taladb 0.2.1 → 0.2.3

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "taladb",
-  "version": "0.2.1",
+  "version": "0.2.3",
   "description": "Local-first document and vector database for React, React Native, and Node.js",
   "main": "dist/index.js",
   "module": "dist/index.mjs",
@@ -48,11 +48,11 @@
     "vitest": "^3.2.0"
   },
   "optionalDependencies": {
-    "@taladb/web": "0.2.1",
-    "@taladb/node": "0.2.1"
+    "@taladb/web": "0.2.3",
+    "@taladb/node": "0.2.3"
   },
   "scripts": {
-    "build": "tsup src/index.ts --format cjs,esm --dts --out-dir dist --external @taladb/web --external @taladb/node --external @taladb/react-native",
+    "build": "tsup",
     "typecheck": "tsc --noEmit",
     "test": "vitest run",
     "test:watch": "vitest"

package/dist/index.d.ts

@@ -1,139 +0,0 @@
-/** Similarity metric used for vector search. */
-type VectorMetric = 'cosine' | 'dot' | 'euclidean';
-interface VectorIndexOptions {
-    /** Number of dimensions in each stored vector. Enforced on insert and search. */
-    dimensions: number;
-    /** Similarity metric. Defaults to `"cosine"`. */
-    metric?: VectorMetric;
-}
-/** A single result returned by `Collection.findNearest`. */
-interface VectorSearchResult<T extends Document = Document> {
-    /** The matched document. */
-    document: T;
-    /**
-     * Similarity score — higher means more similar.
-     * Range depends on metric: cosine ∈ [-1,1], dot ∈ ℝ, euclidean ∈ (0,1].
-     */
-    score: number;
-}
-type Value = null | boolean | number | string | Uint8Array | Value[] | {
-    [key: string]: Value;
-};
-type Document = {
-    _id?: string;
-    [key: string]: Value | undefined;
-};
-type FieldOps<T> = T extends null | undefined ? {
-    $exists?: boolean;
-} : {
-    $eq?: T;
-    $ne?: T;
-    $gt?: T;
-    $gte?: T;
-    $lt?: T;
-    $lte?: T;
-    $in?: T[];
-    $nin?: T[];
-    $exists?: boolean;
-    /** Full-text search: matches documents where this string field contains the given token. */
-    $contains?: string;
-};
-type Filter<T extends Document = Document> = {
-    [K in keyof T]?: T[K] | FieldOps<T[K]>;
-} & {
-    $and?: Filter<T>[];
-    $or?: Filter<T>[];
-    $not?: Filter<T>;
-};
-type Update<T extends Document = Document> = {
-    $set?: Partial<T>;
-    $unset?: {
-        [K in keyof T]?: true;
-    };
-    $inc?: {
-        [K in keyof T]?: number;
-    };
-    $push?: {
-        [K in keyof T]?: Value;
-    };
-    $pull?: {
-        [K in keyof T]?: Value;
-    };
-};
-interface Collection<T extends Document = Document> {
-    insert(doc: Omit<T, '_id'>): Promise<string>;
-    insertMany(docs: Omit<T, '_id'>[]): Promise<string[]>;
-    find(filter?: Filter<T>): Promise<T[]>;
-    findOne(filter: Filter<T>): Promise<T | null>;
-    updateOne(filter: Filter<T>, update: Update<T>): Promise<boolean>;
-    updateMany(filter: Filter<T>, update: Update<T>): Promise<number>;
-    deleteOne(filter: Filter<T>): Promise<boolean>;
-    deleteMany(filter: Filter<T>): Promise<number>;
-    count(filter?: Filter<T>): Promise<number>;
-    createIndex(field: keyof Omit<T, '_id'> & string): Promise<void>;
-    dropIndex(field: keyof Omit<T, '_id'> & string): Promise<void>;
-    /**
-     * Create a vector index on a numeric-array field.
-     *
-     * After creation, `findNearest` can search this field and new inserts/updates
-     * automatically maintain the index. Existing documents are backfilled.
-     *
-     * @example
-     * await articles.createVectorIndex('embedding', { dimensions: 384 });
-     */
-    createVectorIndex(field: keyof Omit<T, '_id'> & string, options: VectorIndexOptions): Promise<void>;
-    /** Drop a vector index. */
-    dropVectorIndex(field: keyof Omit<T, '_id'> & string): Promise<void>;
-    /**
-     * Find the `topK` most similar documents to `vector` using a vector index.
-     *
-     * Optionally combine with a metadata `filter` to restrict the search space
-     * before ranking — e.g. find the 5 most similar english-language articles.
-     *
-     * Results are ordered by descending similarity score (highest first).
-     *
-     * @example
-     * const results = await articles.findNearest('embedding', queryVec, 5);
-     * // results: Array<{ document: Article, score: number }>
-     *
-     * @example with pre-filter
-     * const results = await articles.findNearest('embedding', queryVec, 5, {
-     *   locale: 'en',
-     * });
-     */
-    findNearest(field: keyof Omit<T, '_id'> & string, vector: number[], topK: number, filter?: Filter<T>): Promise<VectorSearchResult<T>[]>;
-    /**
-     * Subscribe to live query results. The callback receives a full snapshot of
-     * matching documents immediately and again after every write that could
-     * affect the result set.
-     *
-     * @returns An unsubscribe function. Call it to stop receiving updates.
-     *
-     * @example
-     * const unsub = users.subscribe({ active: true }, (docs) => {
-     *   console.log('active users:', docs);
-     * });
-     * // later…
-     * unsub();
-     */
-    subscribe(filter: Filter<T>, callback: (docs: T[]) => void): () => void;
-}
-interface TalaDB {
-    collection<T extends Document = Document>(name: string): Collection<T>;
-    close(): Promise<void>;
-}
-
-/**
- * Open a TalaDB database.
- *
- * @param dbName  Name of the database file (used for OPFS and native file paths).
- *                Ignored for in-memory databases.
- *
- * @example
- * const db = await openDB('myapp.db');
- * const users = db.collection<User>('users');
- * const id = await users.insert({ name: 'Alice', age: 30 });
- */
-declare function openDB(dbName?: string): Promise<TalaDB>;
-
-export { type Collection, type Document, type Filter, type TalaDB, type Update, type Value, type VectorIndexOptions, type VectorMetric, type VectorSearchResult, openDB };