Merge pull request #1 from movie-web/dev

Add initial POC of Uber Proxy

README.md

@@ -1,2 +1,5 @@
 # provider-api
-Our provider package exposed as a HTTP API
+Our provider package exposed as a HTTP API. This is not the recommended way to deploy movie-web or use our providers. Instead use the provider package or simple-proxy.
+
+> [!IMPORTANT]
+> This only works on Cloudflare due to CloudFlare specific logic and build processes

package.json

@@ -2,30 +2,30 @@
   "name": "providers-api",
   "version": "1.0.0",
   "private": true,
+  "type": "module",
   "scripts": {
-    "prepare": "nitropack prepare",
-    "dev": "nitropack dev",
-    "build": "nitropack build",
-    "build:cloudflare": "NITRO_PRESET=cloudflare npm run build",
-    "build:aws": "NITRO_PRESET=aws_lambda npm run build",
-    "build:node": "NITRO_PRESET=node-server npm run build",
-    "start": "node .output/server/index.mjs",
-    "lint": "eslint --ext .ts src/",
-    "lint:fix": "eslint --fix --ext .ts src/",
+    "build": "wrangler deploy --dry-run --outdir dist",
+    "dev": "wrangler dev src/index.ts",
+    "deploy": "wrangler deploy --minify src/index.ts",
     "preinstall": "npx only-allow pnpm"
   },
   "dependencies": {
     "@movie-web/providers": "^1.1.5",
-    "h3": "^1.9.0",
-    "nitropack": "latest"
+    "@tsndr/cloudflare-worker-jwt": "^2.3.2",
+    "@types/jsonwebtoken": "^9.0.5",
+    "esbuild": "^0.19.10",
+    "hono": "^3.11.8",
+    "zod": "^3.22.4"
   },
   "devDependencies": {
+    "@cloudflare/workers-types": "^4.20231121.0",
     "@typescript-eslint/eslint-plugin": "^6.14.0",
     "@typescript-eslint/parser": "^6.14.0",
     "eslint": "^8.56.0",
     "eslint-config-airbnb-base": "^15.0.0",
     "eslint-config-prettier": "^9.1.0",
     "eslint-import-resolver-typescript": "^3.6.1",
-    "eslint-plugin-prettier": "^5.0.1"
+    "eslint-plugin-prettier": "^5.0.1",
+    "wrangler": "^3.21.0"
   }
 }

