namespace connections."""
+    from lance.namespace import DirectoryNamespace
+
+    db = lancedb.connect_namespace("dir", {"root": str(tmp_path)})
+    ns_client = db.namespace_client()
+
+    assert isinstance(ns_client, DirectoryNamespace)
+    assert str(tmp_path) in ns_client.namespace_id()

python/python/tests/test_namespace.py

@@ -681,7 +681,7 @@ class TestPushdownOperations:
             {"root": self.temp_dir},
             namespace_client_pushdown_operations=["QueryTable"],
         )
-        assert "QueryTable" in db._pushdown_operations
+        assert "QueryTable" in db._namespace_client_pushdown_operations
 
     def test_create_table_pushdown_stored(self):
         """Test that CreateTable pushdown is stored on sync connection."""
@@ -690,7 +690,7 @@ class TestPushdownOperations:
             {"root": self.temp_dir},
             namespace_client_pushdown_operations=["CreateTable"],
         )
-        assert "CreateTable" in db._pushdown_operations
+        assert "CreateTable" in db._namespace_client_pushdown_operations
 
     def test_both_pushdowns_stored(self):
         """Test that both pushdown operations can be set together."""
@@ -699,13 +699,13 @@ class TestPushdownOperations:
             {"root": self.temp_dir},
             namespace_client_pushdown_operations=["QueryTable", "CreateTable"],
         )
-        assert "QueryTable" in db._pushdown_operations
+        assert "QueryTable" in db._namespace_client_pushdown_operations
-        assert "CreateTable" in db._pushdown_operations
+        assert "CreateTable" in db._namespace_client_pushdown_operations
 
     def test_pushdown_defaults_to_empty(self):
         """Test that pushdown operations default to empty."""
         db = lancedb.connect_namespace("dir", {"root": self.temp_dir})
-        assert len(db._pushdown_operations) == 0
+        assert len(db._namespace_client_pushdown_operations) == 0
 
 
 @pytest.mark.asyncio
@@ -727,7 +727,7 @@ class TestAsyncPushdownOperations:
             {"root": self.temp_dir},
             namespace_client_pushdown_operations=["QueryTable"],
         )
-        assert "QueryTable" in db._pushdown_operations
+        assert "QueryTable" in db._namespace_client_pushdown_operations
 
     async def test_async_create_table_pushdown_stored(self):
         """Test that CreateTable pushdown is stored on async connection."""
@@ -736,9 +736,9 @@ class TestAsyncPushdownOperations:
             {"root": self.temp_dir},
             namespace_client_pushdown_operations=["CreateTable"],
         )
-        assert "CreateTable" in db._pushdown_operations
+        assert "CreateTable" in db._namespace_client_pushdown_operations
 
     async def test_async_pushdown_defaults_to_empty(self):
         """Test that pushdown operations default to empty on async connection."""
         db = lancedb.connect_namespace_async("dir", {"root": self.temp_dir})
-        assert len(db._pushdown_operations) == 0
+        assert len(db._namespace_client_pushdown_operations) == 0

python/python/tests/test_namespace_integration.py

@@ -18,6 +18,9 @@ Tests verify:
 """
 
 import copy
+import shutil
+import sys
+import tempfile
 import time
 import uuid
 from typing import Dict, Optional
@@ -387,6 +390,66 @@ def test_namespace_open_table_with_provider(s3_bucket: str, use_custom: bool):
     assert get_describe_call_count(inner_ns_client) == describe_count_after_open
 
 
+@pytest.mark.skipif(
+    sys.platform == "win32",
+    reason="TODO: fix schema-only namespace metrics test on Windows",
+)
+@pytest.mark.parametrize("use_custom", [False, True], ids=["DirectoryNS", "CustomNS"])
+def test_namespace_create_schema_only_with_provider(use_custom: bool):
+    """
+    Test creating a schema-only table through namespace with storage options provider.
+
+    Verifies:
+    - declare_table is called once to reserve the location
+    - describe_table is not needed during create in create mode
+    - the table can be reopened successfully afterward
+    - opening the table triggers exactly one describe_table call
+    """
+    temp_dir = tempfile.mkdtemp()
+    try:
+        ns_client, inner_ns_client = create_tracking_namespace(
+            bucket_name=temp_dir,
+            storage_options={},
+            credential_expires_in_seconds=3600,
+            use_custom=use_custom,
+        )
+
+        db = LanceNamespaceDBConnection(ns_client)
+
+        namespace_name = f"test_ns_{uuid.uuid4().hex[:8]}"
+        db.create_namespace([namespace_name])
+
+        table_name = f"test_table_{uuid.uuid4().hex}"
+        namespace_path = [namespace_name]
+        schema = pa.schema(
+            [
+                pa.field("id", pa.int64()),
+                pa.field("vector", pa.list_(pa.float32(), 2)),
+                pa.field("text", pa.string()),
+            ]
+        )
+
+        assert get_declare_call_count(inner_ns_client) == 0
+        assert get_describe_call_count(inner_ns_client) == 0
+
+        table = db.create_table(
+            table_name, schema=schema, namespace_path=namespace_path
+        )
+
+        assert table.name == table_name
+        assert table.namespace == namespace_path
+        assert get_declare_call_count(inner_ns_client) == 1
+        assert get_describe_call_count(inner_ns_client) == 0
+
+        reopened_table = db.open_table(table_name, namespace_path=namespace_path)
+
+        assert reopened_table.schema == schema
+        assert get_declare_call_count(inner_ns_client) == 1
+        assert get_describe_call_count(inner_ns_client) == 1
+    finally:
+        shutil.rmtree(temp_dir, ignore_errors=True)
+
+
 @pytest.mark.s3_test
 @pytest.mark.parametrize("use_custom", [False, True], ids=["DirectoryNS", "CustomNS"])
 def test_namespace_credential_refresh_on_read(s3_bucket: str, use_custom: bool):

python/python/tests/test_permutation.py

@@ -522,6 +522,50 @@ def test_no_split_names(some_table: Table):
     assert permutations[1].num_rows == 500
 
 
+def test_permutations_metadata_without_split_names_key(mem_db: DBConnection):
+    """Regression: schema metadata present but missing split_names key must not crash.
+
+    Previously, `.get(b"split_names", None).decode()` was called unconditionally,
+    so any permutation table whose metadata dict had other keys but no split_names
+    raised AttributeError: 'NoneType' has no attribute 'decode'.
+    """
+    base = mem_db.create_table("base_nosplit", pa.table({"x": range(10)}))
+
+    # Build a permutation-like table that carries some metadata but NOT split_names.
+    raw = pa.table(
+        {
+            "row_id": pa.array(range(10), type=pa.uint64()),
+            "split_id": pa.array([0] * 10, type=pa.uint32()),
+        }
+    ).replace_schema_metadata({b"other_key": b"other_value"})
+    perm_tbl = mem_db.create_table("perm_nosplit", raw)
+
+    permutations = Permutations(base, perm_tbl)
+    assert permutations.split_names == []
+    assert permutations.split_dict == {}
+
+
+def test_from_tables_string_split_missing_names_key(mem_db: DBConnection):
+    """Regression: from_tables() with a string split must raise ValueError, not
+    AttributeError.
+
+    Previously, `.get(b"split_names", None).decode()` crashed with AttributeError
+    when the metadata dict existed but had no split_names key.
+    """
+    base = mem_db.create_table("base_strsplit", pa.table({"x": range(10)}))
+
+    raw = pa.table(
+        {
+            "row_id": pa.array(range(10), type=pa.uint64()),
+            "split_id": pa.array([0] * 10, type=pa.uint32()),
+        }
+    ).replace_schema_metadata({b"other_key": b"other_value"})
+    perm_tbl = mem_db.create_table("perm_strsplit", raw)
+
+    with pytest.raises(ValueError, match="no split names are defined"):
+        Permutation.from_tables(base, perm_tbl, split="train")
+
+
 @pytest.fixture
 def some_perm_table(some_table: Table) -> Table:
     return (

python/python/tests/test_pydantic.py

@@ -3,6 +3,7 @@
 
 import json
 from datetime import date, datetime
+from enum import Enum
 from typing import List, Optional, Tuple
 
 import pyarrow as pa
@@ -673,3 +674,29 @@ async def test_aliases_in_lance_model_async(mem_db_async):
     assert hasattr(model, "name")
     assert hasattr(model, "distance")
     assert model.distance < 0.01
+
+
+def test_enum_types():
+    """Enum fields should map to the Arrow type of their value (issue #1846)."""
+
+    class StrStatus(str, Enum):
+        PENDING = "pending"
+        RUNNING = "running"
+        DONE = "done"
+
+    class IntPriority(int, Enum):
+        LOW = 1
+        MEDIUM = 2
+        HIGH = 3
+
+    class TestModel(pydantic.BaseModel):
+        status: StrStatus
+        priority: IntPriority
+        opt_status: Optional[StrStatus] = None
+
+    schema = pydantic_to_schema(TestModel)
+
+    assert schema.field("status").type == pa.dictionary(pa.int32(), pa.utf8())
+    assert schema.field("priority").type == pa.int64()
+    assert schema.field("opt_status").type == pa.dictionary(pa.int32(), pa.utf8())
+    assert schema.field("opt_status").nullable

python/python/tests/test_table.py

@@ -3,6 +3,7 @@
 
 
 import os
+import sys
 from datetime import date, datetime, timedelta
 from time import sleep
 from typing import List
@@ -1049,6 +1050,231 @@ def test_add_with_nans(mem_db: DBConnection):
     assert np.allclose(v, np.array([0.0, 0.0]))
 
