@mastra/chroma 0.0.0-vnext-inngest-20250508122351 → 0.0.0-vnext-20251104230439
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/CHANGELOG.md +715 -4
- package/LICENSE.md +12 -4
- package/README.md +47 -21
- package/dist/index.cjs +301 -105
- package/dist/index.cjs.map +1 -0
- package/dist/index.d.ts +3 -2
- package/dist/index.d.ts.map +1 -0
- package/dist/index.js +299 -103
- package/dist/index.js.map +1 -0
- package/dist/vector/filter.d.ts +19 -0
- package/dist/vector/filter.d.ts.map +1 -0
- package/dist/vector/index.d.ts +70 -0
- package/dist/vector/index.d.ts.map +1 -0
- package/dist/vector/prompt.d.ts +6 -0
- package/dist/vector/prompt.d.ts.map +1 -0
- package/package.json +32 -14
- package/dist/_tsup-dts-rollup.d.cts +0 -75
- package/dist/_tsup-dts-rollup.d.ts +0 -75
- package/dist/index.d.cts +0 -2
- package/docker-compose.yaml +0 -7
- package/eslint.config.js +0 -6
- package/src/index.ts +0 -2
- package/src/vector/filter.test.ts +0 -413
- package/src/vector/filter.ts +0 -115
- package/src/vector/index.test.ts +0 -1657
- package/src/vector/index.ts +0 -231
- package/src/vector/prompt.ts +0 -72
- package/tsconfig.json +0 -5
- package/vitest.config.ts +0 -8
package/src/vector/index.ts
DELETED
|
@@ -1,231 +0,0 @@
|
|
|
1
|
-
import { MastraVector } from '@mastra/core/vector';
|
|
2
|
-
import type {
|
|
3
|
-
QueryResult,
|
|
4
|
-
IndexStats,
|
|
5
|
-
CreateIndexParams,
|
|
6
|
-
UpsertVectorParams,
|
|
7
|
-
QueryVectorParams,
|
|
8
|
-
ParamsToArgs,
|
|
9
|
-
QueryVectorArgs,
|
|
10
|
-
UpsertVectorArgs,
|
|
11
|
-
} from '@mastra/core/vector';
|
|
12
|
-
|
|
13
|
-
import type { VectorFilter } from '@mastra/core/vector/filter';
|
|
14
|
-
import { ChromaClient } from 'chromadb';
|
|
15
|
-
import type { UpdateRecordsParams, Collection } from 'chromadb';
|
|
16
|
-
import { ChromaFilterTranslator } from './filter';
|
|
17
|
-
|
|
18
|
-
interface ChromaUpsertVectorParams extends UpsertVectorParams {
|
|
19
|
-
documents?: string[];
|
|
20
|
-
}
|
|
21
|
-
|
|
22
|
-
type ChromaUpsertArgs = [...UpsertVectorArgs, string[]?];
|
|
23
|
-
|
|
24
|
-
interface ChromaQueryVectorParams extends QueryVectorParams {
|
|
25
|
-
documentFilter?: VectorFilter;
|
|
26
|
-
}
|
|
27
|
-
|
|
28
|
-
type ChromaQueryArgs = [...QueryVectorArgs, VectorFilter?];
|
|
29
|
-
|
|
30
|
-
export class ChromaVector extends MastraVector {
|
|
31
|
-
private client: ChromaClient;
|
|
32
|
-
private collections: Map<string, any>;
|
|
33
|
-
|
|
34
|
-
constructor({
|
|
35
|
-
path,
|
|
36
|
-
auth,
|
|
37
|
-
}: {
|
|
38
|
-
path: string;
|
|
39
|
-
auth?: {
|
|
40
|
-
provider: string;
|
|
41
|
-
credentials: string;
|
|
42
|
-
};
|
|
43
|
-
}) {
|
|
44
|
-
super();
|
|
45
|
-
this.client = new ChromaClient({
|
|
46
|
-
path,
|
|
47
|
-
auth,
|
|
48
|
-
});
|
|
49
|
-
this.collections = new Map();
|
|
50
|
-
}
|
|
51
|
-
|
|
52
|
-
async getCollection(indexName: string, throwIfNotExists: boolean = true) {
|
|
53
|
-
try {
|
|
54
|
-
const collection = await this.client.getCollection({ name: indexName, embeddingFunction: undefined as any });
|
|
55
|
-
this.collections.set(indexName, collection);
|
|
56
|
-
} catch {
|
|
57
|
-
if (throwIfNotExists) {
|
|
58
|
-
throw new Error(`Index ${indexName} does not exist`);
|
|
59
|
-
}
|
|
60
|
-
return null;
|
|
61
|
-
}
|
|
62
|
-
return this.collections.get(indexName);
|
|
63
|
-
}
|
|
64
|
-
|
|
65
|
-
private validateVectorDimensions(vectors: number[][], dimension: number): void {
|
|
66
|
-
