Flowise/packages/components/nodes/documentloaders/Spider/Spider.ts

276 lines
9.6 KiB
TypeScript

import { omit } from 'lodash'
import { TextSplitter } from 'langchain/text_splitter'
import { Document, DocumentInterface } from '@langchain/core/documents'
import { BaseDocumentLoader } from 'langchain/document_loaders/base'
import { INode, INodeData, INodeParams, ICommonObject, INodeOutputsValue } from '../../../src/Interface'
import { getCredentialData, getCredentialParam, handleEscapeCharacters } from '../../../src/utils'
import SpiderApp from './SpiderApp'
interface SpiderLoaderParameters {
url: string
apiKey?: string
mode?: 'crawl' | 'scrape'
limit?: number
additionalMetadata?: Record<string, unknown>
params?: Record<string, unknown>
}
class SpiderLoader extends BaseDocumentLoader {
private apiKey: string
private url: string
private mode: 'crawl' | 'scrape'
private limit?: number
private additionalMetadata?: Record<string, unknown>
private params?: Record<string, unknown>
constructor(loaderParams: SpiderLoaderParameters) {
super()
const { apiKey, url, mode = 'crawl', limit, additionalMetadata, params } = loaderParams
if (!apiKey) {
throw new Error('Spider API key not set. You can set it as SPIDER_API_KEY in your .env file, or pass it to Spider.')
}
this.apiKey = apiKey
this.url = url
this.mode = mode
this.limit = Number(limit)
this.additionalMetadata = additionalMetadata
this.params = params
}
public async load(): Promise<DocumentInterface[]> {
const app = new SpiderApp({ apiKey: this.apiKey })
let spiderDocs: any[]
if (this.mode === 'scrape') {
const response = await app.scrapeUrl(this.url, this.params)
if (!response.success) {
throw new Error(`Spider: Failed to scrape URL. Error: ${response.error}`)
}
spiderDocs = [response.data]
} else if (this.mode === 'crawl') {
if (this.params) {
this.params.limit = this.limit
}
const response = await app.crawlUrl(this.url, this.params)
if (!response.success) {
throw new Error(`Spider: Failed to crawl URL. Error: ${response.error}`)
}
spiderDocs = response.data
} else {
throw new Error(`Unrecognized mode '${this.mode}'. Expected one of 'crawl', 'scrape'.`)
}
return spiderDocs.map(
(doc) =>
new Document({
pageContent: doc.content || '',
metadata: {
...(this.additionalMetadata || {}),
source: doc.url
}
})
)
}
}
class Spider_DocumentLoaders implements INode {
label: string
name: string
description: string
type: string
icon: string
version: number
category: string
baseClasses: string[]
inputs: INodeParams[]
credential: INodeParams
outputs: INodeOutputsValue[]
constructor() {
this.label = 'Spider Document Loaders'
this.name = 'spiderDocumentLoaders'
this.version = 2.0
this.type = 'Document'
this.icon = 'spider.svg'
this.category = 'Document Loaders'
this.description = 'Scrape & Crawl the web with Spider'
this.baseClasses = [this.type]
this.inputs = [
{
label: 'Text Splitter',
name: 'textSplitter',
type: 'TextSplitter',
optional: true
},
{
label: 'Mode',
name: 'mode',
type: 'options',
options: [
{
label: 'Scrape',
name: 'scrape',
description: 'Scrape a single page'
},
{
label: 'Crawl',
name: 'crawl',
description: 'Crawl a website and extract pages within the same domain'
}
],
default: 'scrape'
},
{
label: 'Web Page URL',
name: 'url',
type: 'string',
placeholder: 'https://spider.cloud'
},
{
label: 'Limit',
name: 'limit',
type: 'number',
default: 25
},
{
label: 'Additional Metadata',
name: 'additional_metadata',
type: 'json',
description: 'Additional metadata to be added to the extracted documents',
optional: true,
additionalParams: true
},
{
label: 'Additional Parameters',
name: 'params',
description:
'Find all the available parameters in the <a _target="blank" href="https://spider.cloud/docs/api">Spider API documentation</a>',
additionalParams: true,
placeholder: '{ "anti_bot": true }',
type: 'json',
optional: true
},
{
label: 'Omit Metadata Keys',
name: 'omitMetadataKeys',
type: 'string',
rows: 4,
description:
'Each document loader comes with a default set of metadata keys that are extracted from the document. You can use this field to omit some of the default metadata keys. The value should be a list of keys, seperated by comma. Use * to omit all metadata keys execept the ones you specify in the Additional Metadata field',
placeholder: 'key1, key2, key3.nestedKey1',
optional: true,
additionalParams: true
}
]
this.credential = {
label: 'Credential',
name: 'credential',
type: 'credential',
credentialNames: ['spiderApi']
}
this.outputs = [
{
label: 'Document',
name: 'document',
description: 'Array of document objects containing metadata and pageContent',
baseClasses: [...this.baseClasses, 'json']
},
{
label: 'Text',
name: 'text',
description: 'Concatenated string from pageContent of documents',
baseClasses: ['string', 'json']
}
]
}
async init(nodeData: INodeData, _: string, options: ICommonObject): Promise<any> {
const textSplitter = nodeData.inputs?.textSplitter as TextSplitter
const url = nodeData.inputs?.url as string
const mode = nodeData.inputs?.mode as 'crawl' | 'scrape'
const limit = nodeData.inputs?.limit as number
let additionalMetadata = nodeData.inputs?.additional_metadata
let params = nodeData.inputs?.params || {}
const credentialData = await getCredentialData(nodeData.credential ?? '', options)
const spiderApiKey = getCredentialParam('spiderApiKey', credentialData, nodeData)
const _omitMetadataKeys = nodeData.inputs?.omitMetadataKeys as string
const output = nodeData.outputs?.output as string
let omitMetadataKeys: string[] = []
if (_omitMetadataKeys) {
omitMetadataKeys = _omitMetadataKeys.split(',').map((key) => key.trim())
}
if (typeof params === 'string') {
try {
params = JSON.parse(params)
} catch (e) {
console.error('Invalid JSON string provided for params')
}
}
if (additionalMetadata) {
if (typeof additionalMetadata === 'string') {
try {
additionalMetadata = JSON.parse(additionalMetadata)
} catch (e) {
console.error('Invalid JSON string provided for additional metadata')
}
} else if (typeof additionalMetadata !== 'object') {
console.error('Additional metadata must be a valid JSON object')
}
} else {
additionalMetadata = {}
}
// Ensure return_format is set to markdown
params.return_format = 'markdown'
const input: SpiderLoaderParameters = {
url,
mode: mode as 'crawl' | 'scrape',
apiKey: spiderApiKey,
limit: limit as number,
additionalMetadata: additionalMetadata as Record<string, unknown>,
params: params as Record<string, unknown>
}
const loader = new SpiderLoader(input)
let docs = []
if (textSplitter) {
docs = await loader.loadAndSplit(textSplitter)
} else {
docs = await loader.load()
}
docs = docs.map((doc: DocumentInterface) => ({
...doc,
metadata:
_omitMetadataKeys === '*'
? additionalMetadata
: omit(
{
...doc.metadata,
...additionalMetadata
},
omitMetadataKeys
)
}))
if (output === 'document') {
return docs
} else {
let finaltext = ''
for (const doc of docs) {
finaltext += `${doc.pageContent}\n`
}
return handleEscapeCharacters(finaltext, false)
}
}
}
module.exports = { nodeClass: Spider_DocumentLoaders }