firecrawl 1.18.1 → 1.18.2

package/dist/index.cjs

@@ -42,9 +42,11 @@ var import_isows = require("isows");
 var import_typescript_event_target = require("typescript-event-target");
 var FirecrawlError = class extends Error {
   statusCode;
-  constructor(message, statusCode) {
+  details;
+  constructor(message, statusCode, details) {
     super(message);
     this.statusCode = statusCode;
+    this.details = details;
   }
 };
 var FirecrawlApp = class {
@@ -91,6 +93,20 @@ var FirecrawlApp = class {
         }
       };
     }
+    if (jsonData?.jsonOptions?.schema) {
+      let schema = jsonData.jsonOptions.schema;
+      try {
+        schema = (0, import_zod_to_json_schema.zodToJsonSchema)(schema);
+      } catch (error) {
+      }
+      jsonData = {
+        ...jsonData,
+        jsonOptions: {
+          ...jsonData.jsonOptions,
+          schema
+        }
+      };
+    }
     try {
       const response = await import_axios.default.post(
         this.apiUrl + `/v1/scrape`,
@@ -245,16 +261,26 @@ var FirecrawlApp = class {
    * Checks the status of a crawl job using the Firecrawl API.
    * @param id - The ID of the crawl operation.
    * @param getAllData - Paginate through all the pages of documents, returning the full list of all documents. (default: `false`)
+   * @param nextURL - The `next` URL from the previous crawl status. Only required if you're not manually increasing `skip`. Only used when `getAllData = false`.
+   * @param skip - How many entries to skip to paginate. Only required if you're not providing `nextURL`. Only used when `getAllData = false`.
+   * @param limit - How many entries to return. Only used when `getAllData = false`.
    * @returns The response containing the job status.
    */
-  async checkCrawlStatus(id, getAllData = false) {
+  async checkCrawlStatus(id, getAllData = false, nextURL, skip, limit) {
     if (!id) {
       throw new FirecrawlError("No crawl ID provided", 400);
     }
     const headers = this.prepareHeaders();
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/crawl/${id}`);
+    if (skip !== void 0) {
+      targetURL.searchParams.set("skip", skip.toString());
+    }
+    if (limit !== void 0) {
+      targetURL.searchParams.set("limit", limit.toString());
+    }
     try {
       const response = await this.getRequest(
-        `${this.apiUrl}/v1/crawl/${id}`,
+        targetURL.href,
         headers
       );
       if (response.status === 200) {
@@ -279,6 +305,7 @@ var FirecrawlApp = class {
           total: response.data.total,
           completed: response.data.completed,
           creditsUsed: response.data.creditsUsed,
+          next: getAllData ? void 0 : response.data.next,
           expiresAt: new Date(response.data.expiresAt),
           data: allData
         };
@@ -301,6 +328,28 @@ var FirecrawlApp = class {
     }
     return { success: false, error: "Internal server error." };
   }
+  /**
+   * Returns information about crawl errors.
+   * @param id - The ID of the crawl operation.
+   * @returns Information about crawl errors.
+   */
+  async checkCrawlErrors(id) {
+    const headers = this.prepareHeaders();
+    try {
+      const response = await this.deleteRequest(
+        `${this.apiUrl}/v1/crawl/${id}/errors`,
+        headers
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else {
+        this.handleError(response, "check crawl errors");
+      }
+    } catch (error) {
+      throw new FirecrawlError(error.message, 500);
+    }
+    return { success: false, error: "Internal server error." };
+  }
   /**
    * Cancels a crawl job using the Firecrawl API.
    * @param id - The ID of the crawl operation.
@@ -389,6 +438,20 @@ var FirecrawlApp = class {
         }
       };
     }
+    if (jsonData?.jsonOptions?.schema) {
+      let schema = jsonData.jsonOptions.schema;
+      try {
+        schema = (0, import_zod_to_json_schema.zodToJsonSchema)(schema);
+      } catch (error) {
+      }
+      jsonData = {
+        ...jsonData,
+        jsonOptions: {
+          ...jsonData.jsonOptions,
+          schema
+        }
+      };
+    }
     try {
       const response = await this.postRequest(
         this.apiUrl + `/v1/batch/scrape`,
@@ -452,16 +515,26 @@ var FirecrawlApp = class {
    * Checks the status of a batch scrape job using the Firecrawl API.