 
+def test_add_with_empty_fixed_size_list_drops_bad_rows(mem_db: DBConnection):
+    class Schema(LanceModel):
+        text: str
+        embedding: Vector(16)
+
+    table = mem_db.create_table("test_empty_embeddings", schema=Schema)
+    table.add(
+        [
+            {"text": "hello", "embedding": []},
+            {"text": "bar", "embedding": [0.1] * 16},
+        ],
+        on_bad_vectors="drop",
+    )
+
+    data = table.to_arrow()
+    assert data["text"].to_pylist() == ["bar"]
+    assert np.allclose(data["embedding"].to_pylist()[0], np.array([0.1] * 16))
+
+
+def test_add_with_integer_embeddings_preserves_casting(mem_db: DBConnection):
+    class Schema(LanceModel):
+        text: str
+        embedding: Vector(4)
+
+    table = mem_db.create_table("test_integer_embeddings", schema=Schema)
+    table.add(
+        [{"text": "foo", "embedding": [1, 2, 3, 4]}],
+        on_bad_vectors="drop",
+    )
+
+    assert table.to_arrow()["embedding"].to_pylist() == [[1.0, 2.0, 3.0, 4.0]]
+
+
+def test_on_bad_vectors_does_not_handle_non_vector_fixed_size_lists(
+    mem_db: DBConnection,
+):
+    schema = pa.schema(
+        [
+            pa.field("vector", pa.list_(pa.float32(), 4)),
+            pa.field("bbox", pa.list_(pa.float32(), 4)),
+        ]
+    )
+    table = mem_db.create_table("test_bbox_schema", schema=schema)
+
+    with pytest.raises(RuntimeError, match="FixedSizeListType"):
+        table.add(
+            [{"vector": [1.0, 2.0, 3.0, 4.0], "bbox": [0.0, 1.0]}],
+            on_bad_vectors="drop",
+        )
+
+
+def test_on_bad_vectors_does_not_handle_custom_named_fixed_size_lists(
+    mem_db: DBConnection,
+):
+    schema = pa.schema([pa.field("features", pa.list_(pa.float32(), 16))])
+    table = mem_db.create_table("test_custom_named_fixed_size_vector", schema=schema)
+
+    with pytest.raises(RuntimeError, match="FixedSizeListType"):
+        table.add(
+            [
+                {"features": []},
+                {"features": [0.1] * 16},
+            ],
+            on_bad_vectors="drop",
+        )
+
+
+def test_on_bad_vectors_with_schema_list_vector_still_sanitizes(mem_db: DBConnection):
+    schema = pa.schema([pa.field("vector", pa.list_(pa.float32()))])
+    table = mem_db.create_table("test_schema_list_vector", schema=schema)
+    table.add(
+        [
+            {"vector": [1.0, 2.0]},
+            {"vector": [3.0]},
+            {"vector": [4.0, 5.0]},
+        ],
+        on_bad_vectors="drop",
+    )
+
+    assert table.to_arrow()["vector"].to_pylist() == [[1.0, 2.0], [4.0, 5.0]]
+
+
+def test_on_bad_vectors_handles_typed_custom_fixed_vectors_for_list_schema(
+    mem_db: DBConnection,
+):
+    schema = pa.schema([pa.field("vec", pa.list_(pa.float32()))])
+    table = mem_db.create_table("test_typed_custom_fixed_vector", schema=schema)
+    data = pa.table(
+        {
+            "vec": pa.array(
+                [[float("nan")] * 16, [1.0] * 16],
+                type=pa.list_(pa.float32(), 16),
+            )
+        }
+    )
+
+    table.add(data, on_bad_vectors="drop")
+
+    assert table.to_arrow()["vec"].to_pylist() == [[1.0] * 16]
+
+
+def test_on_bad_vectors_fill_preserves_arrow_nested_vector_type(mem_db: DBConnection):
+    schema = pa.schema([pa.field("vector", pa.list_(pa.float32()))])
+    table = mem_db.create_table("test_fill_arrow_nested_type", schema=schema)
+    data = pa.table(
+        {
+            "vector": pa.array(
+                [[1.0, 2.0], [float("nan"), 3.0]],
+                type=pa.list_(pa.float32(), 2),
+            )
+        }
+    )
+
+    table.add(
+        data,
+        on_bad_vectors="fill",
+        fill_value=0.0,
+    )
+
+    assert table.to_arrow()["vector"].to_pylist() == [[1.0, 2.0], [0.0, 0.0]]
+
+
+@pytest.mark.parametrize(
+    ("table_name", "batch1", "expected"),
+    [
+        (
+            "test_schema_list_vector_empty_prefix",
+            pa.record_batch({"vector": [[], []]}),
+            [[], [], [1.0, 2.0], [3.0, 4.0]],
+        ),
+        (
+            "test_schema_list_vector_all_bad_prefix",
+            pa.record_batch({"vector": [[float("nan")] * 3, [float("nan")] * 3]}),
+            [[1.0, 2.0], [3.0, 4.0]],
+        ),
+    ],
+)
+def test_on_bad_vectors_with_schema_list_vector_ignores_invalid_prefix_batches(
+    mem_db: DBConnection,
+    table_name: str,
+    batch1: pa.RecordBatch,
+    expected: list,
+):
+    schema = pa.schema([pa.field("vector", pa.list_(pa.float32()))])
+    table = mem_db.create_table(table_name, schema=schema)
+    batch2 = pa.record_batch({"vector": [[1.0, 2.0], [3.0, 4.0]]})
+    reader = pa.RecordBatchReader.from_batches(batch1.schema, [batch1, batch2])
+
+    table.add(reader, on_bad_vectors="drop")
+
+    assert table.to_arrow()["vector"].to_pylist() == expected
+
+
+def test_on_bad_vectors_with_multiple_vectors_locks_dim_after_final_drop(
+    mem_db: DBConnection,
+):
+    registry = EmbeddingFunctionRegistry.get_instance()
+    func = MockTextEmbeddingFunction.create()
+    metadata = registry.get_table_metadata(
+        [
+            EmbeddingFunctionConfig(
+                source_column="text1", vector_column="vec1", function=func
+            ),
+            EmbeddingFunctionConfig(
+                source_column="text2", vector_column="vec2", function=func
+            ),
+        ]
+    )
+    schema = pa.schema(
+        [
+            pa.field("vec1", pa.list_(pa.float32())),
+            pa.field("vec2", pa.list_(pa.float32())),
+        ],
+        metadata=metadata,
+    )
+    table = mem_db.create_table("test_multi_vector_dim_lock", schema=schema)
+    batch1 = pa.record_batch(
+        {
+            "vec1": [[1.0, 2.0, 3.0], [10.0, 11.0]],
+            "vec2": [[float("nan"), 0.0], [5.0, 6.0]],
+        }
+    )
+    batch2 = pa.record_batch(
+        {
+            "vec1": [[20.0, 21.0], [30.0, 31.0]],
+            "vec2": [[7.0, 8.0], [9.0, 10.0]],
+        }
+    )
+    reader = pa.RecordBatchReader.from_batches(batch1.schema, [batch1, batch2])
+
+    table.add(reader, on_bad_vectors="drop")
+
+    data = table.to_arrow()
+    assert data["vec1"].to_pylist() == [[10.0, 11.0], [20.0, 21.0], [30.0, 31.0]]
+    assert data["vec2"].to_pylist() == [[5.0, 6.0], [7.0, 8.0], [9.0, 10.0]]
+
+
+def test_on_bad_vectors_does_not_handle_non_vector_list_columns(mem_db: DBConnection):
+    schema = pa.schema([pa.field("embedding_history", pa.list_(pa.float32()))])
+    table = mem_db.create_table("test_non_vector_list_schema", schema=schema)
+    table.add(
+        [
+            {"embedding_history": [1.0, 2.0]},
+            {"embedding_history": [3.0]},
+        ],
+        on_bad_vectors="drop",
+    )
+
+    assert table.to_arrow()["embedding_history"].to_pylist() == [
+        [1.0, 2.0],
+        [3.0],
+    ]
+
+
+def test_on_bad_vectors_all_null_schema_vector_batches_do_not_crash(
+    mem_db: DBConnection,
+):
+    schema = pa.schema([pa.field("vector", pa.list_(pa.float32(), 2), nullable=True)])
+    table = mem_db.create_table("test_all_null_vector_batch", schema=schema)
+
+    table.add([{"vector": None}], on_bad_vectors="drop")
+
+    assert table.to_arrow()["vector"].to_pylist() == [None]
+
+
 def test_restore(mem_db: DBConnection):
     table = mem_db.create_table(
         "my_table",
@@ -1815,6 +2041,13 @@ def test_hybrid_search_metric_type(tmp_db: DBConnection):
 @pytest.mark.parametrize(
     "consistency_interval", [None, timedelta(seconds=0), timedelta(seconds=0.1)]
 )
+@pytest.mark.skipif(
+    sys.platform == "win32",
+    reason=(
+        "TODO: directory namespace is not supported on Windows yet; "
+        "re-enable after that is fixed."
+    ),
+)
 def test_consistency(tmp_path, consistency_interval):
     db = lancedb.connect(tmp_path)
     table = db.create_table("my_table", data=[{"id": 0}])
@@ -1835,7 +2068,6 @@ def test_consistency(tmp_path, consistency_interval):
     elif consistency_interval == timedelta(seconds=0):
         assert table2.version == table.version
     else:
-        # (consistency_interval == timedelta(seconds=0.1)
         assert table2.version == table.version - 1
         sleep(0.1)
         assert table2.version == table.version

python/python/tests/test_util.py

@@ -15,8 +15,10 @@ from lancedb.table import (
     _cast_to_target_schema,
     _handle_bad_vectors,
     _into_pyarrow_reader,
-    _sanitize_data,
     _infer_target_schema,
+    _merge_metadata,
+    _sanitize_data,
+    sanitize_create_table,
 )
 import pyarrow as pa
 import pandas as pd
@@ -304,6 +306,117 @@ def test_handle_bad_vectors_noop():
     assert output["vector"] == vector
 