for (let i = 0; i < vectors.length; i++) {
|
|
67
|
-
if (vectors?.[i]?.length !== dimension) {
|
|
68
|
-
throw new Error(
|
|
69
|
-
`Vector at index ${i} has invalid dimension ${vectors?.[i]?.length}. Expected ${dimension} dimensions.`,
|
|
70
|
-
);
|
|
71
|
-
}
|
|
72
|
-
}
|
|
73
|
-
}
|
|
74
|
-
|
|
75
|
-
async upsert(...args: ParamsToArgs<ChromaUpsertVectorParams> | ChromaUpsertArgs): Promise<string[]> {
|
|
76
|
-
const params = this.normalizeArgs<ChromaUpsertVectorParams, ChromaUpsertArgs>('upsert', args, ['documents']);
|
|
77
|
-
|
|
78
|
-
const { indexName, vectors, metadata, ids, documents } = params;
|
|
79
|
-
|
|
80
|
-
const collection = await this.getCollection(indexName);
|
|
81
|
-
|
|
82
|
-
// Get index stats to check dimension
|
|
83
|
-
const stats = await this.describeIndex(indexName);
|
|
84
|
-
|
|
85
|
-
// Validate vector dimensions
|
|
86
|
-
this.validateVectorDimensions(vectors, stats.dimension);
|
|
87
|
-
|
|
88
|
-
// Generate IDs if not provided
|
|
89
|
-
const generatedIds = ids || vectors.map(() => crypto.randomUUID());
|
|
90
|
-
|
|
91
|
-
// Ensure metadata exists for each vector
|
|
92
|
-
const normalizedMetadata = metadata || vectors.map(() => ({}));
|
|
93
|
-
|
|
94
|
-
await collection.upsert({
|
|
95
|
-
ids: generatedIds,
|
|
96
|
-
embeddings: vectors,
|
|
97
|
-
metadatas: normalizedMetadata,
|
|
98
|
-
documents: documents,
|
|
99
|
-
});
|
|
100
|
-
|
|
101
|
-
return generatedIds;
|
|
102
|
-
}
|
|
103
|
-
|
|
104
|
-
private HnswSpaceMap = {
|
|
105
|
-
cosine: 'cosine',
|
|
106
|
-
euclidean: 'l2',
|
|
107
|
-
dotproduct: 'ip',
|
|
108
|
-
l2: 'euclidean',
|
|
109
|
-
ip: 'dotproduct',
|
|
110
|
-
};
|
|
111
|
-
|
|
112
|
-
async createIndex(...args: ParamsToArgs<CreateIndexParams>): Promise<void> {
|
|
113
|
-
const params = this.normalizeArgs<CreateIndexParams>('createIndex', args);
|
|
114
|
-
const { indexName, dimension, metric = 'cosine' } = params;
|
|
115
|
-
|
|
116
|
-
if (!Number.isInteger(dimension) || dimension <= 0) {
|
|
117
|
-
throw new Error('Dimension must be a positive integer');
|
|
118
|
-
}
|
|
119
|
-
const hnswSpace = this.HnswSpaceMap[metric];
|
|
120
|
-
if (!['cosine', 'l2', 'ip'].includes(hnswSpace)) {
|
|
121
|
-
throw new Error(`Invalid metric: "${metric}". Must be one of: cosine, euclidean, dotproduct`);
|
|
122
|
-
}
|
|
123
|
-
try {
|
|
124
|
-
await this.client.createCollection({
|
|
125
|
-
name: indexName,
|
|
126
|
-
metadata: {
|
|
127
|
-
dimension,
|
|
128
|
-
'hnsw:space': hnswSpace,
|
|
129
|
-
},
|
|
130
|
-
});
|
|
131
|
-
} catch (error: any) {
|
|
132
|
-
// Check for 'already exists' error
|
|
133
|
-
const message = error?.message || error?.toString();
|
|
134
|
-
if (message && message.toLowerCase().includes('already exists')) {
|
|
135
|
-
// Fetch collection info and check dimension
|
|
136
|
-
await this.validateExistingIndex(indexName, dimension, metric);
|
|
137
|
-
return;
|
|
138
|
-
}
|
|
139
|
-
throw error;
|
|
140
|
-
}
|
|
141
|
-
}
|
|
142
|
-
|
|
143
|
-
transformFilter(filter?: VectorFilter) {
|
|
144
|
-
const translator = new ChromaFilterTranslator();
|
|
145
|
-
return translator.translate(filter);
|
|
146
|
-
}
|
|
147
|
-
async query(...args: ParamsToArgs<ChromaQueryVectorParams> | ChromaQueryArgs): Promise<QueryResult[]> {
|
|
148
|
-
const params = this.normalizeArgs<ChromaQueryVectorParams, ChromaQueryArgs>('query', args, ['documentFilter']);
|
|
149
|
-
|
|
150
|
-
const { indexName, queryVector, topK = 10, filter, includeVector = false, documentFilter } = params;