    * @param id - The ID of the batch scrape operation.
    * @param getAllData - Paginate through all the pages of documents, returning the full list of all documents. (default: `false`)
+   * @param nextURL - The `next` URL from the previous batch scrape status. Only required if you're not manually increasing `skip`. Only used when `getAllData = false`.
+   * @param skip - How many entries to skip to paginate. Only used when `getAllData = false`.
+   * @param limit - How many entries to return. Only used when `getAllData = false`.
    * @returns The response containing the job status.
    */
-  async checkBatchScrapeStatus(id, getAllData = false) {
+  async checkBatchScrapeStatus(id, getAllData = false, nextURL, skip, limit) {
     if (!id) {
       throw new FirecrawlError("No batch scrape ID provided", 400);
     }
     const headers = this.prepareHeaders();
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/batch/scrape/${id}`);
+    if (skip !== void 0) {
+      targetURL.searchParams.set("skip", skip.toString());
+    }
+    if (limit !== void 0) {
+      targetURL.searchParams.set("limit", limit.toString());
+    }
     try {
       const response = await this.getRequest(
-        `${this.apiUrl}/v1/batch/scrape/${id}`,
+        targetURL.href,
         headers
       );
       if (response.status === 200) {
@@ -486,6 +559,7 @@ var FirecrawlApp = class {
           total: response.data.total,
           completed: response.data.completed,
           creditsUsed: response.data.creditsUsed,
+          next: getAllData ? void 0 : response.data.next,
           expiresAt: new Date(response.data.expiresAt),
           data: allData
         };
@@ -508,6 +582,28 @@ var FirecrawlApp = class {
     }
     return { success: false, error: "Internal server error." };
   }
+  /**
+   * Returns information about batch scrape errors.
+   * @param id - The ID of the batch scrape operation.
+   * @returns Information about batch scrape errors.
+   */
+  async checkBatchScrapeErrors(id) {
+    const headers = this.prepareHeaders();
+    try {
+      const response = await this.deleteRequest(
+        `${this.apiUrl}/v1/batch/scrape/${id}/errors`,
+        headers
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else {
+        this.handleError(response, "check batch scrape errors");
+      }
+    } catch (error) {
+      throw new FirecrawlError(error.message, 500);
+    }
+    return { success: false, error: "Internal server error." };
+  }
   /**
    * Extracts information from URLs using the Firecrawl API.
    * Currently in Beta. Expect breaking changes on future minor versions.
@@ -533,29 +629,99 @@ var FirecrawlApp = class {
     try {
       const response = await this.postRequest(
         this.apiUrl + `/v1/extract`,
-        { ...jsonData, schema: jsonSchema },
+        { ...jsonData, schema: jsonSchema, origin: params?.origin || "api-sdk" },
         headers
       );
       if (response.status === 200) {
-        const responseData = response.data;
-        if (responseData.success) {
-          return {
-            success: true,
-            data: responseData.data,
-            warning: responseData.warning,
-            error: responseData.error
-          };
-        } else {
-          throw new FirecrawlError(`Failed to scrape URL. Error: ${responseData.error}`, response.status);
-        }
+        const jobId = response.data.id;
+        let extractStatus;
+        do {
+          const statusResponse = await this.getRequest(
+            `${this.apiUrl}/v1/extract/${jobId}`,
+            headers
+          );
+          extractStatus = statusResponse.data;
+          if (extractStatus.status === "completed") {
+            if (extractStatus.success) {
+              return {
+                success: true,
+                data: extractStatus.data,
+                warning: extractStatus.warning,
+                error: extractStatus.error,
+                sources: extractStatus?.sources || void 0
+              };
+            } else {
+              throw new FirecrawlError(`Failed to extract data. Error: ${extractStatus.error}`, statusResponse.status);
+            }
+          } else if (extractStatus.status === "failed" || extractStatus.status === "cancelled") {
+            throw new FirecrawlError(`Extract job ${extractStatus.status}. Error: ${extractStatus.error}`, statusResponse.status);
+          }
+          await new Promise((resolve) => setTimeout(resolve, 1e3));
+        } while (extractStatus.status !== "completed");
       } else {
         this.handleError(response, "extract");
       }
     } catch (error) {
-      throw new FirecrawlError(error.message, 500);
+      throw new FirecrawlError(error.message, 500, error.response?.data?.details);
+    }
+    return { success: false, error: "Internal server error." };
+  }
+  /**
+   * Initiates an asynchronous extract job for a URL using the Firecrawl API.