 
+def test_handle_bad_vectors_updates_reader_schema_for_target_schema():
+    data = pa.table({"vector": [[1, 2, 3, 4]]})
+    target_schema = pa.schema([pa.field("vector", pa.list_(pa.float32(), 4))])
+
+    output = _handle_bad_vectors(
+        data.to_reader(),
+        on_bad_vectors="drop",
+        target_schema=target_schema,
+    )
+
+    assert output.schema == pa.schema([pa.field("vector", pa.list_(pa.float32()))])
+    assert output.read_all()["vector"].to_pylist() == [[1.0, 2.0, 3.0, 4.0]]
+
+
+def test_sanitize_data_keeps_target_field_metadata():
+    source_field = pa.field(
+        "vector",
+        pa.list_(pa.float32(), 2),
+        metadata={b"source": b"drop-me"},
+    )
+    target_field = pa.field(
+        "vector",
+        pa.list_(pa.float32(), 2),
+        metadata={b"target": b"keep-me"},
+    )
+    data = pa.table(
+        {"vector": pa.array([[1.0, 2.0]], type=pa.list_(pa.float32(), 2))},
+        schema=pa.schema([source_field]),
+    )
+
+    output = _sanitize_data(
+        data,
+        target_schema=pa.schema([target_field]),
+        on_bad_vectors="drop",
+    ).read_all()
+
+    assert output.schema.field("vector").metadata == {b"target": b"keep-me"}
+
+
+def test_sanitize_data_uses_separate_embedding_metadata_for_bad_vectors():
+    registry = EmbeddingFunctionRegistry.get_instance()
+    conf = EmbeddingFunctionConfig(
+        source_column="text",
+        vector_column="custom_vector",
+        function=MockTextEmbeddingFunction.create(),
+    )
+    metadata = registry.get_table_metadata([conf])
+    schema = pa.schema(
+        {
+            "text": pa.string(),
+            "custom_vector": pa.list_(pa.float32(), 10),
+        },
+        metadata={b"note": b"keep-me"},
+    )
+    data = pa.table(
+        {
+            "text": ["bad", "good"],
+            "custom_vector": [[1.0] * 9, [2.0] * 10],
+        }
+    )
+
+    output = _sanitize_data(
+        data,
+        target_schema=schema,
+        metadata=metadata,
+        on_bad_vectors="drop",
+    ).read_all()
+
+    assert output["text"].to_pylist() == ["good"]
+    assert output.schema.metadata[b"note"] == b"keep-me"
+    assert b"embedding_functions" in output.schema.metadata
+
+
+def test_sanitize_create_table_merges_and_overrides_embedding_metadata():
+    registry = EmbeddingFunctionRegistry.get_instance()
+    old_conf = EmbeddingFunctionConfig(
+        source_column="text",
+        vector_column="old_vector",
+        function=MockTextEmbeddingFunction.create(),
+    )
+    new_conf = EmbeddingFunctionConfig(
+        source_column="text",
+        vector_column="custom_vector",
+        function=MockTextEmbeddingFunction.create(),
+    )
+    metadata = registry.get_table_metadata([new_conf])
+    schema = pa.schema(
+        {
+            "text": pa.string(),
+            "custom_vector": pa.list_(pa.float32(), 10),
+        },
+        metadata=_merge_metadata(
+            {b"note": b"keep-me"},
+            registry.get_table_metadata([old_conf]),
+        ),
+    )
+
+    data, schema = sanitize_create_table(
+        pa.table({"text": ["good"]}),
+        schema,
+        metadata=metadata,
+        on_bad_vectors="drop",
+    )
+
+    assert schema.metadata[b"note"] == b"keep-me"
+    assert b"embedding_functions" in schema.metadata
+    assert data.schema.metadata[b"note"] == b"keep-me"
+    funcs = EmbeddingFunctionRegistry.get_instance().parse_functions(schema.metadata)
+    assert set(funcs.keys()) == {"custom_vector"}
+
+
 class TestModel(lancedb.pydantic.LanceModel):
     a: Optional[int]
     b: Optional[int]

python/src/connection.rs

@@ -1,11 +1,17 @@
 // SPDX-License-Identifier: Apache-2.0
 // SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
-use std::{collections::HashMap, sync::Arc, time::Duration};
+use std::{
+    collections::{HashMap, HashSet},
+    sync::Arc,
+    time::Duration,
+};
 
 use arrow::{datatypes::Schema, ffi_stream::ArrowArrayStreamReader, pyarrow::FromPyArrow};
 use lancedb::{
     connection::Connection as LanceConnection,
+    connection::NamespaceClientPushdownOperation,
+    database::namespace::LanceNamespaceDatabase,
     database::{CreateTableMode, Database, ReadConsistency},
 };
 use pyo3::{
@@ -39,6 +45,29 @@ impl Connection {
     }
 }
 
+fn parse_namespace_client_pushdown_operations(
+    operations: Option<Vec<String>>,
+) -> PyResult<HashSet<NamespaceClientPushdownOperation>> {
+    let mut parsed = HashSet::new();
+    for operation in operations.unwrap_or_default() {
+        match operation.as_str() {
+            "QueryTable" => {
+                parsed.insert(NamespaceClientPushdownOperation::QueryTable);
+            }
+            "CreateTable" => {
+                parsed.insert(NamespaceClientPushdownOperation::CreateTable);
+            }
+            _ => {
+                return Err(PyValueError::new_err(format!(
+                    "Invalid pushdown operation: {}",
+                    operation
+                )));
+            }
+        }
+    }
+    Ok(parsed)
+}
+
 impl Connection {
     fn parse_create_mode_str(mode: &str) -> PyResult<CreateTableMode> {
         match mode {
@@ -474,6 +503,25 @@ impl Connection {
             })
         })
     }
+
+    /// Get the configuration for constructing an equivalent namespace client.
+    /// Returns a dict with:
+    /// - "impl": "dir" for DirectoryNamespace, "rest" for RestNamespace
+    /// - "properties": configuration properties for the namespace
+    #[pyo3(signature = ())]
+    pub fn namespace_client_config(self_: PyRef<'_, Self>) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.get_inner()?.clone();
+        let py = self_.py();
+        future_into_py(py, async move {
+            let (impl_type, properties) = inner.namespace_client_config().await.infer_error()?;
+            Python::attach(|py| -> PyResult<Py<PyDict>> {
+                let dict = PyDict::new(py);
+                dict.set_item("impl", impl_type)?;
+                dict.set_item("properties", properties)?;
+                Ok(dict.unbind())
+            })
+        })
+    }
 }
 
 #[pyfunction]
@@ -519,6 +567,52 @@ pub fn connect(
     })
 }
 
+#[pyfunction]
+#[pyo3(signature = (
+    namespace_client,
+    read_consistency_interval=None,
+    storage_options=None,
+    session=None,
+    namespace_client_pushdown_operations=None,
+    namespace_client_impl=None,
+    namespace_client_properties=None,
+))]
+#[allow(clippy::too_many_arguments)]
+pub fn connect_namespace_client(
+    py: Python<'_>,
+    namespace_client: Py<PyAny>,
+    read_consistency_interval: Option<f64>,
+    storage_options: Option<HashMap<String, String>>,
+    session: Option<crate::session::Session>,
+    namespace_client_pushdown_operations: Option<Vec<String>>,
+    namespace_client_impl: Option<String>,
+    namespace_client_properties: Option<HashMap<String, String>>,
+) -> PyResult<Connection> {
+    let namespace_client = extract_namespace_arc(py, namespace_client)?;
+    let read_consistency_interval = read_consistency_interval.map(Duration::from_secs_f64);
+    let namespace_client_pushdown_operations =
+        parse_namespace_client_pushdown_operations(namespace_client_pushdown_operations)?;
+    let ns_impl = namespace_client_impl.unwrap_or_else(|| "python".to_string());
+    let ns_properties = namespace_client_properties.unwrap_or_default();
+    let storage_options = storage_options.unwrap_or_default();
+    let session = session.map(|s| s.inner.clone());
+
+    let database = LanceNamespaceDatabase::from_namespace_client(
+        namespace_client,
+        ns_impl,
+        ns_properties,
+        storage_options,
+        read_consistency_interval,
+        session,
+        namespace_client_pushdown_operations,
+    );
+
+    Ok(Connection::new(LanceConnection::new(
+        Arc::new(database),
+        Arc::new(lancedb::embeddings::MemoryRegistry::new()),
+    )))
+}
+
 #[derive(FromPyObject)]
 pub struct PyClientConfig {
     user_agent: String,
@@ -528,6 +622,7 @@ pub struct PyClientConfig {
     id_delimiter: Option<String>,
     tls_config: Option<PyClientTlsConfig>,
     header_provider: Option<Py<PyAny>>,
+    user_id: Option<String>,
 }
 
 #[derive(FromPyObject)]
@@ -612,6 +707,7 @@ impl From<PyClientConfig> for lancedb::remote::ClientConfig {
             id_delimiter: value.id_delimiter,
             tls_config: value.tls_config.map(Into::into),
             header_provider,
+            user_id: value.user_id,
         }
     }
 }

python/src/lib.rs

@@ -2,7 +2,7 @@
 // SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
 use arrow::RecordBatchStream;
-use connection::{Connection, connect};
+use connection::{Connection, connect, connect_namespace_client};
 use env_logger::Env;
 use expr::{PyExpr, expr_col, expr_func, expr_lit};
 use index::IndexConfig;
@@ -58,6 +58,7 @@ pub fn _lancedb(_py: Python, m: &Bound<'_, PyModule>) -> PyResult<()> {
     m.add_class::<PyPermutationReader>()?;
     m.add_class::<PyExpr>()?;
     m.add_function(wrap_pyfunction!(connect, m)?)?;
+    m.add_function(wrap_pyfunction!(connect_namespace_client, m)?)?;
     m.add_function(wrap_pyfunction!(permutation::async_permutation_builder, m)?)?;
     m.add_function(wrap_pyfunction!(util::validate_table_name, m)?)?;
     m.add_function(wrap_pyfunction!(query::fts_query_to_json, m)?)?;

rust-toolchain.toml

@@ -1,2 +1,2 @@
 [toolchain]
-channel = "1.91.0"
+channel = "1.94.0"

rust/lancedb/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb"
-version = "0.27.2"
+version = "0.28.0-beta.7"
 edition.workspace = true
 description = "LanceDB: A serverless, low-latency vector database for AI applications"
 license.workspace = true

rust/lancedb/src/connection.rs

@@ -171,7 +171,7 @@ impl OpenTableBuilder {
     /// Options already set on the connection will be inherited by the table,
     /// but can be overridden here.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_option(mut self, key: impl Into<String>, value: impl Into<String>) -> Self {
         let store_params = self
             .request
@@ -188,7 +188,7 @@ impl OpenTableBuilder {
     /// Options already set on the connection will be inherited by the table,
     /// but can be overridden here.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_options(
         mut self,
         pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
@@ -541,6 +541,16 @@ impl Connection {
         self.internal.namespace_client().await
     }
 
