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

344 lines
12 KiB
TypeScript

import { omit } from 'lodash'
import { ICommonObject, IDocument, INode, INodeData, INodeParams } from '../../../src/Interface'
import { TextSplitter } from 'langchain/text_splitter'
import jsonpointer from 'jsonpointer'
import { getFileFromStorage, handleEscapeCharacters, INodeOutputsValue } from '../../../src'
import { BaseDocumentLoader } from 'langchain/document_loaders/base'
import { Document } from '@langchain/core/documents'
import type { readFile as ReadFileT } from 'node:fs/promises'
const howToUseCode = `
You can add metadata dynamically from the document:
For example, if the document is:
\`\`\`jsonl
{
"source": "www.example.com", "content": "Hello World!"
}
{
"source": "www.example2.com", "content": "Hi World!"
}
\`\`\`
You can have the "source" value as metadata by returning the following:
\`\`\`json
{
"source": "/source"
}
\`\`\``
class Jsonlines_DocumentLoaders implements INode {
label: string
name: string
version: number
description: string
type: string
icon: string
category: string
baseClasses: string[]
inputs: INodeParams[]
outputs: INodeOutputsValue[]
constructor() {
this.label = 'Json Lines File'
this.name = 'jsonlinesFile'
this.version = 3.0
this.type = 'Document'
this.icon = 'jsonlines.svg'
this.category = 'Document Loaders'
this.description = `Load data from JSON Lines files`
this.baseClasses = [this.type]
this.inputs = [
{
label: 'Jsonlines File',
name: 'jsonlinesFile',
type: 'file',
fileType: '.jsonl'
},
{
label: 'Text Splitter',
name: 'textSplitter',
type: 'TextSplitter',
optional: true
},
{
label: 'Pointer Extraction',
name: 'pointerName',
type: 'string',
placeholder: 'key',
description: 'Ex: { "key": "value" }, Pointer Extraction = "key", "value" will be extracted as pageContent of the chunk',
optional: false
},
{
label: 'Additional Metadata',
name: 'metadata',
type: 'json',
description:
'Additional metadata to be added to the extracted documents. You can add metadata dynamically from the document. Ex: { "key": "value", "source": "www.example.com" }. Metadata: { "page": "/source" } will extract the value of the key "source" from the document and add it to the metadata with the key "page"',
hint: {
label: 'How to use',
value: howToUseCode
},
optional: true,
additionalParams: 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.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 jsonLinesFileBase64 = nodeData.inputs?.jsonlinesFile as string
const pointerName = nodeData.inputs?.pointerName as string
const metadata = nodeData.inputs?.metadata
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())
}
let docs: IDocument[] = []
let files: string[] = []
let pointer = '/' + pointerName.trim()
//FILE-STORAGE::["CONTRIBUTING.md","LICENSE.md","README.md"]
if (jsonLinesFileBase64.startsWith('FILE-STORAGE::')) {
const fileName = jsonLinesFileBase64.replace('FILE-STORAGE::', '')
if (fileName.startsWith('[') && fileName.endsWith(']')) {
files = JSON.parse(fileName)
} else {
files = [fileName]
}
const orgId = options.orgId
const chatflowid = options.chatflowid
for (const file of files) {
if (!file) continue
const fileData = await getFileFromStorage(file, orgId, chatflowid)
const blob = new Blob([fileData])
const loader = new JSONLinesLoader(blob, pointer, metadata)
if (textSplitter) {
let splittedDocs = await loader.load()
splittedDocs = await textSplitter.splitDocuments(splittedDocs)
docs.push(...splittedDocs)
} else {
docs.push(...(await loader.load()))
}
}
} else {
if (jsonLinesFileBase64.startsWith('[') && jsonLinesFileBase64.endsWith(']')) {
files = JSON.parse(jsonLinesFileBase64)
} else {
files = [jsonLinesFileBase64]
}
for (const file of files) {
if (!file) continue
const splitDataURI = file.split(',')
splitDataURI.pop()
const bf = Buffer.from(splitDataURI.pop() || '', 'base64')
const blob = new Blob([bf])
const loader = new JSONLinesLoader(blob, pointer, metadata)