src/index.ts

@@ -0,0 +1,272 @@
+import { Context, Env, Hono } from 'hono';
+import { streamSSE } from 'hono/streaming';
+import { cors } from 'hono/cors';
+import {
+  ScrapeMedia,
+  makeProviders,
+  makeStandardFetcher,
+  targets,
+} from '@movie-web/providers';
+import { ZodError, z } from 'zod';
+import { embedSchema, scrapeAllSchema, sourceSchema } from '@/schema';
+import { validateTurnstile } from '@/turnstile';
+
+// hono doesn't export this type, so we retrieve it from a function
+type SSEStreamingApi = Parameters<Parameters<typeof streamSSE>['1']>['0'];
+
+const fetcher = makeStandardFetcher(fetch);
+
+const providers = makeProviders({
+  fetcher,
+  target: targets.BROWSER,
+});
+
+const app = new Hono();
+
+function isTurnstileEnabled(context: Context<Env>) {
+  return context.env?.TURNSTILE_ENABLED === "true"
+}
+
+app.use('*', (context, next) => {
+  const allowedCorsHosts = ((context.env?.CORS_ALLOWED as string) ?? '').split(
+    ',',
+  );
+
+  return cors({
+    origin: (origin) => {
+      try {
+        const hostname = new URL(origin).hostname;
+        if (allowedCorsHosts.includes(hostname)) {
+          return origin;
+        }
+        return '';
+      } catch (_) {
+        // If the Origin URL is not valid, return empty allowed origin
+        return '';
+      }
+    },
+  })(context, next);
+});
+
+let eventId = 0;
+async function writeSSEEvent(
+  stream: SSEStreamingApi,
+  event: string,
+  data: any | undefined,
+) {
+  return await stream.writeSSE({
+    event,
+    data: data ? JSON.stringify(data) : '',
+    id: String(eventId++),
+  });
+}
+
+app.get('/scrape', async (context) => {
+  const queryParams = context.req.query();
+
+  let jwtResponse: string | undefined = undefined;
+
+  if (isTurnstileEnabled(context)) {
+    const turnstileResponse = await validateTurnstile(context);
+
+    if (!turnstileResponse.success) {
+      context.status(401);
+      return context.text(
+        `Turnstile invalid, error codes: ${turnstileResponse.errorCodes.join(
+          ', ',
+        )}`,
+      );
+    }
+
+    jwtResponse = turnstileResponse.jwtToken;
+  }
+
+  let media: ScrapeMedia;
+  try {
+    media = scrapeAllSchema.parse(queryParams);
+  } catch (e) {
+    if (e instanceof ZodError) {
+      context.status(400);
+      return context.json(e.format());
+    }
+    context.status(500);
+    return context.text('An error has occurred!');
+  }
+
+  return streamSSE(context, async (stream) => {
+    if (jwtResponse) {
+      await writeSSEEvent(stream, 'token', jwtResponse);
+    }
+
+    try {
+      const output = await providers.runAll({
+        media,
+        events: {
+          discoverEmbeds(evt) {
+            writeSSEEvent(stream, 'discoverEmbeds', evt);
+          },
+          init(evt) {
+            writeSSEEvent(stream, 'init', evt);
+          },
+          start(evt) {
+            writeSSEEvent(stream, 'start', evt);
+          },
+          update(evt) {
+            writeSSEEvent(stream, 'update', evt);
+          },
+        },
+      });
+
+      if (output) {
+        await writeSSEEvent(stream, 'completed', output);
+        return await stream.close();
+      }
+
+      await writeSSEEvent(stream, 'noOutput', '');
+      return await stream.close();
+    } catch (e: any) {
+      await writeSSEEvent(stream, 'error', {
+        name: e.name,
+        message: e.message,
+        stack: e.stack,
+      });
+      return await stream.close();
+    }
+  });
+});
+
+app.get('/scrape/embed', async (context) => {
+  const queryParams = context.req.query();
+
+  let jwtResponse: string | undefined = undefined;
+
+  if (isTurnstileEnabled(context)) {
+    const turnstileResponse = await validateTurnstile(context);
+
+    if (!turnstileResponse.success) {
+      context.status(401);
+      return context.text(
+        `Turnstile invalid, error codes: ${turnstileResponse.errorCodes.join(
+          ', ',
+        )}`,
+      );
+    }
+
+    jwtResponse = turnstileResponse.jwtToken;
+  }
+
+  let embedInput: z.infer<typeof embedSchema>;
+  try {
+    embedInput = embedSchema.parse(queryParams);
+  } catch (e) {
+    if (e instanceof ZodError) {
+      context.status(400);
+      return context.json(e.format());
+    }
+    context.status(500);
+    return context.text('An error has occurred!');
+  }
+
+  return streamSSE(context, async (stream) => {
+    if (jwtResponse) {
+      await writeSSEEvent(stream, 'token', jwtResponse);
+    }
+    try {
+      const output = await providers.runEmbedScraper({
+        id: embedInput.id,
+        url: embedInput.url,
+        events: {
+          update(evt) {
+            writeSSEEvent(stream, 'update', evt);
+          },
+        },
+      });
+
+      if (output) {
+        await writeSSEEvent(stream, 'completed', output);
+        return await stream.close();
+      }
+
+      await writeSSEEvent(stream, 'noOutput', '');
+      return await stream.close();
+    } catch (e: any) {
+      await writeSSEEvent(stream, 'error', {
+        name: e.name,
+        message: e.message,
+        stack: e.stack,
+      });
+      return await stream.close();
+    }
+  });
+});
+
+app.get('/scrape/source', async (context) => {
+  const queryParams = context.req.query();
+
+  let jwtResponse: string | undefined = undefined;
+
+  if (isTurnstileEnabled(context)) {
+    const turnstileResponse = await validateTurnstile(context);
+
+    if (!turnstileResponse.success) {
+      context.status(401);
+      return context.text(
+        `Turnstile invalid, error codes: ${turnstileResponse.errorCodes.join(
+          ', ',
+        )}`,
+      );
+    }
+
+    jwtResponse = turnstileResponse.jwtToken;
+  }
+
+  let sourceInput: z.infer<typeof sourceSchema>;
+  try {
+    sourceInput = sourceSchema.parse(queryParams);
+  } catch (e) {
+    if (e instanceof ZodError) {
+      context.status(400);
+      return context.json(e.format());
+    }
+    context.status(500);
+    return context.text('An error has occurred!');
+  }
+
+  return streamSSE(context, async (stream) => {
+    if (jwtResponse) {
+      await writeSSEEvent(stream, 'token', jwtResponse);
+    }
+    try {
+      const output = await providers.runSourceScraper({
+        id: sourceInput.id,
+        media: sourceInput,
+        events: {
+          update(evt) {
+            writeSSEEvent(stream, 'update', evt);
+          },
+        },
+      });
+
+      if (output) {
+        await writeSSEEvent(stream, 'completed', output);
+        return await stream.close();
+      }
+
+      await writeSSEEvent(stream, 'noOutput', '');
+      return await stream.close();
+    } catch (e: any) {
+      await writeSSEEvent(stream, 'error', {
+        name: e.name,
+        message: e.message,
+        stack: e.stack,
+      });
+      return await stream.close();
+    }
+  });
+});
+
+app.get('/metadata', async (context) => {
+  return context.json([providers.listEmbeds(), providers.listSources()]);
+});
+
+export default app;