+    /// Get the configuration for constructing an equivalent namespace client.
+    /// Returns (impl_type, properties) where:
+    /// - impl_type: "dir" for DirectoryNamespace, "rest" for RestNamespace
+    /// - properties: configuration properties for the namespace
+    pub async fn namespace_client_config(
+        &self,
+    ) -> Result<(String, std::collections::HashMap<String, String>)> {
+        self.internal.namespace_client_config().await
+    }
+
     /// List tables with pagination support
     pub async fn list_tables(&self, request: ListTablesRequest) -> Result<ListTablesResponse> {
         self.internal.list_tables(request).await
@@ -572,6 +582,14 @@ pub struct ConnectRequest {
     /// Database specific options
     pub options: HashMap<String, String>,
 
+    /// Extra properties for the equivalent namespace client.
+    ///
+    /// For a local [`ListingDatabase`], these are merged into the backing
+    /// `DirectoryNamespace` properties. This is useful for namespace-specific
+    /// settings such as `table_version_tracking_enabled` that are distinct from
+    /// storage options.
+    pub namespace_client_properties: HashMap<String, String>,
+
     /// The interval at which to check for updates from other processes.
     ///
     /// If None, then consistency is not checked. For performance
@@ -611,6 +629,7 @@ impl ConnectBuilder {
                 client_config: Default::default(),
                 read_consistency_interval: None,
                 options: HashMap::new(),
+                namespace_client_properties: HashMap::new(),
                 session: None,
             },
             embedding_registry: None,
@@ -728,7 +747,7 @@ impl ConnectBuilder {
 
     /// Set an option for the storage layer.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_option(mut self, key: impl Into<String>, value: impl Into<String>) -> Self {
         self.request.options.insert(key.into(), value.into());
         self
@@ -736,7 +755,7 @@ impl ConnectBuilder {
 
     /// Set multiple options for the storage layer.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_options(
         mut self,
         pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
@@ -747,6 +766,31 @@ impl ConnectBuilder {
         self
     }
 
+    /// Set an additional property for the equivalent namespace client.
+    pub fn namespace_client_property(
+        mut self,
+        key: impl Into<String>,
+        value: impl Into<String>,
+    ) -> Self {
+        self.request
+            .namespace_client_properties
+            .insert(key.into(), value.into());
+        self
+    }
+
+    /// Set multiple additional properties for the equivalent namespace client.
+    pub fn namespace_client_properties(
+        mut self,
+        pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
+    ) -> Self {
+        for (key, value) in pairs {
+            self.request
+                .namespace_client_properties
+                .insert(key.into(), value.into());
+        }
+        self
+    }
+
     /// The interval at which to check for updates from other processes. This
     /// only affects LanceDB OSS.
     ///
@@ -871,7 +915,7 @@ use std::collections::HashSet;
 /// These operations will be executed on the namespace server instead of locally
 /// when enabled via [`ConnectNamespaceBuilder::pushdown_operations`].
 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
-pub enum PushdownOperation {
+pub enum NamespaceClientPushdownOperation {
     /// Execute queries on the namespace server via `query_table()` instead of locally.
     QueryTable,
     /// Execute table creation on the namespace server via `create_table()`
@@ -883,10 +927,11 @@ pub struct ConnectNamespaceBuilder {
     ns_impl: String,
     properties: HashMap<String, String>,
     storage_options: HashMap<String, String>,
+    namespace_client_properties: HashMap<String, String>,
     read_consistency_interval: Option<std::time::Duration>,
     embedding_registry: Option<Arc<dyn EmbeddingRegistry>>,
     session: Option<Arc<lance::session::Session>>,
-    pushdown_operations: HashSet<PushdownOperation>,
+    pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
 }
 
 impl ConnectNamespaceBuilder {
@@ -895,6 +940,7 @@ impl ConnectNamespaceBuilder {
             ns_impl: ns_impl.to_string(),
             properties,
             storage_options: HashMap::new(),
+            namespace_client_properties: HashMap::new(),
             read_consistency_interval: None,
             embedding_registry: None,
             session: None,
@@ -904,7 +950,7 @@ impl ConnectNamespaceBuilder {
 
     /// Set an option for the storage layer.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_option(mut self, key: impl Into<String>, value: impl Into<String>) -> Self {
         self.storage_options.insert(key.into(), value.into());
         self
@@ -912,7 +958,7 @@ impl ConnectNamespaceBuilder {
 
     /// Set multiple options for the storage layer.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_options(
         mut self,
         pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
@@ -923,6 +969,29 @@ impl ConnectNamespaceBuilder {
         self
     }
 
+    /// Set an additional namespace client property.
+    pub fn namespace_client_property(
+        mut self,
+        key: impl Into<String>,
+        value: impl Into<String>,
+    ) -> Self {
+        self.namespace_client_properties
+            .insert(key.into(), value.into());
+        self
+    }
+
+    /// Set multiple additional namespace client properties.
+    pub fn namespace_client_properties(
+        mut self,
+        pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
+    ) -> Self {
+        for (key, value) in pairs {
+            self.namespace_client_properties
+                .insert(key.into(), value.into());
+        }
+        self
+    }
+
     /// The interval at which to check for updates from other processes.
     ///
     /// If left unset, consistency is not checked. For maximum read
@@ -960,11 +1029,11 @@ impl ConnectNamespaceBuilder {
     /// and leveraging server-side compute resources.
     ///
     /// Available operations:
-    /// - [`PushdownOperation::QueryTable`]: Execute queries via `namespace.query_table()`
+    /// - [`NamespaceClientPushdownOperation::QueryTable`]: Execute queries via `namespace.query_table()`
-    /// - [`PushdownOperation::CreateTable`]: Execute table creation via `namespace.create_table()`
+    /// - [`NamespaceClientPushdownOperation::CreateTable`]: Execute table creation via `namespace.create_table()`
     ///
     /// By default, no operations are pushed down (all executed locally).
-    pub fn pushdown_operation(mut self, operation: PushdownOperation) -> Self {
+    pub fn pushdown_operation(mut self, operation: NamespaceClientPushdownOperation) -> Self {
         self.pushdown_operations.insert(operation);
         self
     }
@@ -974,7 +1043,7 @@ impl ConnectNamespaceBuilder {
     /// See [`Self::pushdown_operation`] for details.
     pub fn pushdown_operations(
         mut self,
-        operations: impl IntoIterator<Item = PushdownOperation>,
+        operations: impl IntoIterator<Item = NamespaceClientPushdownOperation>,
     ) -> Self {
         self.pushdown_operations.extend(operations);
         self
@@ -984,10 +1053,13 @@ impl ConnectNamespaceBuilder {
     pub async fn execute(self) -> Result<Connection> {
         use crate::database::namespace::LanceNamespaceDatabase;
 
+        let mut properties = self.properties;
+        properties.extend(self.namespace_client_properties);
+
         let internal = Arc::new(
             LanceNamespaceDatabase::connect(
                 &self.ns_impl,
-                self.properties,
+                properties,
                 self.storage_options,
                 self.read_consistency_interval,
                 self.session,
@@ -1107,6 +1179,31 @@ mod tests {
         assert_eq!(db.uri(), relative_uri.to_str().unwrap().to_string());
     }
 
+    #[tokio::test]
+    async fn test_connect_with_namespace_client_properties() {
+        let tmp_dir = tempdir().unwrap();
+        let uri = tmp_dir.path().to_str().unwrap();
+
+        let db = connect(uri)
+            .namespace_client_property("table_version_tracking_enabled", "true")
+            .namespace_client_property("manifest_enabled", "true")
+            .execute()
+            .await
+            .unwrap();
+
+        let (ns_impl, properties) = db.namespace_client_config().await.unwrap();
+        assert_eq!(ns_impl, "dir");
+        assert_eq!(properties.get("root"), Some(&uri.to_string()));
+        assert_eq!(
+            properties.get("table_version_tracking_enabled"),
+            Some(&"true".to_string())
+        );
+        assert_eq!(
+            properties.get("manifest_enabled"),
+            Some(&"true".to_string())
+        );
+    }
+
     #[tokio::test]
     async fn test_table_names() {
         let tc = new_test_connection().await.unwrap();

rust/lancedb/src/connection/create_table.rs

@@ -55,7 +55,7 @@ impl CreateTableBuilder {
     /// Options already set on the connection will be inherited by the table,
     /// but can be overridden here.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_option(mut self, key: impl Into<String>, value: impl Into<String>) -> Self {
         let store_params = self
             .request
@@ -73,7 +73,7 @@ impl CreateTableBuilder {
     /// Options already set on the connection will be inherited by the table,
     /// but can be overridden here.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_options(
         mut self,
         pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,

rust/lancedb/src/database.rs

@@ -265,4 +265,13 @@ pub trait Database:
     /// For ListingDatabase, it is the equivalent DirectoryNamespace.
     /// For RemoteDatabase, it is the equivalent RestNamespace.
     async fn namespace_client(&self) -> Result<Arc<dyn LanceNamespace>>;
+
+    /// Get the configuration for constructing an equivalent namespace client.
+    /// Returns (impl_type, properties) where:
+    /// - impl_type: "dir" for DirectoryNamespace, "rest" for RestNamespace
+    /// - properties: configuration properties for the namespace
+    ///
+    /// This is useful for Python bindings where we want to return a Python
+    /// namespace object rather than a Rust trait object.
+    async fn namespace_client_config(&self) -> Result<(String, HashMap<String, String>)>;
 }

rust/lancedb/src/database/listing.rs

@@ -20,6 +20,7 @@ use snafu::ResultExt;
 
 use crate::connection::ConnectRequest;
 use crate::database::ReadConsistency;
+use crate::database::namespace::LanceNamespaceDatabase;
 use crate::error::{CreateDirSnafu, Error, Result};
 use crate::io::object_store::MirroringObjectStoreWrapper;
 use crate::table::NativeTable;
@@ -73,7 +74,7 @@ pub struct ListingDatabaseOptions {
     /// These are used to create/list tables and they are inherited by all tables
     /// opened by this database.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub storage_options: HashMap<String, String>,
 }
 
