prehistoric-systems/omnivore
1
0
Fork 0
mirror of https://github.com/omnivore-app/omnivore.git synced 2026-03-11 08:54:26 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Remove dom param in preHandler method

Browse source
This commit is contained in:
Hongbo Wu 2022-10-06 13:16:19 +08:00
parent f607215826
commit c9400ca517
9 changed files with 18 additions and 18 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
packages/content-handler/src/websites/apple-news-handler.ts
View file

@ -8,12 +8,12 @@ export class AppleNewsHandler extends ContentHandler {
this.name = 'Apple News'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
const u = new URL(url)
return u.hostname === 'apple.news'
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
const MOBILE_USER_AGENT =
'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/94.0.4606.61 Safari/537.36'
const response = await axios.get(url, {

4
packages/content-handler/src/websites/bloomberg-handler.ts
View file

@ -8,13 +8,13 @@ export class BloombergHandler extends ContentHandler {
this.name = 'Bloomberg'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
const BLOOMBERG_URL_MATCH =
/https?:\/\/(www\.)?bloomberg.[a-zA-Z0-9()]{1,6}\b([-a-zA-Z0-9()@:%_+.~#?&/=]*)/
return BLOOMBERG_URL_MATCH.test(url.toString())
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
console.log('prehandling bloomberg url', url)
try {

4
packages/content-handler/src/websites/derstandard-handler.ts
View file

@ -8,12 +8,12 @@ export class DerstandardHandler extends ContentHandler {
this.name = 'Derstandard'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
const u = new URL(url)
return u.hostname === 'www.derstandard.at'
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
const response = await axios.get(url, {
// set cookie to give consent to get the article
headers: {

4
packages/content-handler/src/websites/image-handler.ts
View file

@ -6,12 +6,12 @@ export class ImageHandler extends ContentHandler {
this.name = 'Image'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
const IMAGE_URL_PATTERN = /(https?:\/\/.*\.(?:jpg|jpeg|png|webp))/i
return IMAGE_URL_PATTERN.test(url.toString())
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
const title = url.toString().split('/').pop() || 'Image'
const content = `
<html>

4
packages/content-handler/src/websites/medium-handler.ts
View file

@ -6,12 +6,12 @@ export class MediumHandler extends ContentHandler {
this.name = 'Medium'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
const u = new URL(url)
return u.hostname.endsWith('medium.com')
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
console.log('prehandling medium url', url)
try {

4
packages/content-handler/src/websites/pdf-handler.ts
View file

@ -6,13 +6,13 @@ export class PdfHandler extends ContentHandler {
this.name = 'PDF'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
const u = new URL(url)
const path = u.pathname.replace(u.search, '')
return path.endsWith('.pdf')
}
async preHandle(_url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
return Promise.resolve({ contentType: 'application/pdf' })
}
}

4
packages/content-handler/src/websites/scrapingBee-handler.ts
View file

@ -8,14 +8,14 @@ export class ScrapingBeeHandler extends ContentHandler {
this.name = 'ScrapingBee'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
const u = new URL(url)
const hostnames = ['nytimes.com', 'news.google.com']
return hostnames.some((h) => u.hostname.endsWith(h))
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
console.log('prehandling url with scrapingbee', url)
try {

4
packages/content-handler/src/websites/twitter-handler.ts
View file

@ -140,11 +140,11 @@ export class TwitterHandler extends ContentHandler {
this.name = 'Twitter'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
return !!TWITTER_BEARER_TOKEN && TWITTER_URL_MATCH.test(url.toString())
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
const tweetId = tweetIdFromStatusUrl(url)
if (!tweetId) {
throw new Error('could not find tweet id in url')

4
packages/content-handler/src/websites/youtube-handler.ts
View file

@ -24,11 +24,11 @@ export class YoutubeHandler extends ContentHandler {
this.name = 'Youtube'
}
shouldPreHandle(url: string, dom?: Document): boolean {
shouldPreHandle(url: string): boolean {
return YOUTUBE_URL_MATCH.test(url.toString())
}
async preHandle(url: string, document?: Document): Promise<PreHandleResult> {
async preHandle(url: string): Promise<PreHandleResult> {
const videoId = getYoutubeVideoId(url)
if (!videoId) {
return {}