@dakera-ai/dakera 0.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/LICENSE ADDED
@@ -0,0 +1,21 @@
1
+ MIT License
2
+
3
+ Copyright (c) 2025 Dakera AI
4
+
5
+ Permission is hereby granted, free of charge, to any person obtaining a copy
6
+ of this software and associated documentation files (the "Software"), to deal
7
+ in the Software without restriction, including without limitation the rights
8
+ to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
9
+ copies of the Software, and to permit persons to whom the Software is
10
+ furnished to do so, subject to the following conditions:
11
+
12
+ The above copyright notice and this permission notice shall be included in all
13
+ copies or substantial portions of the Software.
14
+
15
+ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16
+ IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17
+ FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18
+ AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19
+ LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20
+ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21
+ SOFTWARE.
package/README.md ADDED
@@ -0,0 +1,313 @@
1
+ # Dakera TypeScript SDK
2
+
3
+ [![CI](https://github.com/dakera-ai/dakera-js/actions/workflows/ci.yml/badge.svg)](https://github.com/dakera-ai/dakera-js/actions/workflows/ci.yml)
4
+ [![npm](https://img.shields.io/npm/v/dakera)](https://www.npmjs.com/package/dakera)
5
+ [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)
6
+ [![TypeScript](https://img.shields.io/badge/TypeScript-5.0+-blue)](https://www.typescriptlang.org/)
7
+
8
+ Official TypeScript/JavaScript client for [Dakera](https://github.com/dakera/dakera) - a high-performance vector database.
9
+
10
+ ## Installation
11
+
12
+ ```bash
13
+ npm install dakera
14
+ # or
15
+ yarn add dakera
16
+ # or
17
+ pnpm add dakera
18
+ ```
19
+
20
+ ## Quick Start
21
+
22
+ ```typescript
23
+ import { DakeraClient } from 'dakera';
24
+
25
+ // Connect to Dakera
26
+ const client = new DakeraClient('http://localhost:3000');
27
+
28
+ // Upsert vectors
29
+ await client.upsert('my-namespace', [
30
+ { id: 'vec1', values: [0.1, 0.2, 0.3], metadata: { label: 'a' } },
31
+ { id: 'vec2', values: [0.4, 0.5, 0.6], metadata: { label: 'b' } },
32
+ ]);
33
+
34
+ // Query similar vectors
35
+ const results = await client.query('my-namespace', [0.1, 0.2, 0.3], {
36
+ topK: 10,
37
+ });
38
+
39
+ for (const result of results.results) {
40
+ console.log(`${result.id}: ${result.score}`);
41
+ }
42
+ ```
43
+
44
+ ## Features
45
+
46
+ - **Full TypeScript Support**: Complete type definitions for all operations
47
+ - **Vector Operations**: Upsert, query, delete, fetch vectors
48
+ - **Full-Text Search**: Index documents and perform BM25 search
49
+ - **Hybrid Search**: Combine vector and text search with configurable weights
50
+ - **Namespace Management**: Create, list, delete namespaces
51
+ - **Metadata Filtering**: Filter queries by metadata fields
52
+ - **Automatic Retries**: Built-in retry logic with exponential backoff
53
+ - **Error Handling**: Typed exceptions for different error scenarios
54
+
55
+ ## Usage Examples
56
+
57
+ ### Vector Operations
58
+
59
+ ```typescript
60
+ import { DakeraClient } from 'dakera';
61
+
62
+ const client = new DakeraClient({
63
+ baseUrl: 'http://localhost:3000',
64
+ apiKey: 'your-api-key', // optional
65
+ timeout: 30000,
66
+ });
67
+
68
+ // Upsert vectors
69
+ await client.upsert('my-namespace', [
70
+ { id: 'vec1', values: [0.1, 0.2, 0.3], metadata: { category: 'A' } },
71
+ { id: 'vec2', values: [0.4, 0.5, 0.6], metadata: { category: 'B' } },
72
+ ]);
73
+
74
+ // Query with metadata filter
75