@@ -185,7 +186,7 @@ impl ListingDatabaseOptionsBuilder {
 
     /// Set an option for the storage layer.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_option(mut self, key: impl Into<String>, value: impl Into<String>) -> Self {
         self.options
             .storage_options
@@ -195,7 +196,7 @@ impl ListingDatabaseOptionsBuilder {
 
     /// Set multiple options for the storage layer.
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     pub fn storage_options(
         mut self,
         pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
@@ -255,6 +256,9 @@ pub struct ListingDatabase {
 
     // Session for object stores and caching
     session: Arc<lance::session::Session>,
+
+    // Namespace-backed database for child namespace operations
+    namespace_database: Arc<LanceNamespaceDatabase>,
 }
 
 impl std::fmt::Display for ListingDatabase {
@@ -281,6 +285,44 @@ const MIRRORED_STORE: &str = "mirroredStore";
 
 /// A connection to LanceDB
 impl ListingDatabase {
+    fn build_namespace_client_properties(
+        uri: &str,
+        storage_options: &HashMap<String, String>,
+        namespace_client_properties: HashMap<String, String>,
+    ) -> HashMap<String, String> {
+        let mut properties = namespace_client_properties;
+        properties.insert("root".to_string(), uri.to_string());
+        for (key, value) in storage_options {
+            properties.insert(format!("storage.{}", key), value.clone());
+        }
+        properties
+    }
+
+    async fn connect_namespace_database(
+        uri: &str,
+        storage_options: HashMap<String, String>,
+        namespace_client_properties: HashMap<String, String>,
+        read_consistency_interval: Option<std::time::Duration>,
+        session: Arc<lance::session::Session>,
+    ) -> Result<Arc<LanceNamespaceDatabase>> {
+        let ns_properties = Self::build_namespace_client_properties(
+            uri,
+            &storage_options,
+            namespace_client_properties,
+        );
+        Ok(Arc::new(
+            LanceNamespaceDatabase::connect(
+                "dir",
+                ns_properties,
+                storage_options,
+                read_consistency_interval,
+                Some(session),
+                HashSet::new(),
+            )
+            .await?,
+        ))
+    }
+
     /// Connect to a listing database
     ///
     /// The URI should be a path to a directory where the tables are stored.
@@ -300,6 +342,7 @@ impl ListingDatabase {
                     uri,
                     request.read_consistency_interval,
                     options.new_table_config,
+                    request.namespace_client_properties.clone(),
                     request.session.clone(),
                 )
                 .await
@@ -387,6 +430,15 @@ impl ListingDatabase {
                     None => None,
                 };
 
+                let namespace_database = Self::connect_namespace_database(
+                    &table_base_uri,
+                    options.storage_options.clone(),
+                    request.namespace_client_properties.clone(),
+                    request.read_consistency_interval,
+                    session.clone(),
+                )
+                .await?;
+
                 Ok(Self {
                     uri: table_base_uri,
                     query_string,
@@ -398,6 +450,7 @@ impl ListingDatabase {
                     storage_options_provider: None,
                     new_table_config: options.new_table_config,
                     session,
+                    namespace_database,
                 })
             }
             Err(_) => {
@@ -405,6 +458,7 @@ impl ListingDatabase {
                     uri,
                     request.read_consistency_interval,
                     options.new_table_config,
+                    request.namespace_client_properties.clone(),
                     request.session.clone(),
                 )
                 .await
@@ -416,6 +470,7 @@ impl ListingDatabase {
         path: &str,
         read_consistency_interval: Option<std::time::Duration>,
         new_table_config: NewTableConfig,
+        namespace_client_properties: HashMap<String, String>,
         session: Option<Arc<lance::session::Session>>,
     ) -> Result<Self> {
         let session = session.unwrap_or_else(|| Arc::new(lance::session::Session::default()));
@@ -429,6 +484,15 @@ impl ListingDatabase {
             Self::try_create_dir(path).context(CreateDirSnafu { path })?;
         }
 
+        let namespace_database = Self::connect_namespace_database(
+            path,
+            HashMap::new(),
+            namespace_client_properties,
+            read_consistency_interval,
+            session.clone(),
+        )
+        .await?;
+
         Ok(Self {
             uri: path.to_string(),
             query_string: None,
@@ -440,6 +504,7 @@ impl ListingDatabase {
             storage_options_provider: None,
             new_table_config,
             session,
+            namespace_database,
         })
     }
 
@@ -497,6 +562,10 @@ impl ListingDatabase {
         Ok(uri)
     }
 
+    fn namespace_database(&self) -> Arc<LanceNamespaceDatabase> {
+        self.namespace_database.clone()
+    }
+
     async fn drop_tables(&self, names: Vec<String>) -> Result<()> {
         let object_store_params = ObjectStoreParams {
             storage_options_accessor: if self.storage_options.is_empty() {
@@ -696,16 +765,7 @@ impl Database for ListingDatabase {
         &self,
         request: ListNamespacesRequest,
     ) -> Result<ListNamespacesResponse> {
-        if request.id.as_ref().map(|v| !v.is_empty()).unwrap_or(false) {
+        self.namespace_database().list_namespaces(request).await
-            return Err(Error::NotSupported {
-                message: "Namespace operations are not supported for listing database".into(),
-            });
-        }
-
-        Ok(ListNamespacesResponse {
-            namespaces: Vec::new(),
-            page_token: None,
-        })
     }
 
     fn uri(&self) -> &str {
@@ -726,36 +786,26 @@ impl Database for ListingDatabase {
 
     async fn create_namespace(
         &self,
-        _request: CreateNamespaceRequest,
+        request: CreateNamespaceRequest,
     ) -> Result<CreateNamespaceResponse> {
-        Err(Error::NotSupported {
+        self.namespace_database().create_namespace(request).await
-            message: "Namespace operations are not supported for listing database".into(),
-        })
     }
 
-    async fn drop_namespace(
+    async fn drop_namespace(&self, request: DropNamespaceRequest) -> Result<DropNamespaceResponse> {
-        &self,
+        self.namespace_database().drop_namespace(request).await
-        _request: DropNamespaceRequest,
-    ) -> Result<DropNamespaceResponse> {
-        Err(Error::NotSupported {
-            message: "Namespace operations are not supported for listing database".into(),
-        })
     }
 
     async fn describe_namespace(
         &self,
-        _request: DescribeNamespaceRequest,
+        request: DescribeNamespaceRequest,
     ) -> Result<DescribeNamespaceResponse> {
-        Err(Error::NotSupported {
+        self.namespace_database().describe_namespace(request).await
-            message: "Namespace operations are not supported for listing database".into(),
-        })
     }
 
+    #[allow(deprecated)]
     async fn table_names(&self, request: TableNamesRequest) -> Result<Vec<String>> {
         if !request.namespace_path.is_empty() {
-            return Err(Error::NotSupported {
+            return self.namespace_database().table_names(request).await;
-                message: "Namespace parameter is not supported for listing database. Only root namespace is supported.".into(),
-            });
         }
         let mut f = self
             .object_store
@@ -788,9 +838,7 @@ impl Database for ListingDatabase {
 
     async fn list_tables(&self, request: ListTablesRequest) -> Result<ListTablesResponse> {
         if request.id.as_ref().map(|v| !v.is_empty()).unwrap_or(false) {
-            return Err(Error::NotSupported {
+            return self.namespace_database().list_tables(request).await;
-                message: "Namespace parameter is not supported for listing database. Only root namespace is supported.".into(),
-            });
         }
         let mut f = self
             .object_store
@@ -838,11 +886,8 @@ impl Database for ListingDatabase {
     }
 
     async fn create_table(&self, request: CreateTableRequest) -> Result<Arc<dyn BaseTable>> {
-        // When namespace is not empty, location must be provided
+        if !request.namespace_path.is_empty() {
-        if !request.namespace_path.is_empty() && request.location.is_none() {
+            return self.namespace_database().create_table(request).await;
-            return Err(Error::InvalidInput {
-                message: "Location must be provided when namespace is not empty".into(),
-            });
         }
         // Use provided location if available, otherwise derive from table name
         let table_uri = request
@@ -959,11 +1004,8 @@ impl Database for ListingDatabase {
     }
 
     async fn open_table(&self, mut request: OpenTableRequest) -> Result<Arc<dyn BaseTable>> {
-        // When namespace is not empty, location must be provided
+        if !request.namespace_path.is_empty() {
-        if !request.namespace_path.is_empty() && request.location.is_none() {
+            return self.namespace_database().open_table(request).await;
-            return Err(Error::InvalidInput {
-                message: "Location must be provided when namespace is not empty".into(),
-            });
         }
         // Use provided location if available, otherwise derive from table name
         let table_uri = request
@@ -1059,9 +1101,10 @@ impl Database for ListingDatabase {
 
     async fn drop_table(&self, name: &str, namespace_path: &[String]) -> Result<()> {
         if !namespace_path.is_empty() {
-            return Err(Error::NotSupported {
+            return self
-                message: "Namespace parameter is not supported for listing database.".into(),
+                .namespace_database()
-            });
+                .drop_table(name, namespace_path)
+                .await;
         }
         self.drop_tables(vec![name.to_string()]).await
     }
@@ -1070,9 +1113,10 @@ impl Database for ListingDatabase {
     async fn drop_all_tables(&self, namespace_path: &[String]) -> Result<()> {
         // Check if namespace parameter is provided
         if !namespace_path.is_empty() {
-            return Err(Error::NotSupported {
+            return self
-                message: "Namespace parameter is not supported for listing database.".into(),
+                .namespace_database()
-            });
+                .drop_all_tables(namespace_path)
+                .await;
         }
         let tables = self.table_names(TableNamesRequest::default()).await?;
         self.drop_tables(tables).await
@@ -1083,21 +1127,11 @@ impl Database for ListingDatabase {
     }
 
     async fn namespace_client(&self) -> Result<Arc<dyn lance_namespace::LanceNamespace>> {
-        // Create a DirectoryNamespace pointing to the same root with the same storage options
+        self.namespace_database.namespace_client().await
-        let mut builder = lance_namespace_impls::DirectoryNamespaceBuilder::new(&self.uri);
+    }
 
