Benature/omnivore
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

do not pass browser instance to content-handler

Browse Source
This commit is contained in:
Hongbo Wu
2024-05-13 13:10:02 +08:00
parent 86e637febd
commit 0ac5299c32
4 changed files with 4 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
packages/content-handler/package.json
View File

@ -36,7 +36,6 @@
"linkedom": "^0.14.16", "linkedom": "^0.14.16",
"lodash": "^4.17.21", "lodash": "^4.17.21",
"luxon": "^3.0.4", "luxon": "^3.0.4",
"puppeteer-core": "^20.9.0",
"underscore": "^1.13.6", "underscore": "^1.13.6",
"uuid": "^9.0.0" "uuid": "^9.0.0"
}, },

3
packages/content-handler/src/content-handler.ts
View File

@ -1,7 +1,6 @@
import addressparser from 'addressparser' import addressparser from 'addressparser'
import axios from 'axios' import axios from 'axios'
import { parseHTML } from 'linkedom' import { parseHTML } from 'linkedom'
import { Browser } from 'puppeteer-core'
import { v4 as uuid } from 'uuid' import { v4 as uuid } from 'uuid'
interface Unsubscribe { interface Unsubscribe {
@ -61,7 +60,7 @@ export abstract class ContentHandler {
return false return false
} }
async preHandle(url: string, browser?: Browser): Promise<PreHandleResult> { async preHandle(url: string): Promise<PreHandleResult> {
return Promise.resolve({ url }) return Promise.resolve({ url })
} }

6
packages/content-handler/src/index.ts
View File

@ -1,5 +1,4 @@
import { parseHTML } from 'linkedom' import { parseHTML } from 'linkedom'
import { Browser } from 'puppeteer-core'
import { import {
ContentHandler, ContentHandler,
NewsletterInput, NewsletterInput,
@ -104,8 +103,7 @@ const newsletterHandlers: ContentHandler[] = [
] ]
export const preHandleContent = async ( export const preHandleContent = async (
url: string, url: string
browser: Browser
): Promise<PreHandleResult | undefined> => { ): Promise<PreHandleResult | undefined> => {
// Before we run the regular handlers we check to see if we need tp // Before we run the regular handlers we check to see if we need tp
// pre-resolve the URL. TODO: This should probably happen recursively, // pre-resolve the URL. TODO: This should probably happen recursively,
@ -129,7 +127,7 @@ export const preHandleContent = async (
for (const handler of contentHandlers) { for (const handler of contentHandlers) {
if (handler.shouldPreHandle(url)) { if (handler.shouldPreHandle(url)) {
console.log('preHandleContent', handler.name, url) console.log('preHandleContent', handler.name, url)
return handler.preHandle(url, browser) return handler.preHandle(url)
} }
} }
return undefined return undefined

3
packages/puppeteer-parse/src/index.ts
View File

@ -163,8 +163,7 @@ export const fetchContent = async (
// pre handle url with custom handlers // pre handle url with custom handlers
try { try {
const browser = await getBrowserPromise const result = await preHandleContent(url)
const result = await preHandleContent(url, browser)
if (result && result.url) { if (result && result.url) {
validateUrlString(url) validateUrlString(url)
url = result.url url = result.url