+ const results = await client.query('my-namespace', [0.1, 0.2, 0.3], {
76
+ topK: 5,
77
+ filter: { category: { $eq: 'A' } },
78
+ includeMetadata: true,
79
+ });
80
+
81
+ // Batch query
82
+ const batchResults = await client.batchQuery('my-namespace', [
83
+ { vector: [0.1, 0.2, 0.3], topK: 5 },
84
+ { vector: [0.4, 0.5, 0.6], topK: 3 },
85
+ ]);
86
+
87
+ // Fetch vectors by ID
88
+ const vectors = await client.fetch('my-namespace', ['vec1', 'vec2']);
89
+
90
+ // Delete vectors
91
+ await client.delete('my-namespace', { ids: ['vec1', 'vec2'] });
92
+ await client.delete('my-namespace', { filter: { category: { $eq: 'obsolete' } } });
93
+ ```
94
+
95
+ ### Full-Text Search
96
+
97
+ ```typescript
98
+ // Index documents
99
+ await client.indexDocuments('my-namespace', [
100
+ { id: 'doc1', content: 'Machine learning is transforming industries' },
101
+ { id: 'doc2', content: 'Vector databases enable semantic search' },
102
+ ]);
103
+
104
+ // Search
105
+ const results = await client.fulltextSearch('my-namespace', 'machine learning', {
106
+ topK: 10,
107
+ });
108
+
109
+ for (const result of results) {
110
+ console.log(`${result.id}: ${result.score}`);
111
+ }
112
+ ```
113
+
114
+ ### Hybrid Search
115
+
116
+ ```typescript
117
+ // Combine vector and text search
118
+ const results = await client.hybridSearch(
119
+ 'my-namespace',
120
+ [0.1, 0.2, 0.3], // Query vector
121
+ 'machine learning', // Text query
122
+ {
123
+ topK: 10,
124
+ alpha: 0.7, // 0 = pure vector, 1 = pure text
125
+ }
126
+ );
127
+
128
+ for (const result of results) {
129
+ console.log(`${result.id}: score=${result.score}, vector=${result.vectorScore}, text=${result.textScore}`);
130
+ }
131
+ ```
132
+
133
+ ### Namespace Management
134
+
135
+ ```typescript
136
+ // Create namespace
137
+ await client.createNamespace('embeddings', {
138
+ dimensions: 384,
139
+ indexType: 'hnsw',
140
+ });
141
+
142
+ // List namespaces
143
+ const namespaces = await client.listNamespaces();
144
+ for (const ns of namespaces) {
145
+ console.log(`${ns.name}: ${ns.vectorCount} vectors`);
146
+ }
147
+
148
+ // Get namespace info
149
+ const info = await client.getNamespace('embeddings');
150
+ console.log(`Dimensions: ${info.dimensions}, Index: ${info.indexType}`);
151
+
152
+ // Delete namespace
153
+ await client.deleteNamespace('old-namespace');
154
+ ```
155
+
156
+ ### Metadata Filtering
157
+
158
+ Dakera supports rich metadata filtering:
159
+
160
+ ```typescript
161
+ // Equality
162
+ const filter1 = { status: { $eq: 'active' } };
163
+
164
+ // Comparison
165
+ const filter2 = { price: { $gt: 100, $lt: 500 } };
166
+
167
+ // In list
168
+ const filter3 = { category: { $in: ['electronics', 'books'] } };
169
+
170
+ // Logical operators
171
+ const filter4 = {
172
+ $and: [
173
+ { status: { $eq: 'active' } },
174
+ { price: { $lt: 1000 } },
175
+ ],
176
+ };
177
+
178
+ const results = await client.query('products', queryVector, {
179
+ filter: filter4,
180
+ topK: 20,
181
+ });
182
+ ```
183
+
184
+ ### Error Handling
185
+
186
+ ```typescript
187
+ import {
188
+ DakeraClient,
189
+ NotFoundError,
190
+ ValidationError,
191
+ RateLimitError,
192
+ ServerError,
193
+ } from 'dakera';
194
+
195
+ const client = new DakeraClient('http://localhost:3000');
196
+
197
+ try {
198
+ const results = await client.query('nonexistent', [0.1, 0.2]);
199
+ } catch (error) {
200
+ if (error instanceof NotFoundError) {
201
+ console.log(`Namespace not found: ${error.message}`);
202
+ } else if (error instanceof ValidationError) {
203
+ console.log(`Invalid request: ${error.message}`);
204
+ } else if (error instanceof RateLimitError) {
205
+ console.log(`Rate limited, retry after ${error.retryAfter} seconds`);