-        // Add storage options
+    async fn namespace_client_config(&self) -> Result<(String, HashMap<String, String>)> {
-        if !self.storage_options.is_empty() {
+        self.namespace_database.namespace_client_config().await
-            builder = builder.storage_options(self.storage_options.clone());
-        }
-
-        // Use the same session
-        builder = builder.session(self.session.clone());
-
-        let namespace = builder.build().await.map_err(|e| Error::Runtime {
-            message: format!("Failed to create namespace client: {}", e),
-        })?;
-        Ok(Arc::new(namespace) as Arc<dyn lance_namespace::LanceNamespace>)
     }
 }
 
@@ -1123,6 +1157,7 @@ mod tests {
             #[cfg(feature = "remote")]
             client_config: Default::default(),
             options: Default::default(),
+            namespace_client_properties: Default::default(),
             read_consistency_interval: None,
             session: None,
         };
@@ -1256,6 +1291,7 @@ mod tests {
             #[cfg(feature = "remote")]
             client_config: Default::default(),
             options: options.clone(),
+            namespace_client_properties: Default::default(),
             read_consistency_interval: None,
             session: None,
         };
@@ -1790,6 +1826,7 @@ mod tests {
             #[cfg(feature = "remote")]
             client_config: Default::default(),
             options,
+            namespace_client_properties: Default::default(),
             read_consistency_interval: None,
             session: None,
         };
@@ -1895,6 +1932,7 @@ mod tests {
             #[cfg(feature = "remote")]
             client_config: Default::default(),
             options,
+            namespace_client_properties: Default::default(),
             read_consistency_interval: None,
             session: None,
         };
@@ -1966,6 +2004,7 @@ mod tests {
             #[cfg(feature = "remote")]
             client_config: Default::default(),
             options,
+            namespace_client_properties: Default::default(),
             read_consistency_interval: None,
             session: None,
         };
@@ -2099,4 +2138,209 @@ mod tests {
         assert!(tables.contains(&"table1".to_string()));
         assert!(tables.contains(&"table2".to_string()));
     }
+
+    #[tokio::test]
+    async fn test_listing_database_namespace_operations() {
+        let (_tempdir, db) = setup_database().await;
+
+        db.create_namespace(CreateNamespaceRequest {
+            id: Some(vec!["parent".to_string()]),
+            ..Default::default()
+        })
+        .await
+        .unwrap();
+
+        db.create_namespace(CreateNamespaceRequest {
+            id: Some(vec!["parent".to_string(), "child".to_string()]),
+            ..Default::default()
+        })
+        .await
+        .unwrap();
+
+        let root_namespaces = db
+            .list_namespaces(ListNamespacesRequest {
+                id: Some(vec![]),
+                ..Default::default()
+            })
+            .await
+            .unwrap();
+        assert!(root_namespaces.namespaces.contains(&"parent".to_string()));
+
+        let child_namespaces = db
+            .list_namespaces(ListNamespacesRequest {
+                id: Some(vec!["parent".to_string()]),
+                ..Default::default()
+            })
+            .await
+            .unwrap();
+        assert!(child_namespaces.namespaces.contains(&"child".to_string()));
+
+        db.describe_namespace(DescribeNamespaceRequest {
+            id: Some(vec!["parent".to_string(), "child".to_string()]),
+            ..Default::default()
+        })
+        .await
+        .unwrap();
+    }
+
+    #[tokio::test]
+    #[cfg(not(windows))] // TODO: support Windows once directory namespace-backed listing DB tests are supported.
+    async fn test_listing_database_with_namespace_client_properties() {
+        let tempdir = tempdir().unwrap();
+        let uri = tempdir.path().to_str().unwrap();
+
+        let mut namespace_client_properties = HashMap::new();
+        namespace_client_properties.insert(
+            "table_version_tracking_enabled".to_string(),
+            "true".to_string(),
+        );
+        namespace_client_properties.insert("manifest_enabled".to_string(), "true".to_string());
+
+        let request = ConnectRequest {
+            uri: uri.to_string(),
+            #[cfg(feature = "remote")]
+            client_config: Default::default(),
+            options: Default::default(),
+            namespace_client_properties,
+            read_consistency_interval: None,
+            session: None,
+        };
+
+        let db = ListingDatabase::connect_with_options(&request)
+            .await
+            .unwrap();
+        let namespace_path = vec!["test_ns".to_string()];
+
+        db.create_namespace(CreateNamespaceRequest {
+            id: Some(namespace_path.clone()),
+            ..Default::default()
+        })
+        .await
+        .unwrap();
+
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("name", DataType::Utf8, false),
+        ]));
+
+        db.create_table(CreateTableRequest {
+            name: "managed_table".to_string(),
+            namespace_path: namespace_path.clone(),
+            data: Box::new(RecordBatch::new_empty(schema)) as Box<dyn Scannable>,
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+            location: None,
+            namespace_client: None,
+        })
+        .await
+        .unwrap();
+
+        let namespace_client = db.namespace_client().await.unwrap();
+        let describe = namespace_client
+            .describe_table(lance_namespace::models::DescribeTableRequest {
+                id: Some(vec!["test_ns".to_string(), "managed_table".to_string()]),
+                ..Default::default()
+            })
+            .await
+            .unwrap();
+
+        assert_eq!(describe.managed_versioning, Some(true));
+    }
+
+    #[tokio::test]
+    async fn test_listing_database_nested_namespace_table_ops() {
+        let (_tempdir, db) = setup_database().await;
+        let namespace_path = vec!["parent".to_string(), "child".to_string()];
+
+        db.create_namespace(CreateNamespaceRequest {
+            id: Some(vec!["parent".to_string()]),
+            ..Default::default()
+        })
+        .await
+        .unwrap();
+        db.create_namespace(CreateNamespaceRequest {
+            id: Some(namespace_path.clone()),
+            ..Default::default()
+        })
+        .await
+        .unwrap();
+
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("name", DataType::Utf8, false),
+        ]));
+
+        db.create_table(CreateTableRequest {
+            name: "nested_table".to_string(),
+            namespace_path: namespace_path.clone(),
+            data: Box::new(RecordBatch::new_empty(schema)) as Box<dyn Scannable>,
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+            location: None,
+            namespace_client: None,
+        })
+        .await
+        .unwrap();
+
+        let namespace_client = db.namespace_client().await.unwrap();
+        let describe = namespace_client
+            .describe_table(lance_namespace::models::DescribeTableRequest {
+                id: Some(vec![
+                    "parent".to_string(),
+                    "child".to_string(),
+                    "nested_table".to_string(),
+                ]),
+                ..Default::default()
+            })
+            .await
+            .unwrap();
+        assert!(describe.location.is_some());
+
+        let table = db
+            .open_table(OpenTableRequest {
+                name: "nested_table".to_string(),
+                namespace_path: namespace_path.clone(),
+                index_cache_size: None,
+                lance_read_params: None,
+                location: None,
+                namespace_client: None,
+                managed_versioning: None,
+            })
+            .await
+            .unwrap();
+        assert_eq!(table.name(), "nested_table");
+
+        #[allow(deprecated)]
+        let table_names = db
+            .table_names(TableNamesRequest {
+                namespace_path: namespace_path.clone(),
+                start_after: None,
+                limit: None,
+            })
+            .await
+            .unwrap();
+        assert_eq!(table_names, vec!["nested_table".to_string()]);
+
+        let list_tables = db
+            .list_tables(ListTablesRequest {
+                id: Some(namespace_path.clone()),
+                ..Default::default()
+            })
+            .await
+            .unwrap();
+        assert_eq!(list_tables.tables, vec!["nested_table".to_string()]);
+
+        db.drop_table("nested_table", &namespace_path)
+            .await
+            .unwrap();
+
+        let post_drop = db
+            .list_tables(ListTablesRequest {
+                id: Some(namespace_path),
+                ..Default::default()
+            })
+            .await
+            .unwrap();
+        assert!(post_drop.tables.is_empty());
+    }
 }

rust/lancedb/src/database/namespace.rs

@@ -22,7 +22,7 @@ use lance_namespace_impls::ConnectBuilder;
 use lance_table::io::commit::CommitHandler;
 use lance_table::io::commit::external_manifest::ExternalManifestCommitHandler;
 
