feat(js-sdk): add crawlUrlAndWatch

apps/js-sdk/example.js

@@ -29,5 +29,21 @@ if (job.data) {
   console.log(job.data[0].markdown);
 }
 
+// Map a website:
 const mapResult = await app.map('https://firecrawl.dev');
 console.log(mapResult)
+
+// Crawl a website with WebSockets:
+const watch = await app.crawlUrlAndWatch('mendable.ai', { excludePaths: ['blog/*'], limit: 5});
+
+watch.addEventListener("document", doc => {
+  console.log("DOC", doc.detail);
+});
+
+watch.addEventListener("error", err => {
+  console.error("ERR", err.detail.error);
+});
+
+watch.addEventListener("done", state => {
+  console.log("DONE", state.detail.status);
+});

apps/js-sdk/example.ts

@@ -32,8 +32,24 @@ const main = async () => {
     console.log(checkStatus.data[0].markdown);
   }
 
+  // Map a website:
   const mapResult = await app.mapUrl('https://firecrawl.dev');
   console.log(mapResult)
+
+  // Crawl a website with WebSockets:
+  const watch = await app.crawlUrlAndWatch('mendable.ai', { excludePaths: ['blog/*'], limit: 5});
+
+  watch.addEventListener("document", doc => {
+    console.log("DOC", doc.detail);
+  });
+
+  watch.addEventListener("error", err => {
+    console.error("ERR", err.detail.error);
+  });
+
+  watch.addEventListener("done", state => {
+    console.log("DONE", state.detail.status);
+  });
 }
 
 main()

apps/js-sdk/firecrawl/package-lock.json

@@ -1,16 +1,18 @@
 {
   "name": "@mendable/firecrawl-js",
-  "version": "0.0.36",
+  "version": "1.0.3",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "@mendable/firecrawl-js",
-      "version": "0.0.36",
+      "version": "1.0.3",
       "license": "MIT",
       "dependencies": {
         "axios": "^1.6.8",
         "dotenv": "^16.4.5",
+        "isows": "^1.0.4",
+        "typescript-event-target": "^1.1.1",
         "uuid": "^9.0.1",
         "zod": "^3.23.8",
         "zod-to-json-schema": "^3.23.0"
@@ -2137,6 +2139,20 @@
       "integrity": "sha512-RHxMLp9lnKHGHRng9QFhRCMbYAcVpn69smSGcq3f36xjgVVWThj4qqLbTLlq7Ssj8B+fIQ1EuCEGI2lKsyQeIw==",
       "dev": true
     },
+    "node_modules/isows": {
+      "version": "1.0.4",
+      "resolved": "https://registry.npmjs.org/isows/-/isows-1.0.4.tgz",
+      "integrity": "sha512-hEzjY+x9u9hPmBom9IIAqdJCwNLax+xrPb51vEPpERoFlIxgmZcHzsT5jKG06nvInKOBGvReAVz80Umed5CczQ==",
+      "funding": [
+        {
+          "type": "github",
+          "url": "https://github.com/sponsors/wagmi-dev"
+        }
+      ],
+      "peerDependencies": {
+        "ws": "*"
+      }
+    },
     "node_modules/istanbul-lib-coverage": {
       "version": "3.2.2",
       "resolved": "https://registry.npmjs.org/istanbul-lib-coverage/-/istanbul-lib-coverage-3.2.2.tgz",
@@ -3733,6 +3749,11 @@
         "node": ">=14.17"
       }
     },
+    "node_modules/typescript-event-target": {
+      "version": "1.1.1",
+      "resolved": "https://registry.npmjs.org/typescript-event-target/-/typescript-event-target-1.1.1.tgz",
+      "integrity": "sha512-dFSOFBKV6uwaloBCCUhxlD3Pr/P1a/tJdcmPrTXCHlEFD3faj0mztjcGn6VBAhQ0/Bdy8K3VWrrqwbt/ffsYsg=="
+    },
     "node_modules/undici-types": {
       "version": "5.26.5",
       "resolved": "https://registry.npmjs.org/undici-types/-/undici-types-5.26.5.tgz",
@@ -3855,6 +3876,27 @@
         "node": "^12.13.0 || ^14.15.0 || >=16.0.0"
       }
     },
+    "node_modules/ws": {
+      "version": "8.18.0",
+      "resolved": "https://registry.npmjs.org/ws/-/ws-8.18.0.tgz",
+      "integrity": "sha512-8VbfWfHLbbwu3+N6OKsOMpBdT4kXPDDB9cJk2bJ6mh9ucxdlnNvH1e+roYkKmN9Nxw2yjz7VzeO9oOz2zJ04Pw==",
+      "peer": true,
+      "engines": {
+        "node": ">=10.0.0"
+      },
+      "peerDependencies": {
+        "bufferutil": "^4.0.1",
+        "utf-8-validate": ">=5.0.2"
+      },
+      "peerDependenciesMeta": {
+        "bufferutil": {
+          "optional": true
+        },
+        "utf-8-validate": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/y18n": {
       "version": "5.0.8",
       "resolved": "https://registry.npmjs.org/y18n/-/y18n-5.0.8.tgz",

apps/js-sdk/firecrawl/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mendable/firecrawl-js",
-  "version": "1.0.3",
+  "version": "1.0.4",
   "description": "JavaScript SDK for Firecrawl API",
   "main": "build/cjs/index.js",
   "types": "types/index.d.ts",
@@ -30,6 +30,8 @@
   "dependencies": {
     "axios": "^1.6.8",
     "dotenv": "^16.4.5",
+    "isows": "^1.0.4",
+    "typescript-event-target": "^1.1.1",
     "uuid": "^9.0.1",
     "zod": "^3.23.8",
     "zod-to-json-schema": "^3.23.0"

apps/js-sdk/firecrawl/src/index.ts

@@ -1,6 +1,8 @@
 import axios, { AxiosResponse, AxiosRequestHeaders } from "axios";
 import { z } from "zod";
 import { zodToJsonSchema } from "zod-to-json-schema";
+import { WebSocket } from "isows";
+import { TypedEventTarget } from "typescript-event-target";
 