if (textSplitter) {
let splittedDocs = await loader.load()
splittedDocs = await textSplitter.splitDocuments(splittedDocs)
docs.push(...splittedDocs)
} else {
docs.push(...(await loader.load()))
}
}
}
if (metadata) {
let parsedMetadata = typeof metadata === 'object' ? metadata : JSON.parse(metadata)
parsedMetadata = removeValuesStartingWithSlash(parsedMetadata)
docs = docs.map((doc) => ({
...doc,
metadata:
_omitMetadataKeys === '*'
? {
...parsedMetadata
}
: omit(
{
...doc.metadata,
...parsedMetadata
},
omitMetadataKeys
)
}))
} else {
docs = docs.map((doc) => ({
...doc,
metadata:
_omitMetadataKeys === '*'
? {}
: omit(
{
...doc.metadata
},
omitMetadataKeys
)
}))
}
if (output === 'document') {
return docs
} else {
let finaltext = ''
for (const doc of docs) {
finaltext += `${doc.pageContent}\n`
}
return handleEscapeCharacters(finaltext, false)
}
}
}
const removeValuesStartingWithSlash = (obj: Record<string, any>): Record<string, any> => {
const result: Record<string, any> = {}
for (const key in obj) {
const value = obj[key]
if (typeof value === 'string' && value.startsWith('/')) {
continue
}
result[key] = value
}
return result
}
class TextLoader extends BaseDocumentLoader {
constructor(public filePathOrBlob: string | Blob) {
super()
}
protected async parse(raw: string): Promise<{ pageContent: string; metadata: ICommonObject }[]> {
return [{ pageContent: raw, metadata: {} }]
}
public async load(): Promise<Document[]> {
let text: string
let metadata: Record<string, string>
if (typeof this.filePathOrBlob === 'string') {
const { readFile } = await TextLoader.imports()
text = await readFile(this.filePathOrBlob, 'utf8')
metadata = { source: this.filePathOrBlob }
} else {
text = await this.filePathOrBlob.text()
metadata = { source: 'blob', blobType: this.filePathOrBlob.type }
}
const parsed = await this.parse(text)
parsed.forEach((parsedData, i) => {
const { pageContent } = parsedData
if (typeof pageContent !== 'string') {
throw new Error(`Expected string, at position ${i} got ${typeof pageContent}`)
}
})
return parsed.map((parsedData, i) => {
const { pageContent, metadata: additionalMetadata } = parsedData
return new Document({
pageContent,
metadata:
parsed.length === 1
? { ...metadata, ...additionalMetadata }
: {
...metadata,
line: i + 1,
...additionalMetadata
}
})
})
}
static async imports(): Promise<{
readFile: typeof ReadFileT
}> {
try {
const { readFile } = await import('node:fs/promises')
return { readFile }
} catch (e) {
console.error(e)
throw new Error(`Failed to load fs/promises. Make sure you are running in Node.js environment.`)
}
}
}
class JSONLinesLoader extends TextLoader {
metadata?: ICommonObject
additionalMetadata: ICommonObject[] = []
constructor(filePathOrBlob: string | Blob, public pointer: string, metadata?: any) {
super(filePathOrBlob)
if (metadata) {
this.metadata = typeof metadata === 'object' ? metadata : JSON.parse(metadata)
}
}
async getAdditionalMetadata(): Promise<ICommonObject[]> {
return this.additionalMetadata
}
protected async parse(raw: string): Promise<{ pageContent: string; metadata: ICommonObject }[]> {
const lines = raw.split('\n')
const jsons = lines
.map((line) => line.trim())
.filter(Boolean)
.map((line) => JSON.parse(line))
const pointer = jsonpointer.compile(this.pointer)
if (this.metadata) {
const values = Object.values(this.metadata).filter((value) => typeof value === 'string' && value.startsWith('/'))
let newJsons = []
for (const json of jsons) {
let metadata = {}
for (const value of values) {
if (value) {
const key = Object.keys(this.metadata).find((key) => this.metadata?.[key] === value)
if (key) {
metadata = {
...metadata,
[key]: jsonpointer.get(json, value)
}
}
}
}
newJsons.push({ pageContent: pointer.get(json), metadata })
}
return newJsons
}
return jsons.map((json) => {
return { pageContent: pointer.get(json), metadata: {} }
})
}
}
module.exports = { nodeClass: Jsonlines_DocumentLoaders }