112 lines
5.0 KiB
TypeScript
112 lines
5.0 KiB
TypeScript
import * as path from 'path'
|
|
import { RunnableToolFunction } from 'openai/lib/RunnableFunction'
|
|
|
|
import puppeteer from 'puppeteer'
|
|
import TurndownService from 'turndown'
|
|
|
|
import { toolLogger } from '../../index.js'
|
|
import { IKBotTask } from '../../types.js'
|
|
|
|
const turndown = new TurndownService()
|
|
|
|
export const tools = (target: string, options: IKBotTask): Array<any> => {
|
|
const logger = toolLogger('web', options)
|
|
return [
|
|
{
|
|
type: 'function',
|
|
function: {
|
|
name: 'browse_page',
|
|
description: 'Browse a webpage and return its content as markdown, all links, images and pages main image',
|
|
parameters: {
|
|
type: 'object',
|
|
properties: {
|
|
url: {
|
|
type: 'string',
|
|
description: 'URL of the webpage to browse'
|
|
}
|
|
},
|
|
required: ['url']
|
|
},
|
|
function: async (params: any) => {
|
|
try {
|
|
logger.debug(`Tool::BrowsePage Browsing ${params.url}`);
|
|
const browser = await puppeteer.launch({
|
|
headless: true,
|
|
args: ['--no-sandbox', '--disable-setuid-sandbox']
|
|
})
|
|
|
|
try {
|
|
const page = await browser.newPage()
|
|
logger.debug(`Tool::Web::BrowsePage Opening page ${params.url}`)
|
|
await page.goto(params.url, {
|
|
waitUntil: 'networkidle2'
|
|
})
|
|
|
|
const pageData = await page.evaluate((selector) => {
|
|
const elementsToRemove = document.querySelectorAll(
|
|
'script, style, link, meta, noscript, iframe, [style*="display:none"],[style*="display: none"], .hidden'
|
|
)
|
|
elementsToRemove.forEach(el => el.remove())
|
|
|
|
const links = Array.from(document.querySelectorAll('a'))
|
|
.map(a => ({
|
|
text: a.textContent?.trim() || '',
|
|
href: a.href
|
|
}))
|
|
.filter(link => link.href && link.href.startsWith('http'))
|
|
.slice(0, 20)
|
|
|
|
const images = Array.from(document.querySelectorAll('img'))
|
|
.map(img => ({
|
|
src: img.src,
|
|
alt: img.alt || '',
|
|
width: img.width,
|
|
height: img.height
|
|
}))
|
|
.filter(img => img.src && img.src.startsWith('http'))
|
|
.slice(0, 20)
|
|
|
|
const mainImage = document.querySelector('meta[property="og:image"]')?.getAttribute('content') ||
|
|
document.querySelector('meta[name="og:image"]')?.getAttribute('content')
|
|
|
|
let content
|
|
const body = document.body
|
|
content = body ? body.innerHTML : ''
|
|
return {
|
|
content,
|
|
links,
|
|
images,
|
|
ogImage: mainImage
|
|
}
|
|
}, null)
|
|
|
|
const markdown = turndown.turndown(pageData.content)
|
|
await browser.close()
|
|
const ret = {
|
|
success: true,
|
|
markdown: markdown,
|
|
links: pageData.links,
|
|
images: pageData.images,
|
|
mainImage: pageData.ogImage,
|
|
url: params.url
|
|
};
|
|
return ret
|
|
} catch (error: any) {
|
|
logger.debug('Error browsing page:', error.message, error);
|
|
await browser.close()
|
|
throw error
|
|
}
|
|
} catch (error: any) {
|
|
logger.debug('Error browsing page:', error.message);
|
|
return {
|
|
success: false,
|
|
error: error.message,
|
|
url: params.url
|
|
};
|
|
}
|
|
},
|
|
parse: JSON.parse
|
|
}
|
|
} as RunnableToolFunction<any>
|
|
]
|
|
} |