Flowise/packages/components/nodes/vectorstores/Couchbase/Couchbase.ts

235 lines
8.9 KiB
TypeScript

/*
* Temporary disabled due to the incompatibility with the docker node-alpine:
* https://github.com/FlowiseAI/Flowise/pull/2303
import { flatten } from 'lodash'
import { Embeddings } from '@langchain/core/embeddings'
import { Document } from '@langchain/core/documents'
import { CouchbaseVectorStore, CouchbaseVectorStoreArgs } from '@langchain/community/vectorstores/couchbase'
import { Cluster } from 'couchbase'
import { ICommonObject, INode, INodeData, INodeOutputsValue, INodeParams, IndexingResult } from '../../../src/Interface'
import { getBaseClasses, getCredentialData, getCredentialParam } from '../../../src/utils'
import { resolveVectorStoreOrRetriever } from '../VectorStoreUtils'
class Couchbase_VectorStores implements INode {
label: string
name: string
version: number
description: string
type: string
icon: string
category: string
badge: string
baseClasses: string[]
inputs: INodeParams[]
credential: INodeParams
outputs: INodeOutputsValue[]
constructor() {
this.label = 'Couchbase'
this.name = 'couchbase'
this.version = 1.0
this.type = 'Couchbase'
this.icon = 'couchbase.svg'
this.category = 'Vector Stores'
this.description = `Upsert embedded data and load existing index using Couchbase, a award-winning distributed NoSQL database`
this.baseClasses = [this.type, 'VectorStoreRetriever', 'BaseRetriever']
this.credential = {
label: 'Connect Credential',
name: 'credential',
type: 'credential',
credentialNames: ['couchbaseApi']
}
this.inputs = [
{
label: 'Document',
name: 'document',
type: 'Document',
list: true,
optional: true
},
{
label: 'Embeddings',
name: 'embeddings',
type: 'Embeddings'
},
{
label: 'Bucket Name',
name: 'bucketName',
placeholder: '<DB_BUCKET_NAME>',
type: 'string'
},
{
label: 'Scope Name',
name: 'scopeName',
placeholder: '<SCOPE_NAME>',
type: 'string'
},
{
label: 'Collection Name',
name: 'collectionName',
placeholder: '<COLLECTION_NAME>',
type: 'string'
},
{
label: 'Index Name',
name: 'indexName',
placeholder: '<VECTOR_INDEX_NAME>',
type: 'string'
},
{
label: 'Content Field',
name: 'textKey',
description: 'Name of the field (column) that contains the actual content',
type: 'string',
default: 'text',
additionalParams: true,
optional: true
},
{
label: 'Embedded Field',
name: 'embeddingKey',
description: 'Name of the field (column) that contains the Embedding',
type: 'string',
default: 'embedding',
additionalParams: true,
optional: true
},
{
label: 'Couchbase Metadata Filter',
name: 'couchbaseMetadataFilter',
type: 'json',
optional: true,
additionalParams: true
},
{
label: 'Top K',
name: 'topK',
description: 'Number of top results to fetch. Default to 4',
placeholder: '4',
type: 'number',
additionalParams: true,
optional: true
}
]
this.outputs = [
{
label: 'Couchbase Retriever',
name: 'retriever',
baseClasses: this.baseClasses
},
{
label: 'Couchbase Vector Store',
name: 'vectorStore',
baseClasses: [this.type, ...getBaseClasses(CouchbaseVectorStore)]
}
]
}
//@ts-ignore
vectorStoreMethods = {
async upsert(nodeData: INodeData, options: ICommonObject): Promise<Partial<IndexingResult>> {
const credentialData = await getCredentialData(nodeData.credential ?? '', options)
const bucketName = nodeData.inputs?.bucketName as string
const scopeName = nodeData.inputs?.scopeName as string
const collectionName = nodeData.inputs?.collectionName as string
const indexName = nodeData.inputs?.indexName as string
let textKey = nodeData.inputs?.textKey as string
let embeddingKey = nodeData.inputs?.embeddingKey as string
const embeddings = nodeData.inputs?.embeddings as Embeddings
let connectionString = getCredentialParam('connectionString', credentialData, nodeData)
let databaseUsername = getCredentialParam('username', credentialData, nodeData)
let databasePassword = getCredentialParam('password', credentialData, nodeData)
const docs = nodeData.inputs?.document as Document[]
const flattenDocs = docs && docs.length ? flatten(docs) : []
const finalDocs = []
for (let i = 0; i < flattenDocs.length; i += 1) {
if (flattenDocs[i] && flattenDocs[i].pageContent) {
const document = new Document(flattenDocs[i])
finalDocs.push(document)
}
}
const couchbaseClient = await Cluster.connect(connectionString, {
username: databaseUsername,
password: databasePassword,
configProfile: 'wanDevelopment'
})
const couchbaseConfig: CouchbaseVectorStoreArgs = {
cluster: couchbaseClient,
bucketName: bucketName,
scopeName: scopeName,
collectionName: collectionName,
indexName: indexName,
textKey: textKey,
embeddingKey: embeddingKey
}
try {
if (!textKey || textKey === '') couchbaseConfig.textKey = 'text'
if (!embeddingKey || embeddingKey === '') couchbaseConfig.embeddingKey = 'embedding'
await CouchbaseVectorStore.fromDocuments(finalDocs, embeddings, couchbaseConfig)
return { numAdded: finalDocs.length, addedDocs: finalDocs }
} catch (e) {
throw new Error(e)
}
}
}
async init(nodeData: INodeData, _: string, options: ICommonObject): Promise<any> {
const credentialData = await getCredentialData(nodeData.credential ?? '', options)
const bucketName = nodeData.inputs?.bucketName as string
const scopeName = nodeData.inputs?.scopeName as string
const collectionName = nodeData.inputs?.collectionName as string
const indexName = nodeData.inputs?.indexName as string
let textKey = nodeData.inputs?.textKey as string
let embeddingKey = nodeData.inputs?.embeddingKey as string
const embeddings = nodeData.inputs?.embeddings as Embeddings
const couchbaseMetadataFilter = nodeData.inputs?.couchbaseMetadataFilter
let connectionString = getCredentialParam('connectionString', credentialData, nodeData)
let databaseUsername = getCredentialParam('username', credentialData, nodeData)
let databasePassword = getCredentialParam('password', credentialData, nodeData)
let metadatafilter
const couchbaseClient = await Cluster.connect(connectionString, {
username: databaseUsername,
password: databasePassword,
configProfile: 'wanDevelopment'
})
const couchbaseConfig: CouchbaseVectorStoreArgs = {
cluster: couchbaseClient,
bucketName: bucketName,
scopeName: scopeName,
collectionName: collectionName,
indexName: indexName,
textKey: textKey,
embeddingKey: embeddingKey
}
try {
if (!textKey || textKey === '') couchbaseConfig.textKey = 'text'
if (!embeddingKey || embeddingKey === '') couchbaseConfig.embeddingKey = 'embedding'
if (couchbaseMetadataFilter) {
metadatafilter = typeof couchbaseMetadataFilter === 'object' ? couchbaseMetadataFilter : JSON.parse(couchbaseMetadataFilter)
}
const vectorStore = await CouchbaseVectorStore.initialize(embeddings, couchbaseConfig)
return resolveVectorStoreOrRetriever(nodeData, vectorStore, metadatafilter)
} catch (e) {
throw new Error(e)
}
}
}
module.exports = { nodeClass: Couchbase_VectorStores }
*/