|
|
151
|
-
|
|
152
|
-
const collection = await this.getCollection(indexName, true);
|
|
153
|
-
|
|
154
|
-
const defaultInclude = ['documents', 'metadatas', 'distances'];
|
|
155
|
-
|
|
156
|
-
const translatedFilter = this.transformFilter(filter);
|
|
157
|
-
const results = await collection.query({
|
|
158
|
-
queryEmbeddings: [queryVector],
|
|
159
|
-
nResults: topK,
|
|
160
|
-
where: translatedFilter,
|
|
161
|
-
whereDocument: documentFilter,
|
|
162
|
-
include: includeVector ? [...defaultInclude, 'embeddings'] : defaultInclude,
|
|
163
|
-
});
|
|
164
|
-
|
|
165
|
-
// Transform ChromaDB results to QueryResult format
|
|
166
|
-
return (results.ids[0] || []).map((id: string, index: number) => ({
|
|
167
|
-
id,
|
|
168
|
-
score: results.distances?.[0]?.[index] || 0,
|
|
169
|
-
metadata: results.metadatas?.[0]?.[index] || {},
|
|
170
|
-
document: results.documents?.[0]?.[index],
|
|
171
|
-
...(includeVector && { vector: results.embeddings?.[0]?.[index] || [] }),
|
|
172
|
-
}));
|
|
173
|
-
}
|
|
174
|
-
|
|
175
|
-
async listIndexes(): Promise<string[]> {
|
|
176
|
-
const collections = await this.client.listCollections();
|
|
177
|
-
return collections.map(collection => collection);
|
|
178
|
-
}
|
|
179
|
-
|
|
180
|
-
async describeIndex(indexName: string): Promise<IndexStats> {
|
|
181
|
-
const collection = await this.getCollection(indexName);
|
|
182
|
-
const count = await collection.count();
|
|
183
|
-
const metadata = collection.metadata;
|
|
184
|
-
|
|
185
|
-
const hnswSpace = metadata?.['hnsw:space'] as 'cosine' | 'l2' | 'ip';
|
|
186
|
-
|
|
187
|
-
return {
|
|
188
|
-
dimension: metadata?.dimension || 0,
|
|
189
|
-
count,
|
|
190
|
-
metric: this.HnswSpaceMap[hnswSpace] as 'cosine' | 'euclidean' | 'dotproduct',
|
|
191
|
-
};
|
|
192
|
-
}
|
|
193
|
-
|
|
194
|
-
async deleteIndex(indexName: string): Promise<void> {
|
|
195
|
-
await this.client.deleteCollection({ name: indexName });
|
|
196
|
-
this.collections.delete(indexName);
|
|
197
|
-
}
|
|
198
|
-
|
|
199
|
-
async updateIndexById(
|
|
200
|
-
indexName: string,
|
|
201
|
-
id: string,
|
|
202
|
-
update: { vector?: number[]; metadata?: Record<string, any> },
|
|
203
|
-
): Promise<void> {
|
|
204
|
-
if (!update.vector && !update.metadata) {
|
|
205
|
-
throw new Error('No updates provided');
|
|
206
|
-
}
|
|
207
|
-
|
|
208
|
-
const collection: Collection = await this.getCollection(indexName, true);
|
|
209
|
-
|
|
210
|
-
const updateOptions: UpdateRecordsParams = { ids: [id] };
|
|
211
|
-
|
|
212
|
-
if (update?.vector) {
|
|
213
|
-
updateOptions.embeddings = [update.vector];
|
|
214
|
-
}
|
|
215
|
-
|
|
216
|
-
if (update?.metadata) {
|
|
217
|
-
updateOptions.metadatas = [update.metadata];
|
|
218
|
-
}
|
|
219
|
-
|
|
220
|
-
return await collection.update(updateOptions);
|
|
221
|
-
}
|
|
222
|
-
|
|
223
|
-
async deleteIndexById(indexName: string, id: string): Promise<void> {
|
|
224
|
-
try {
|
|
225
|
-
const collection: Collection = await this.getCollection(indexName, true);
|
|
226
|
-
await collection.delete({ ids: [id] });
|
|
227
|
-
} catch (error: any) {
|
|
228
|
-
throw new Error(`Failed to delete index by id: ${id} for index name: ${indexName}: ${error.message}`);
|
|
229
|
-
}
|
|
230
|
-
}
|
|
231
|
-
}
|
package/src/vector/prompt.ts
DELETED
|
@@ -1,72 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* Vector store specific prompt that details supported operators and examples.