 /**
  * Configuration interface for FirecrawlApp.
@@ -315,8 +317,8 @@ export interface SearchResponseV0 {
  * Provides methods for scraping, searching, crawling, and mapping web content.
  */
 export default class FirecrawlApp<T extends "v0" | "v1"> {
-  private apiKey: string;
-  private apiUrl: string;
+  public apiKey: string;
+  public apiUrl: string;
   public version: T;
 
   /**
@@ -561,6 +563,21 @@ export default class FirecrawlApp<T extends "v0" | "v1"> {
       } as this['version'] extends 'v0' ? CrawlStatusResponseV0 : CrawlStatusResponse);
   }
 
+  async crawlUrlAndWatch(
+    url: string,
+    params?: this['version'] extends 'v0' ? CrawlParamsV0 : CrawlParams,
+    idempotencyKey?: string,
+  ) {
+    if (this.version === 'v0') {
+      throw new Error("crawlUrlAndWatch is only available on v1");
+    }
+
+    const crawl = await this.crawlUrl(url, params, false, 0, idempotencyKey);
+    const id = this.version === 'v0' ? (crawl as CrawlResponseV0).jobId : (crawl as CrawlResponse).id;
+
+    return new CrawlWatcher(id as string, this as FirecrawlApp<"v1">);
+  }
+
   async mapUrl(url: string, params?: MapParams): Promise<MapResponse> {
     if (this.version == 'v0') {
       throw new Error("Map is not supported in v0");
@@ -696,3 +713,111 @@ export default class FirecrawlApp<T extends "v0" | "v1"> {
     }
   }
 }
+
+interface CrawlWatcherEvents {
+  document: CustomEvent<FirecrawlDocument>,
+  done: CustomEvent<{
+    status: CrawlStatusResponse["status"];
+    data: FirecrawlDocument[];
+  }>,
+  error: CustomEvent<{
+    status: CrawlStatusResponse["status"],
+    data: FirecrawlDocument[],
+    error: string,
+  }>,
+}
+
+export class CrawlWatcher extends TypedEventTarget<CrawlWatcherEvents> {
+  private ws: WebSocket;
+  public data: FirecrawlDocument[];
+  public status: CrawlStatusResponse["status"];
+
+  constructor(id: string, app: FirecrawlApp<"v1">) {
+    super();
+    this.ws = new WebSocket(`${app.apiUrl}/v1/crawl/${id}`, app.apiKey);
+    this.status = "scraping";
+    this.data = [];
+
+    type ErrorMessage = {
+      type: "error",
+      error: string,
+    }
+    
+    type CatchupMessage = {
+      type: "catchup",
+      data: CrawlStatusResponse,
+    }
+    
+    type DocumentMessage = {
+      type: "document",
+      data: FirecrawlDocument,
+    }
+    
+    type DoneMessage = { type: "done" }
+    
+    type Message = ErrorMessage | CatchupMessage | DoneMessage | DocumentMessage;
+
+    const messageHandler = (msg: Message) => {
+      if (msg.type === "done") {
+        this.status = "completed";
+        this.dispatchTypedEvent("done", new CustomEvent("done", {
+          detail: {
+            status: this.status,
+            data: this.data,
+          },
+        }));
+      } else if (msg.type === "error") {
+        this.status = "failed";
+        this.dispatchTypedEvent("error", new CustomEvent("error", {
+          detail: {
+            status: this.status,
+            data: this.data,
+            error: msg.error,
+          },
+        }));
+      } else if (msg.type === "catchup") {
+        this.status = msg.data.status;
+        this.data.push(...(msg.data.data ?? []));
+        for (const doc of this.data) {
+          this.dispatchTypedEvent("document", new CustomEvent("document", {
+            detail: doc,
+          }));
+        }
+      } else if (msg.type === "document") {
+        this.dispatchTypedEvent("document", new CustomEvent("document", {
+          detail: msg.data,
+        }));
+      }
+    }
+
+    this.ws.onmessage = ((ev: MessageEvent) => {
+      if (typeof ev.data !== "string") {
+        this.ws.close();
+        return;
+      }
+
+      const msg = JSON.parse(ev.data) as Message;
+      messageHandler(msg);
+    }).bind(this);
+
+    this.ws.onclose = ((ev: CloseEvent) => {
+      const msg = JSON.parse(ev.reason) as Message;
+      messageHandler(msg);
+    }).bind(this);
+
+    this.ws.onerror = ((_: Event) => {
+      this.status = "failed"
+      this.dispatchTypedEvent("error", new CustomEvent("error", {
+        detail: {
+          status: this.status,
+          data: this.data,
+          error: "WebSocket error",
+        },
+      }));
+    }).bind(this);
+  }
+
+  close() {
+    this.ws.close();
+  }
+}