src/routes/index.ts

@@ -1,39 +0,0 @@
-import { getBodyBuffer } from '@/utils/body';
-import {
-  getProxyHeaders,
-  getAfterResponseHeaders,
-  cleanupHeadersBeforeProxy,
-} from '@/utils/headers';
-
-export default defineEventHandler(async (event) => {
-  // handle cors, if applicable
-  if (isPreflightRequest(event)) return handleCors(event, {});
-
-  // parse destination URL
-  const destination = getQuery<{ destination?: string }>(event).destination;
-  if (!destination)
-    return await sendJson({
-      event,
-      status: 400,
-      data: {
-        error: 'destination query parameter invalid',
-      },
-    });
-
-  // read body
-  const body = await getBodyBuffer(event);
-
-  // proxy
-  cleanupHeadersBeforeProxy(event);
-  await proxyRequest(event, destination, {
-    fetchOptions: {
-      redirect: 'follow',
-      headers: getProxyHeaders(event.headers),
-      body,
-    },
-    onResponse(outputEvent, response) {
-      const headers = getAfterResponseHeaders(response.headers, response.url);
-      setResponseHeaders(outputEvent, headers);
-    },
-  });
-});

src/schema.ts

@@ -0,0 +1,52 @@
+import { z } from 'zod';
+
+export const tmdbIdSchema = z.string().regex(/^\d+$/);
+
+export const scrapeAllSchema = z
+  .discriminatedUnion('type', [
+    z.object({
+      type: z.literal('movie'),
+      title: z.string().min(1),
+      releaseYear: z.coerce.number().int().gt(0),
+      tmdbId: tmdbIdSchema,
+    }),
+    z.object({
+      type: z.literal('show'),
+      title: z.string().min(1),
+      releaseYear: z.coerce.number().int().gt(0),
+      tmdbId: tmdbIdSchema,
+      episodeNumber: z.coerce.number().int(),
+      episodeTmdbId: tmdbIdSchema,
+      seasonNumber: z.coerce.number().int(),
+      seasonTmdbId: tmdbIdSchema,
+    }),
+  ])
+  .transform((query) => {
+    if (query.type == 'movie') return query;
+
+    return {
+      type: query.type,
+      title: query.title,
+      releaseYear: query.releaseYear,
+      tmdbId: query.tmdbId,
+      episode: {
+        number: query.episodeNumber,
+        tmdbId: query.episodeTmdbId,
+      },
+      season: {
+        number: query.seasonNumber,
+        tmdbId: query.seasonTmdbId,
+      },
+    };
+  });
+
+export const embedSchema = z.object({
+  id: z.string(),
+  url: z.string(),
+});
+
+export const sourceSchema = scrapeAllSchema.and(
+  z.object({
+    id: z.string(),
+  }),
+);

src/turnstile.ts

@@ -0,0 +1,81 @@
+import { Context, Env } from 'hono';
+import jsonwebtoken from '@tsndr/cloudflare-worker-jwt';
+
+export async function validateTurnstile(context: Context<Env>) {
+  const turnstileSecret = context.env?.TURNSTILE_SECRET as string | undefined;
+  const jwtSecret = (context.env?.JWT_SECRET as string | undefined) ?? '';
+
+  const token = context.req.query('token') || '';
+
+  const ip = context.req.header('CF-Connecting-IP') || '';
+
+  if (token.startsWith('jwt|')) {
+    try {
+      const isValid = await jsonwebtoken.verify(
+        token.slice('jwt|'.length),
+        jwtSecret,
+        {
+          algorithm: 'HS256',
+        },
+      );
+
+      if (!isValid) {
+        return {
+          success: false,
+          errorCodes: ['jwt-invalid'],
+        };
+      }
+
+      const { payload } = jsonwebtoken.decode(token.slice('jwt|'.length));
+
+      if (!payload || payload.ip !== ip) {
+        return {
+          success: false,
+          errorCodes: ['jwt-ip-invalid'],
+        };
+      }
+
+      return {
+        success: true,
+        errorCodes: [],
+      };
+    } catch (e: any) {}
+  }
+
+  if (!token.startsWith('turnstile|')) {
+    return {
+      success: false,
+      errorCodes: ['invalid-token-type'],
+    };
+  }
+
+  const formData = new FormData();
+  formData.append('secret', turnstileSecret || '');
+  formData.append('response', token.slice('turnstile|'.length));
+  formData.append('remoteip', ip);
+
+  const url = 'https://challenges.cloudflare.com/turnstile/v0/siteverify';
+  const result = await fetch(url, {
+    body: formData,
+    method: 'POST',
+  });
+
+  const outcome = await result.json<any>();
+
+  let jwt: string | undefined = undefined;
+  if (outcome.success) {
+    jwt = await jsonwebtoken.sign(
+      {
+        ip,
+        exp: Math.floor(Date.now() / 1000) + 60 * 10, // 10 Minutes
+      },
+      jwtSecret,
+    );
+  }
+
+  return {
+    success: outcome.success as boolean,
+    errorCodes: outcome['error-codes'] as string[],
+    jwtToken: jwt,
+  };
+}