+   * @param url - The URL to extract data from.
+   * @param params - Additional parameters for the extract request.
+   * @param idempotencyKey - Optional idempotency key for the request.
+   * @returns The response from the extract operation.
+   */
+  async asyncExtract(urls, params, idempotencyKey) {
+    const headers = this.prepareHeaders(idempotencyKey);
+    let jsonData = { urls, ...params };
+    let jsonSchema;
+    try {
+      if (params?.schema instanceof zt.ZodType) {
+        jsonSchema = (0, import_zod_to_json_schema.zodToJsonSchema)(params.schema);
+      } else {
+        jsonSchema = params?.schema;
+      }
+    } catch (error) {
+      throw new FirecrawlError("Invalid schema. Schema must be either a valid Zod schema or JSON schema object.", 400);
+    }
+    try {
+      const response = await this.postRequest(
+        this.apiUrl + `/v1/extract`,
+        { ...jsonData, schema: jsonSchema },
+        headers
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else {
+        this.handleError(response, "start extract job");
+      }
+    } catch (error) {
+      throw new FirecrawlError(error.message, 500, error.response?.data?.details);
     }
     return { success: false, error: "Internal server error." };
   }
+  /**
+   * Retrieves the status of an extract job.
+   * @param jobId - The ID of the extract job.
+   * @returns The status of the extract job.
+   */
+  async getExtractStatus(jobId) {
+    try {
+      const response = await this.getRequest(
+        `${this.apiUrl}/v1/extract/${jobId}`,
+        this.prepareHeaders()
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else {
+        this.handleError(response, "get extract status");
+      }
+    } catch (error) {
+      throw new FirecrawlError(error.message, 500);
+    }
+  }
   /**
    * Prepares the headers for an API request.
    * @param idempotencyKey - Optional key to ensure idempotency.
@@ -670,11 +836,13 @@ var FirecrawlApp = class {
    * @param {string} action - The action being performed when the error occurred.
    */
   handleError(response, action) {
-    if ([402, 408, 409, 500].includes(response.status)) {
+    if ([400, 402, 408, 409, 500].includes(response.status)) {
       const errorMessage = response.data.error || "Unknown error occurred";
+      const details = response.data.details ? ` - ${JSON.stringify(response.data.details)}` : "";
       throw new FirecrawlError(
-        `Failed to ${action}. Status code: ${response.status}. Error: ${errorMessage}`,
-        response.status
+        `Failed to ${action}. Status code: ${response.status}. Error: ${errorMessage}${details}`,
+        response.status,
+        response?.data?.details
       );
     } else {
       throw new FirecrawlError(
@@ -683,6 +851,198 @@ var FirecrawlApp = class {
       );
     }
   }
+  /**
+   * Initiates a deep research operation on a given topic and polls until completion.
+   * @param params - Parameters for the deep research operation.
+   * @returns The final research results.
+   */
+  async __deepResearch(topic, params) {
+    try {
+      const response = await this.__asyncDeepResearch(topic, params);
+      if (!response.success || "error" in response) {
+        return { success: false, error: "error" in response ? response.error : "Unknown error" };
+      }
+      if (!response.id) {
+        throw new FirecrawlError(`Failed to start research. No job ID returned.`, 500);
+      }
+      const jobId = response.id;
+      let researchStatus;
+      while (true) {
+        researchStatus = await this.__checkDeepResearchStatus(jobId);
+        if ("error" in researchStatus && !researchStatus.success) {
+          return researchStatus;
+        }
+        if (researchStatus.status === "completed") {
+          return researchStatus;
+        }
+        if (researchStatus.status === "failed") {
+          throw new FirecrawlError(
+            `Research job ${researchStatus.status}. Error: ${researchStatus.error}`,
+            500
+          );
+        }
+        if (researchStatus.status !== "processing") {
+          break;
+        }
+        await new Promise((resolve) => setTimeout(resolve, 2e3));
+      }
+      return { success: false, error: "Research job terminated unexpectedly" };
+    } catch (error) {
+      throw new FirecrawlError(error.message, 500, error.response?.data?.details);
+    }
+  }
+  /**
+   * Initiates a deep research operation on a given topic without polling.