206
+ } else if (error instanceof ServerError) {
207
+ console.log(`Server error: ${error.message}`);
208
+ }
209
+ }
210
+ ```
211
+
212
+ ## Configuration
213
+
214
+ | Option | Type | Default | Description |
215
+ |--------|------|---------|-------------|
216
+ | `baseUrl` | string | required | Dakera server URL |
217
+ | `apiKey` | string | undefined | API key for authentication |
218
+ | `timeout` | number | 30000 | Request timeout in milliseconds |
219
+ | `maxRetries` | number | 3 | Max retries for failed requests |
220
+ | `headers` | object | undefined | Additional HTTP headers |
221
+
222
+ ## API Reference
223
+
224
+ ### DakeraClient
225
+
226
+ #### Vector Operations
227
+ - `upsert(namespace, vectors)` - Insert or update vectors
228
+ - `query(namespace, vector, options?)` - Query similar vectors
229
+ - `delete(namespace, options)` - Delete vectors
230
+ - `fetch(namespace, ids, options?)` - Fetch vectors by ID
231
+ - `batchQuery(namespace, queries)` - Execute multiple queries
232
+
233
+ #### Full-Text Operations
234
+ - `indexDocuments(namespace, documents)` - Index documents
235
+ - `fulltextSearch(namespace, query, options?)` - Text search
236
+ - `hybridSearch(namespace, vector, query, options?)` - Hybrid search
237
+
238
+ #### Namespace Operations
239
+ - `listNamespaces()` - List all namespaces
240
+ - `getNamespace(namespace)` - Get namespace info
241
+ - `createNamespace(namespace, options?)` - Create namespace
242
+ - `deleteNamespace(namespace)` - Delete namespace
243
+
244
+ #### Admin Operations
245
+ - `health()` - Check server health
246
+ - `getIndexStats(namespace)` - Get index statistics
247
+ - `compact(namespace)` - Trigger compaction
248
+ - `flush(namespace)` - Flush pending writes
249
+
250
+ ## TypeScript Types
251
+
252
+ All types are exported for use in your application:
253
+
254
+ ```typescript
255
+ import type {
256
+ Vector,
257
+ VectorInput,
258
+ QueryResult,
259
+ SearchResult,
260
+ NamespaceInfo,
261
+ IndexStats,
262
+ Document,
263
+ FilterExpression,
264
+ QueryOptions,
265
+ ClientOptions,
266
+ } from 'dakera';
267
+ ```
268
+
269
+ ## Browser Support
270
+
271
+ This SDK uses the Fetch API and works in:
272
+ - Node.js 18+
273
+ - Modern browsers (Chrome, Firefox, Safari, Edge)
274
+ - Deno
275
+ - Bun
276
+
277
+ ## Development
278
+
279
+ ```bash
280
+ # Install dependencies
281
+ npm install
282
+
283
+ # Build
284
+ npm run build
285
+
286
+ # Run tests
287
+ npm test
288
+
289
+ # Type check
290
+ npm run typecheck
291
+
292
+ # Lint
293
+ npm run lint
294
+ ```
295
+
296
+ ## Related Repositories
297
+
298
+ | Repository | Description |
299
+ |------------|-------------|
300
+ | [dakera](https://github.com/dakera-ai/dakera) | Core vector database engine (Rust) |
301
+ | [dakera-py](https://github.com/dakera-ai/dakera-py) | Python SDK |
302
+ | [dakera-go](https://github.com/dakera-ai/dakera-go) | Go SDK |
303
+ | [dakera-rs](https://github.com/dakera-ai/dakera-rs) | Rust SDK |
304
+ | [dakera-cli](https://github.com/dakera-ai/dakera-cli) | Command-line interface |
305
+ | [dakera-mcp](https://github.com/dakera-ai/dakera-mcp) | MCP Server for AI agent memory |
306
+ | [dakera-dashboard](https://github.com/dakera-ai/dakera-dashboard) | Admin dashboard (Leptos/WASM) |
307
+ | [dakera-docs](https://github.com/dakera-ai/dakera-docs) | Documentation |
308
+ | [dakera-deploy](https://github.com/dakera-ai/dakera-deploy) | Deployment configs and Docker Compose |
309
+ | [dakera-cortex](https://github.com/dakera-ai/dakera-cortex) | Flagship demo with AI agents |
310
+
311
+ ## License
312
+
313
+ MIT License - see [LICENSE](LICENSE) for details.