src/utils/body.ts

@@ -1,13 +0,0 @@
-import { H3Event } from 'h3';
-
-export function hasBody(event: H3Event) {
-  const method = event.method.toUpperCase();
-  return ['PUT', 'POST', 'PATCH', 'DELETE'].includes(method);
-}
-
-export async function getBodyBuffer(
-  event: H3Event,
-): Promise<Buffer | undefined> {
-  if (!hasBody(event)) return;
-  return await readRawBody(event, false);
-}

src/utils/headers.ts

@@ -1,73 +0,0 @@
-import { H3Event } from 'h3';
-
-const blacklistedHeaders = [
-  'cf-connecting-ip',
-  'cf-worker',
-  'cf-ray',
-  'cf-visitor',
-  'cf-ew-via',
-  'x-forwarded-for',
-  'x-forwarded-host',
-  'x-forwarded-proto',
-  'forwarded',
-  'x-real-ip',
-];
-
-function copyHeader(
-  headers: Headers,
-  outputHeaders: Headers,
-  inputKey: string,
-  outputKey: string,
-) {
-  if (headers.has(inputKey))
-    outputHeaders.set(outputKey, headers.get(inputKey) ?? '');
-}
-
-export function getProxyHeaders(headers: Headers): Headers {
-  const output = new Headers();
-
-  const headerMap: Record<string, string> = {
-    'X-Cookie': 'Cookie',
-    'X-Referer': 'Referer',
-    'X-Origin': 'Origin',
-  };
-  Object.entries(headerMap).forEach((entry) => {
-    copyHeader(headers, output, entry[0], entry[1]);
-  });
-
-  output.set(
-    'User-Agent',
-    'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:93.0) Gecko/20100101 Firefox/93.0',
-  );
-
-  return output;
-}
-
-export function getAfterResponseHeaders(
-  headers: Headers,
-  finalUrl: string,
-): Record<string, string> {
-  const output: Record<string, string> = {};
-
-  if (headers.has('Set-Cookie'))
-    output['X-Set-Cookie'] = headers.get('Set-Cookie') ?? '';
-
-  return {
-    'Access-Control-Allow-Origin': '*',
-    'Access-Control-Expose-Headers': '*',
-    Vary: 'Origin',
-    'X-Final-Destination': finalUrl,
-  };
-}
-
-export function removeHeadersFromEvent(event: H3Event, key: string) {
-  const normalizedKey = key.toLowerCase();
-  if (event.node.req.headers[normalizedKey])
-    delete event.node.req.headers[normalizedKey];
-}
-
-export function cleanupHeadersBeforeProxy(event: H3Event) {
-  blacklistedHeaders.forEach((key) => {
-    removeHeadersFromEvent(event, key);
-  });
-}

src/utils/sending.ts

@@ -1,10 +0,0 @@
-import { H3Event, EventHandlerRequest } from 'h3';
-
-export async function sendJson(ops: {
-  event: H3Event<EventHandlerRequest>;
-  data: Record<string, any>;
-  status?: number;
-}) {
-  setResponseStatus(ops.event, ops.status ?? 200);
-  await send(ops.event, JSON.stringify(ops.data, null, 2), 'application/json');
-}

tsconfig.json

@@ -0,0 +1,25 @@
+{
+  "compilerOptions": {
+    "target": "ESNext",
+    "lib": ["dom", "dom.iterable", "esnext"],
+    "allowJs": true,
+    "skipLibCheck": true,
+    "esModuleInterop": true,
+    "allowSyntheticDefaultImports": true,
+    "strict": true,
+    "forceConsistentCasingInFileNames": true,
+    "noFallthroughCasesInSwitch": true,
+    "module": "ESNext",
+    "moduleResolution": "node",
+    "resolveJsonModule": true,
+    "isolatedModules": true,
+    "noEmit": true,
+    "baseUrl": "./src",
+    "paths": {
+      "@/*": ["./*"]
+    },
+    "types": [
+      "@cloudflare/workers-types"
+    ],
+  },
+}

wrangler.toml

@@ -0,0 +1,4 @@
+name = "providers-api"
+main = "./src/index.ts"
+workers_dev = true
+compatibility_date = "2023-12-17"