|
|
3
|
-
* This prompt helps users construct valid filters for Chroma Vector.
|
|
4
|
-
*/
|
|
5
|
-
export const CHROMA_PROMPT = `When querying Chroma, you can ONLY use the operators listed below. Any other operators will be rejected.
|
|
6
|
-
Important: Don't explain how to construct the filter - use the specified operators and fields to search the content and return relevant results.
|
|
7
|
-
If a user tries to give an explicit operator that is not supported, reject the filter entirely and let them know that the operator is not supported.
|
|
8
|
-
|
|
9
|
-
Basic Comparison Operators:
|
|
10
|
-
- $eq: Exact match (default when using field: value)
|
|
11
|
-
Example: { "category": "electronics" }
|
|
12
|
-
- $ne: Not equal
|
|
13
|
-
Example: { "category": { "$ne": "electronics" } }
|
|
14
|
-
- $gt: Greater than
|
|
15
|
-
Example: { "price": { "$gt": 100 } }
|
|
16
|
-
- $gte: Greater than or equal
|
|
17
|
-
Example: { "price": { "$gte": 100 } }
|
|
18
|
-
- $lt: Less than
|
|
19
|
-
Example: { "price": { "$lt": 100 } }
|
|
20
|
-
- $lte: Less than or equal
|
|
21
|
-
Example: { "price": { "$lte": 100 } }
|
|
22
|
-
|
|
23
|
-
Array Operators:
|
|
24
|
-
- $in: Match any value in array
|
|
25
|
-
Example: { "category": { "$in": ["electronics", "books"] } }
|
|
26
|
-
- $nin: Does not match any value in array
|
|
27
|
-
Example: { "category": { "$nin": ["electronics", "books"] } }
|
|
28
|
-
|
|
29
|
-
Logical Operators:
|
|
30
|
-
- $and: Logical AND
|
|
31
|
-
Example: { "$and": [{ "price": { "$gt": 100 } }, { "category": "electronics" }] }
|
|
32
|
-
- $or: Logical OR
|
|
33
|
-
Example: { "$or": [{ "price": { "$lt": 50 } }, { "category": "books" }] }
|
|
34
|
-
|
|
35
|
-
Restrictions:
|
|
36
|
-
- Regex patterns are not supported
|
|
37
|
-
- Element operators are not supported
|
|
38
|
-
- Only $and and $or logical operators are supported
|
|
39
|
-
- Nested fields are supported using dot notation
|
|
40
|
-
- Multiple conditions on the same field are supported with both implicit and explicit $and
|
|
41
|
-
- Empty arrays in $in/$nin will return no results
|
|
42
|
-
- If multiple top-level fields exist, they're wrapped in $and
|
|
43
|
-
- Only logical operators ($and, $or) can be used at the top level
|
|
44
|
-
- All other operators must be used within a field condition
|
|
45
|
-
Valid: { "field": { "$gt": 100 } }
|
|
46
|
-
Valid: { "$and": [...] }
|
|
47
|
-
Invalid: { "$gt": 100 }
|
|
48
|
-
Invalid: { "$in": [...] }
|
|
49
|
-
- Logical operators must contain field conditions, not direct operators
|
|
50
|
-
Valid: { "$and": [{ "field": { "$gt": 100 } }] }
|
|
51
|
-
Invalid: { "$and": [{ "$gt": 100 }] }
|
|
52
|
-
- Logical operators ($and, $or):
|
|
53
|
-
- Can only be used at top level or nested within other logical operators
|
|
54
|
-
- Can not be used on a field level, or be nested inside a field
|
|
55
|
-
- Can not be used inside an operator
|
|
56
|
-
- Valid: { "$and": [{ "field": { "$gt": 100 } }] }
|
|
57
|
-
- Valid: { "$or": [{ "$and": [{ "field": { "$gt": 100 } }] }] }
|
|
58
|
-
- Invalid: { "field": { "$and": [{ "$gt": 100 }] } }
|
|
59
|
-
- Invalid: { "field": { "$or": [{ "$gt": 100 }] } }
|
|
60
|
-
- Invalid: { "field": { "$gt": { "$and": [{...}] } } }
|
|
61
|
-
|
|
62
|
-
Example Complex Query:
|
|
63
|
-
{
|
|
64
|
-
"$and": [
|
|
65
|
-
{ "category": { "$in": ["electronics", "computers"] } },
|
|
66
|
-
{ "price": { "$gte": 100, "$lte": 1000 } },
|
|
67
|
-
{ "$or": [
|
|
68
|
-
{ "inStock": true },
|
|
69
|
-
{ "preorder": true }
|
|
70
|
-
]}
|
|
71
|
-
]
|
|
72
|
-
}`;
|
package/tsconfig.json
DELETED