-use crate::connection::PushdownOperation;
+use crate::connection::NamespaceClientPushdownOperation;
 use crate::database::ReadConsistency;
 use crate::error::{Error, Result};
 use crate::table::NativeTable;
@@ -44,17 +44,42 @@ pub struct LanceNamespaceDatabase {
     // database URI
     uri: String,
     // Operations to push down to the namespace server
-    pushdown_operations: HashSet<PushdownOperation>,
+    pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
+    // Namespace implementation type (e.g., "dir", "rest")
+    ns_impl: String,
+    // Namespace properties used to construct the namespace client
+    ns_properties: HashMap<String, String>,
 }
 
 impl LanceNamespaceDatabase {
+    pub fn from_namespace_client(
+        namespace_client: Arc<dyn LanceNamespace>,
+        namespace_client_impl: String,
+        namespace_client_properties: HashMap<String, String>,
+        storage_options: HashMap<String, String>,
+        read_consistency_interval: Option<std::time::Duration>,
+        session: Option<Arc<lance::session::Session>>,
+        namespace_client_pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
+    ) -> Self {
+        Self {
+            namespace: namespace_client,
+            storage_options,
+            read_consistency_interval,
+            session,
+            uri: format!("namespace://{}", namespace_client_impl),
+            pushdown_operations: namespace_client_pushdown_operations,
+            ns_impl: namespace_client_impl,
+            ns_properties: namespace_client_properties,
+        }
+    }
+
     pub async fn connect(
         ns_impl: &str,
         ns_properties: HashMap<String, String>,
         storage_options: HashMap<String, String>,
         read_consistency_interval: Option<std::time::Duration>,
         session: Option<Arc<lance::session::Session>>,
-        pushdown_operations: HashSet<PushdownOperation>,
+        pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
     ) -> Result<Self> {
         let mut builder = ConnectBuilder::new(ns_impl);
         for (key, value) in ns_properties.clone() {
@@ -74,6 +99,8 @@ impl LanceNamespaceDatabase {
             session,
             uri: format!("namespace://{}", ns_impl),
             pushdown_operations,
+            ns_impl: ns_impl.to_string(),
+            ns_properties,
         })
     }
 }
@@ -157,22 +184,15 @@ impl Database for LanceNamespaceDatabase {
     async fn create_table(&self, request: DbCreateTableRequest) -> Result<Arc<dyn BaseTable>> {
         let mut table_id = request.namespace_path.clone();
         table_id.push(request.name.clone());
-        let describe_request = DescribeTableRequest {
-            id: Some(table_id.clone()),
-            ..Default::default()
-        };
-
-        let describe_result = self.namespace.describe_table(describe_request).await;
 
         match request.mode {
-            CreateTableMode::Create => {
+            CreateTableMode::Create => {}
-                if describe_result.is_ok() {
-                    return Err(Error::TableAlreadyExists {
-                        name: request.name.clone(),
-                    });
-                }
-            }
             CreateTableMode::Overwrite => {
+                let describe_request = DescribeTableRequest {
+                    id: Some(table_id.clone()),
+                    ..Default::default()
+                };
+                let describe_result = self.namespace.describe_table(describe_request).await;
                 if describe_result.is_ok() {
                     // Drop the existing table - must succeed
                     let drop_request = DropTableRequest {
@@ -188,6 +208,11 @@ impl Database for LanceNamespaceDatabase {
                 }
             }
             CreateTableMode::ExistOk(_) => {
+                let describe_request = DescribeTableRequest {
+                    id: Some(table_id.clone()),
+                    ..Default::default()
+                };
+                let describe_result = self.namespace.describe_table(describe_request).await;
                 if describe_result.is_ok() {
                     let native_table = NativeTable::open_from_namespace(
                         self.namespace.clone(),
@@ -215,7 +240,26 @@ impl Database for LanceNamespaceDatabase {
         };
 
         let (location, initial_storage_options, managed_versioning) = {
-            let response = self.namespace.declare_table(declare_request).await?;
+            let response = self
+                .namespace
+                .declare_table(declare_request)
+                .await
+                .map_err(|e| {
+                    let err_str = e.to_string();
+                    if matches!(request.mode, CreateTableMode::Create)
+                        && (err_str.contains("already exists")
+                            || err_str.contains("TableAlreadyExists")
+                            || err_str.contains("table already exists"))
+                    {
+                        Error::TableAlreadyExists {
+                            name: request.name.clone(),
+                        }
+                    } else {
+                        Error::Runtime {
+                            message: format!("Failed to declare table: {}", e),
+                        }
+                    }
+                })?;
             let loc = response.location.ok_or_else(|| Error::Runtime {
                 message: "Table location is missing from declare_table response".to_string(),
             })?;
@@ -345,6 +389,10 @@ impl Database for LanceNamespaceDatabase {
     async fn namespace_client(&self) -> Result<Arc<dyn LanceNamespace>> {
         Ok(self.namespace.clone())
     }
+
+    async fn namespace_client_config(&self) -> Result<(String, HashMap<String, String>)> {
+        Ok((self.ns_impl.clone(), self.ns_properties.clone()))
+    }
 }
 
 #[cfg(test)]
@@ -440,6 +488,47 @@ mod tests {
         ));
     }
 
+    #[tokio::test]
+    async fn test_namespace_connection_with_namespace_client_properties() {
+        let tmp_dir = tempdir().unwrap();
+        let root_path = tmp_dir.path().to_str().unwrap().to_string();
+
+        let mut properties = HashMap::new();
+        properties.insert("root".to_string(), root_path);
+
+        let conn = connect_namespace("dir", properties)
+            .namespace_client_property("table_version_tracking_enabled", "true")
+            .namespace_client_property("manifest_enabled", "true")
+            .execute()
+            .await
+            .expect("Failed to connect to namespace");
+
+        conn.create_namespace(CreateNamespaceRequest {
+            id: Some(vec!["test_ns".into()]),
+            ..Default::default()
+        })
+        .await
+        .expect("Failed to create namespace");
+
+        let test_data = create_test_data();
+        conn.create_table("test_table", test_data)
+            .namespace(vec!["test_ns".into()])
+            .execute()
+            .await
+            .expect("Failed to create table");
+
+        let namespace_client = conn.namespace_client().await.unwrap();
+        let describe = namespace_client
+            .describe_table(DescribeTableRequest {
+                id: Some(vec!["test_ns".into(), "test_table".into()]),
+                ..Default::default()
+            })
+            .await
+            .expect("Failed to describe table");
+
+        assert_eq!(describe.managed_versioning, Some(true));
+    }
+
     #[tokio::test]
     async fn test_namespace_create_table_basic() {
         // Setup: Create a temporary directory for the namespace

rust/lancedb/src/embeddings/bedrock.rs

@@ -177,6 +177,7 @@ impl BedrockEmbeddingFunction {
                         ))
                         .send()
                         .await
+                        .map_err(Box::new)
                 })
             })
             .unwrap();

rust/lancedb/src/lib.rs

@@ -69,7 +69,7 @@
 //! It treats [`FixedSizeList<Float16/Float32>`](https://docs.rs/arrow/latest/arrow/array/struct.FixedSizeListArray.html)
 //! columns as vector columns.
 //!
-//! For more details, please refer to the [LanceDB documentation](https://lancedb.com/docs).
+//! For more details, please refer to the [LanceDB documentation](https://docs.lancedb.com).
 //!
 //! #### Create a table
 //!

rust/lancedb/src/remote/client.rs

@@ -52,6 +52,13 @@ pub struct ClientConfig {
     pub tls_config: Option<TlsConfig>,
     /// Provider for custom headers to be added to each request
     pub header_provider: Option<Arc<dyn HeaderProvider>>,
+    /// User identifier for tracking purposes.
+    ///
+    /// This is sent as the `x-lancedb-user-id` header in requests to LanceDB Cloud/Enterprise.
+    /// It can be set directly, or via the `LANCEDB_USER_ID` environment variable.
+    /// Alternatively, set `LANCEDB_USER_ID_ENV_KEY` to specify another environment
+    /// variable that contains the user ID value.
+    pub user_id: Option<String>,
 }
 
 impl std::fmt::Debug for ClientConfig {
@@ -67,6 +74,7 @@ impl std::fmt::Debug for ClientConfig {
                 "header_provider",
                 &self.header_provider.as_ref().map(|_| "Some(...)"),
             )
+            .field("user_id", &self.user_id)
             .finish()
     }
 }
@@ -81,10 +89,41 @@ impl Default for ClientConfig {
             id_delimiter: None,
             tls_config: None,
             header_provider: None,
+            user_id: None,
         }
     }
 }
 
+impl ClientConfig {
+    /// Resolve the user ID from the config or environment variables.
+    ///
+    /// Resolution order:
+    /// 1. If `user_id` is set in the config, use that value
+    /// 2. If `LANCEDB_USER_ID` environment variable is set, use that value
+    /// 3. If `LANCEDB_USER_ID_ENV_KEY` is set, read the env var it points to
+    /// 4. Otherwise, return None
+    pub fn resolve_user_id(&self) -> Option<String> {
+        if self.user_id.is_some() {
+            return self.user_id.clone();
+        }
+
+        if let Ok(user_id) = std::env::var("LANCEDB_USER_ID")
+            && !user_id.is_empty()
+        {
+            return Some(user_id);
+        }
+
+        if let Ok(env_key) = std::env::var("LANCEDB_USER_ID_ENV_KEY")
+            && let Ok(user_id) = std::env::var(&env_key)
+            && !user_id.is_empty()
+        {
+            return Some(user_id);
+        }
+
+        None
+    }
+}
+
 /// How to handle timeouts for HTTP requests.
 #[derive(Clone, Default, Debug)]
 pub struct TimeoutConfig {
@@ -464,6 +503,15 @@ impl<S: HttpSend> RestfulLanceDbClient<S> {
             );
         }
 
+        if let Some(user_id) = config.resolve_user_id() {
+            headers.insert(
+                HeaderName::from_static("x-lancedb-user-id"),
+                HeaderValue::from_str(&user_id).map_err(|_| Error::InvalidInput {
+                    message: format!("non-ascii user_id '{}' provided", user_id),
+                })?,
+            );
+        }
+
         Ok(headers)
     }
 
