Merge pull request #59 from ccbikai/main

README.md

@@ -172,7 +172,7 @@ Yes, `substats` is now version `v2.0-beta`! Most of the updates are under-the-ho
 - [x] Updated to CloudFlare's module workers.
 - [x] Worker is built with `esbuild` instead of `webpack`, extra fast!
 - [x] Support for Newsblur has been deprecated ~~(seems nobody uses it)~~.
-- [x] KV storages are now supported, some routes including `instagram` and `inoreader` depends on this for storing cookies (wip).
+- [x] KV storages are now supported, some routes including `instagram` depends on this for storing cookies (wip).
 - [x] Caching is ported to module workers in 2.0 and supported as always.
 - [x] New documentation and query builder.
 

worker/bindings.d.ts

@@ -6,9 +6,6 @@ interface Env {
   STEAM_API_KEY: string
   TG_BOT_TOKEN: string
   UNSPLASH_ACCESS_TOKEN: string
-  INOREADER_EMAIL: string
-  INOREADER_PASSWORD: string
-  INOREADER_COOKIE: string
   INSTAGRAM_COOKIE: string
 
   // Worker KV namespaces

worker/src/providers/inoreader.ts

@@ -1,99 +1,20 @@
 import type { SubstatsResponse } from '@/types'
 import { providerErrorHandler } from '.'
 
-// https%3A%2F%2Fnnw.ranchero.com%2Ffeed.xml
-type InoreaderRawResponseOnSuccess = { xjxobj: Array<Record<string, string>> }
-
-async function getInoreaderCookie(env: Env): Promise<string> {
-  const { INOREADER_EMAIL, INOREADER_PASSWORD } = env
-  const { KV_COOKIES } = env
-
-  const cookie = await KV_COOKIES.get('inoreader')
-  if (cookie) {
-    return cookie
-  }
-
-  // If Inoreader's cookie is not set, we need to login and set it with an
-  // expiration inside the worker's KV storage.
-  const body = new URLSearchParams({
-    warp_action: 'login',
-    username: INOREADER_EMAIL,
-    password: INOREADER_PASSWORD,
-    remember_me: 'on',
-  })
-
-  // TODO: this is not working yet, fetch here does not return the required
-  // 'Set-Cookie' header, it somehow stores the cookie somewhere and returns a
-  // header without any one of the required cookies. (We at least require the
-  // ssid)
-  const resp = await fetch('https://www.inoreader.com/', {
-    headers: {
-      'Content-Type': 'application/x-www-form-urlencoded',
-    },
-    body: body,
-    method: 'POST',
-  })
-
-  const freshCookie = resp.headers.get('set-cookie')
-  console.log(freshCookie)
-
-  if (freshCookie) {
-    // const expires = freshCookie.match(/expires=([^;]+);/)?.[1] ?? ''
-    await KV_COOKIES.put('inoreader', freshCookie)
-    return freshCookie
-  }
-  return ''
-}
-
 export default async function inoreaderProvider(
   key: string,
-  env: Env,
 ): Promise<SubstatsResponse> {
-  // This route uses Inoreader's search API, which is a POST request with an
-  // x-www-form-urlencoded body containing the search query. It returns a JSON
-  // object with an 'xjxobj' key, which contains raw HTML. We can extract the
-  // feed's follower count from this raw HTML.
-  const requestBody = new URLSearchParams()
-  requestBody.append('xjxfun', 'build_searcher_content')
-  requestBody.append('xjxr', '1645356116453')
-  requestBody.append(
-    'xjxargs[]',
-    `{"tab":"feeds","term":"${decodeURIComponent(key)}","offset":0}`,
-  )
-
-  // Attempt to get the Inoreader cookie from either env or KV storage.
-  const inoreaderCookie =
-    env.INOREADER_COOKIE ?? (await getInoreaderCookie(env))
-
   try {
-    const resp = await fetch('https://www.inoreader.com/', {
+    const resp = await fetch(`https://www.innoreader.com/feed/${encodeURIComponent(key)}`, {
       headers: {
-        'content-type': 'application/x-www-form-urlencoded',
-        cookie: inoreaderCookie ?? '',
+        'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36'
       },
-      body: requestBody,
-      method: 'POST',
       cf: { cacheEverything: true },
     })
 
-    const data = await resp.json<InoreaderRawResponseOnSuccess>()
-
-    // We are looking for the returned element with the 'search_content' prop as
-    // 'id' and 'innerHTML' prop as 'prop', this object should contain 'data' as
-    // an HTML string for us to parse. A sample response:
-    // {
-    //   xjxobj: [
-    //     { cmd: 'js', data: '...'},
-    //     { cmd: 'jc', func: 'xxx', data: '...'},
-    //     { cmd: 'as', id: 'search_content', prop: 'innerHTML', data: '...' },
-    //     { /* ... */ },
-    //   ]
-    // }
-    const predicate = (x: Record<string, string>) =>
-      'id' in x && x?.id === 'search_content' && 'data' in x
-    const followerHTML = data.xjxobj.find(predicate)?.data
+    const followerHTML = await resp.text()
 
-    if (!followerHTML) {
+    if (!followerHTML || /Page not found/.test(followerHTML)) {
       throw new Error('Feed not found on Inoreader')
     }