feat(fts): support custom stop-word lists (#3734)

## What

Expose custom FTS stop-word lists in the Python and TypeScript public
APIs, including their standalone tokenize helpers and remote index
creation.

This PR supports concrete string lists only. It does not add file or
LanceDB-table stop-word sources.

## Why

Rust already exposes Lance's custom stop-word list option. The Python
and TypeScript APIs did not pass it through, and local index details did
not retain the full tokenizer parameters needed by index-backed
tokenization after reopening a table.

## How

- Add `custom_stop_words` / `customStopWords` to the Python and
TypeScript FTS and tokenize options.
- Preserve `None` / `undefined`, empty lists, and list contents without
normalization.
- Load the persisted FTS segment parameters when returning local index
details.
- Serialize the concrete list in remote create-index requests.
- Keep Python and TypeScript tests thin; behavior, persistence, query
tokenization, and remote JSON coverage live primarily in Rust.

## Validation

- `cargo check --quiet --features remote --tests --examples`
- `cargo clippy --quiet --features remote --tests --examples`
- `cargo test --quiet --features remote --tests`
- Python extension rebuild with `uv` and `maturin`
- Targeted Python tests: 4 passed
- Python `ruff format --check` and `ruff check`
- TypeScript build, typecheck, Biome lint, generated docs, and targeted
tests

---------

Co-authored-by: Yang Cen <yangcen@Yangs-Mac-mini.local>

nodejs/__test__/remote.test.ts

@@ -226,7 +226,7 @@ describe("remote connection", () => {
     );
   });
 
-  it("sends the FTS posting block size to remote tables", async () => {
+  it("sends FTS options to remote tables", async () => {
     let createIndexBody: Record<string, unknown> | undefined;
 
     await withMockDatabase(
@@ -264,7 +264,11 @@ describe("remote connection", () => {
       async (db) => {
         const table = await db.openTable("t");
         await table.createIndex("text", {
-          config: Index.fts({ blockSize: 256 }),
+          config: Index.fts({
+            blockSize: 256,
+            removeStopWords: true,
+            customStopWords: ["the"],
+          }),
         });
       },
     );
@@ -272,6 +276,7 @@ describe("remote connection", () => {
     expect(createIndexBody?.["column"]).toBe("text");
     expect(createIndexBody?.["index_type"]).toBe("FTS");
     expect(createIndexBody?.["block_size"]).toBe(256);
+    expect(createIndexBody?.["custom_stop_words"]).toEqual(["the"]);
   });
 
   it("diffs and merges remote branches", async () => {

nodejs/__test__/table.test.ts

@@ -2769,6 +2769,15 @@ describe.each([arrow15, arrow16, arrow17, arrow18])(
   },
 );
 
+test("tokenize supports custom stop words", async () => {
+  const tokens = await tokenize("the lance data", {
+    stem: false,
+    removeStopWords: true,
+    customStopWords: ["lance"],
+  });
+  expect(tokens.map((token) => token.text)).toEqual(["the", "data"]);
+});
+
 describe("when calling explainPlan", () => {
   let tmpDir: tmp.DirResult;
   let table: Table;

nodejs/examples/full_text_search.test.ts

@@ -29,8 +29,14 @@ test("full text search", async () => {
     const tbl = await db.createTable("myVectors", data, { mode: "overwrite" });
 
     await tbl.createIndex("doc", {
-      config: lancedb.Index.fts(),
+      config: lancedb.Index.fts({
+        stem: false,
+        removeStopWords: true,
+        customStopWords: ["banana"],
+      }),
     });
+    const tokens = await tbl.tokenize("apple banana", { column: "doc" });
+    expect(tokens.map((token) => token.text)).toEqual(["apple"]);
 
     // --8<-- [start:full_text_search]
     const result = await tbl

nodejs/lancedb/index.ts

@@ -194,6 +194,16 @@ export interface TokenizeOptions {
   /** Whether to remove stop words. */
   removeStopWords?: boolean;
 
+  /**
+   * Custom stop words that replace the built-in list for `language`.
+   *
+   * This option only affects tokenization when `removeStopWords` is true.
+   *
+   * `undefined` keeps the built-in language list. An empty array explicitly
+   * replaces it with no stop words.
+   */
+  customStopWords?: string[];
+
   /** Whether to fold ASCII characters. */
   asciiFolding?: boolean;
 
@@ -225,6 +235,7 @@ export async function tokenize(
     options?.lowercase,
     options?.stem,
     options?.removeStopWords,
+    options?.customStopWords,
     options?.asciiFolding,
     options?.ngramMinLength,
     options?.ngramMaxLength,

nodejs/lancedb/indices.ts

@@ -553,6 +553,16 @@ export interface FtsOptions {
    */
   removeStopWords?: boolean;
 
+  /**
+   * Custom stop words that replace the built-in list for `language`.
+   *
+   * This option only affects tokenization when `removeStopWords` is true.
+   *
+   * `undefined` keeps the built-in language list. An empty array explicitly
+   * replaces it with no stop words.
+   */
+  customStopWords?: string[];
+
   /**
    * whether to remove punctuation
    */
@@ -755,6 +765,7 @@ export class Index {
         options?.lowercase,
         options?.stem,
         options?.removeStopWords,
+        options?.customStopWords,
         options?.asciiFolding,
         options?.ngramMinLength,
         options?.ngramMaxLength,

nodejs/src/index.rs

@@ -43,6 +43,7 @@ pub fn tokenize(
     lower_case: Option<bool>,
     stem: Option<bool>,
     remove_stop_words: Option<bool>,
+    custom_stop_words: Option<Vec<String>>,
     ascii_folding: Option<bool>,
     ngram_min_length: Option<u32>,
     ngram_max_length: Option<u32>,
@@ -72,6 +73,7 @@ pub fn tokenize(
     if let Some(remove_stop_words) = remove_stop_words {
         opts = opts.remove_stop_words(remove_stop_words);
     }
+    opts = opts.custom_stop_words(custom_stop_words);
     if let Some(ascii_folding) = ascii_folding {
         opts = opts.ascii_folding(ascii_folding);
     }
@@ -222,6 +224,7 @@ impl Index {
         lower_case: Option<bool>,
         stem: Option<bool>,
         remove_stop_words: Option<bool>,
+        custom_stop_words: Option<Vec<String>>,
         ascii_folding: Option<bool>,
         ngram_min_length: Option<u32>,
         ngram_max_length: Option<u32>,
@@ -250,6 +253,7 @@ impl Index {
         if let Some(remove_stop_words) = remove_stop_words {
             opts = opts.remove_stop_words(remove_stop_words);
         }
+        opts = opts.custom_stop_words(custom_stop_words);
         if let Some(ascii_folding) = ascii_folding {
             opts = opts.ascii_folding(ascii_folding);
         }