@@ -1072,4 +1120,91 @@ mod tests {
             _ => panic!("Expected Runtime error"),
         }
     }
+
+    #[test]
+    fn test_resolve_user_id_direct_value() {
+        let config = ClientConfig {
+            user_id: Some("direct-user-id".to_string()),
+            ..Default::default()
+        };
+        assert_eq!(config.resolve_user_id(), Some("direct-user-id".to_string()));
+    }
+
+    #[test]
+    fn test_resolve_user_id_none() {
+        let config = ClientConfig::default();
+        // Clear env vars that might be set from other tests
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::remove_var("LANCEDB_USER_ID");
+            std::env::remove_var("LANCEDB_USER_ID_ENV_KEY");
+        }
+        assert_eq!(config.resolve_user_id(), None);
+    }
+
+    #[test]
+    fn test_resolve_user_id_from_env() {
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::set_var("LANCEDB_USER_ID", "env-user-id");
+        }
+        let config = ClientConfig::default();
+        assert_eq!(config.resolve_user_id(), Some("env-user-id".to_string()));
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::remove_var("LANCEDB_USER_ID");
+        }
+    }
+
+    #[test]
+    fn test_resolve_user_id_from_env_key() {
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::remove_var("LANCEDB_USER_ID");
+            std::env::set_var("LANCEDB_USER_ID_ENV_KEY", "MY_CUSTOM_USER_ID");
+            std::env::set_var("MY_CUSTOM_USER_ID", "custom-env-user-id");
+        }
+        let config = ClientConfig::default();
+        assert_eq!(
+            config.resolve_user_id(),
+            Some("custom-env-user-id".to_string())
+        );
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::remove_var("LANCEDB_USER_ID_ENV_KEY");
+            std::env::remove_var("MY_CUSTOM_USER_ID");
+        }
+    }
+
+    #[test]
+    fn test_resolve_user_id_direct_takes_precedence() {
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::set_var("LANCEDB_USER_ID", "env-user-id");
+        }
+        let config = ClientConfig {
+            user_id: Some("direct-user-id".to_string()),
+            ..Default::default()
+        };
+        assert_eq!(config.resolve_user_id(), Some("direct-user-id".to_string()));
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::remove_var("LANCEDB_USER_ID");
+        }
+    }
+
+    #[test]
+    fn test_resolve_user_id_empty_env_ignored() {
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::set_var("LANCEDB_USER_ID", "");
+            std::env::remove_var("LANCEDB_USER_ID_ENV_KEY");
+        }
+        let config = ClientConfig::default();
+        assert_eq!(config.resolve_user_id(), None);
+        // SAFETY: This is only called in tests
+        unsafe {
+            std::env::remove_var("LANCEDB_USER_ID");
+        }
+    }
 }

rust/lancedb/src/remote/db.rs

@@ -97,7 +97,7 @@ pub struct RemoteDatabaseOptions {
     pub host_override: Option<String>,
     /// Storage options configure the storage layer (e.g. S3, GCS, Azure, etc.)
     ///
-    /// See available options at <https://lancedb.com/docs/storage/>
+    /// See available options at <https://docs.lancedb.com/storage/>
     ///
     /// These options are only used for LanceDB Enterprise and only a subset of options
     /// are supported.
@@ -777,6 +777,32 @@ impl<S: HttpSend> Database for RemoteDatabase<S> {
         let namespace = builder.build();
         Ok(Arc::new(namespace) as Arc<dyn lance_namespace::LanceNamespace>)
     }
+
+    async fn namespace_client_config(&self) -> Result<(String, HashMap<String, String>)> {
+        let mut properties = HashMap::new();
+        properties.insert("uri".to_string(), self.client.host().to_string());
+        properties.insert("delimiter".to_string(), self.client.id_delimiter.clone());
+        for (key, value) in &self.namespace_headers {
+            properties.insert(format!("header.{}", key), value.clone());
+        }
+        // Add TLS configuration if present
+        if let Some(tls_config) = &self.tls_config {
+            if let Some(cert_file) = &tls_config.cert_file {
+                properties.insert("tls.cert_file".to_string(), cert_file.clone());
+            }
+            if let Some(key_file) = &tls_config.key_file {
+                properties.insert("tls.key_file".to_string(), key_file.clone());
+            }
+            if let Some(ssl_ca_cert) = &tls_config.ssl_ca_cert {
+                properties.insert("tls.ssl_ca_cert".to_string(), ssl_ca_cert.clone());
+            }
+            properties.insert(
+                "tls.assert_hostname".to_string(),
+                tls_config.assert_hostname.to_string(),
+            );
+        }
+        Ok(("rest".to_string(), properties))
+    }
 }
 
 /// RemoteOptions contains a subset of StorageOptions that are compatible with Remote LanceDB connections

rust/lancedb/src/table.rs

@@ -47,7 +47,7 @@ use std::format;
 use std::path::Path;
 use std::sync::Arc;
 
-use crate::connection::PushdownOperation;
+use crate::connection::NamespaceClientPushdownOperation;
 
 use crate::data::scannable::{PeekedScannable, Scannable, estimate_write_partitions};
 use crate::database::Database;
@@ -1272,7 +1272,7 @@ pub struct NativeTable {
     pub(crate) namespace_client: Option<Arc<dyn LanceNamespace>>,
     // Operations to push down to the namespace server.
     // pub(crate) so query.rs can access the field for server-side query execution.
-    pub(crate) pushdown_operations: HashSet<PushdownOperation>,
+    pub(crate) pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
 }
 
 impl std::fmt::Debug for NativeTable {
@@ -1359,7 +1359,7 @@ impl NativeTable {
         params: Option<ReadParams>,
         read_consistency_interval: Option<std::time::Duration>,
         namespace_client: Option<Arc<dyn LanceNamespace>>,
-        pushdown_operations: HashSet<PushdownOperation>,
+        pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
         managed_versioning: Option<bool>,
     ) -> Result<Self> {
         let params = params.unwrap_or_default();
@@ -1470,7 +1470,7 @@ impl NativeTable {
         write_store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
         params: Option<ReadParams>,
         read_consistency_interval: Option<std::time::Duration>,
-        pushdown_operations: HashSet<PushdownOperation>,
+        pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
         session: Option<Arc<lance::session::Session>>,
     ) -> Result<Self> {
         let mut params = params.unwrap_or_default();
@@ -1518,7 +1518,7 @@ impl NativeTable {
         let id = Self::build_id(&namespace, name);
 
         let stored_namespace_client =
-            if pushdown_operations.contains(&PushdownOperation::QueryTable) {
+            if pushdown_operations.contains(&NamespaceClientPushdownOperation::QueryTable) {
                 Some(namespace_client)
             } else {
                 None
@@ -1588,7 +1588,7 @@ impl NativeTable {
         params: Option<WriteParams>,
         read_consistency_interval: Option<std::time::Duration>,
         namespace_client: Option<Arc<dyn LanceNamespace>>,
-        pushdown_operations: HashSet<PushdownOperation>,
+        pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
     ) -> Result<Self> {
         // Default params uses format v1.
         let params = params.unwrap_or(WriteParams {
@@ -1635,7 +1635,7 @@ impl NativeTable {
         params: Option<WriteParams>,
         read_consistency_interval: Option<std::time::Duration>,
         namespace_client: Option<Arc<dyn LanceNamespace>>,
-        pushdown_operations: HashSet<PushdownOperation>,
+        pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
     ) -> Result<Self> {
         let data: Box<dyn Scannable> = Box::new(RecordBatch::new_empty(schema));
         Self::create(
@@ -1685,7 +1685,7 @@ impl NativeTable {
         write_store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
         params: Option<WriteParams>,
         read_consistency_interval: Option<std::time::Duration>,
-        pushdown_operations: HashSet<PushdownOperation>,
+        pushdown_operations: HashSet<NamespaceClientPushdownOperation>,
         session: Option<Arc<lance::session::Session>>,
     ) -> Result<Self> {
         // Build table_id from namespace + name for the storage options provider
@@ -1738,7 +1738,7 @@ impl NativeTable {
         let id = Self::build_id(&namespace, name);
 
         let stored_namespace_client =
-            if pushdown_operations.contains(&PushdownOperation::QueryTable) {
+            if pushdown_operations.contains(&NamespaceClientPushdownOperation::QueryTable) {
                 Some(namespace_client)
             } else {
                 None

rust/lancedb/src/table/dataset.rs

@@ -285,7 +285,10 @@ mod tests {
 
     use super::*;
 
-    use crate::{connect, io::object_store::io_tracking::IoStatsHolder, table::WriteOptions};
+    use crate::{
+        connect, io::object_store::io_tracking::IoStatsHolder, table::WriteOptions,
+        utils::background_cache::clock,
+    };
 
     async fn create_test_dataset(uri: &str) -> Dataset {
         let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
@@ -463,6 +466,10 @@ mod tests {
         let uri = dir.path().to_str().unwrap();
         let ds = create_test_dataset(uri).await;
 
+        // Other tests use a thread-local mock clock. Simulate leaked state from a
+        // previous test to ensure this wrapper starts from real time.
+        clock::advance_by(Duration::from_secs(60));
+
         let wrapper = DatasetConsistencyWrapper::new_latest(ds, Some(Duration::from_millis(200)));
 
         // Populate the cache

rust/lancedb/src/table/query.rs

@@ -4,7 +4,7 @@
 use std::sync::Arc;
 
 use super::NativeTable;
-use crate::connection::PushdownOperation;
+use crate::connection::NamespaceClientPushdownOperation;
 use crate::error::{Error, Result};
 use crate::expr::expr_to_sql_string;
 use crate::query::{
@@ -44,7 +44,7 @@ pub async fn execute_query(
     // If QueryTable pushdown is enabled and namespace client is configured, use server-side query execution
     if table
         .pushdown_operations
-        .contains(&PushdownOperation::QueryTable)
+        .contains(&NamespaceClientPushdownOperation::QueryTable)
         && let Some(ref namespace_client) = table.namespace_client
     {
         return execute_namespace_query(table, namespace_client.clone(), query, options).await;

rust/lancedb/src/utils/background_cache.rs

@@ -107,6 +107,14 @@ where
             refresh_window < ttl,
             "refresh_window ({refresh_window:?}) must be less than ttl ({ttl:?})"
         );
+        #[cfg(test)]
+        {
+            // Tests may advance the thread-local mock clock and leave it behind for
+            // the next test that happens to run on the same worker thread. Each new
+            // cache should start from a clean clock state instead of inheriting
+            // unrelated mock time from a previous test.
+            clock::clear_mock();
+        }
         Self {
             inner: Arc::new(Mutex::new(CacheInner {
                 state: State::Empty,