+   * @param params - Parameters for the deep research operation.
+   * @returns The response containing the research job ID.
+   */
+  async __asyncDeepResearch(topic, params) {
+    const headers = this.prepareHeaders();
+    try {
+      const response = await this.postRequest(
+        `${this.apiUrl}/v1/deep-research`,
+        { topic, ...params },
+        headers
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else {
+        this.handleError(response, "start deep research");
+      }
+    } catch (error) {
+      if (error.response?.data?.error) {
+        throw new FirecrawlError(`Request failed with status code ${error.response.status}. Error: ${error.response.data.error} ${error.response.data.details ? ` - ${JSON.stringify(error.response.data.details)}` : ""}`, error.response.status);
+      } else {
+        throw new FirecrawlError(error.message, 500);
+      }
+    }
+    return { success: false, error: "Internal server error." };
+  }
+  /**
+   * Checks the status of a deep research operation.
+   * @param id - The ID of the deep research operation.
+   * @returns The current status and results of the research operation.
+   */
+  async __checkDeepResearchStatus(id) {
+    const headers = this.prepareHeaders();
+    try {
+      const response = await this.getRequest(
+        `${this.apiUrl}/v1/deep-research/${id}`,
+        headers
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else if (response.status === 404) {
+        throw new FirecrawlError("Deep research job not found", 404);
+      } else {
+        this.handleError(response, "check deep research status");
+      }
+    } catch (error) {
+      if (error.response?.data?.error) {
+        throw new FirecrawlError(`Request failed with status code ${error.response.status}. Error: ${error.response.data.error} ${error.response.data.details ? ` - ${JSON.stringify(error.response.data.details)}` : ""}`, error.response.status);
+      } else {
+        throw new FirecrawlError(error.message, 500);
+      }
+    }
+    return { success: false, error: "Internal server error." };
+  }
+  /**
+   * Generates LLMs.txt for a given URL and polls until completion.
+   * @param url - The URL to generate LLMs.txt from.
+   * @param params - Parameters for the LLMs.txt generation operation.
+   * @returns The final generation results.
+   */
+  async generateLLMsText(url, params) {
+    try {
+      const response = await this.asyncGenerateLLMsText(url, params);
+      if (!response.success || "error" in response) {
+        return { success: false, error: "error" in response ? response.error : "Unknown error" };
+      }
+      if (!response.id) {
+        throw new FirecrawlError(`Failed to start LLMs.txt generation. No job ID returned.`, 500);
+      }
+      const jobId = response.id;
+      let generationStatus;
+      while (true) {
+        generationStatus = await this.checkGenerateLLMsTextStatus(jobId);
+        if ("error" in generationStatus && !generationStatus.success) {
+          return generationStatus;
+        }
+        if (generationStatus.status === "completed") {
+          return generationStatus;
+        }
+        if (generationStatus.status === "failed") {
+          throw new FirecrawlError(
+            `LLMs.txt generation job ${generationStatus.status}. Error: ${generationStatus.error}`,
+            500
+          );
+        }
+        if (generationStatus.status !== "processing") {
+          break;
+        }
+        await new Promise((resolve) => setTimeout(resolve, 2e3));
+      }
+      return { success: false, error: "LLMs.txt generation job terminated unexpectedly" };
+    } catch (error) {
+      throw new FirecrawlError(error.message, 500, error.response?.data?.details);
+    }
+  }
+  /**
+   * Initiates a LLMs.txt generation operation without polling.
+   * @param url - The URL to generate LLMs.txt from.
+   * @param params - Parameters for the LLMs.txt generation operation.
+   * @returns The response containing the generation job ID.
+   */
+  async asyncGenerateLLMsText(url, params) {
+    const headers = this.prepareHeaders();
+    try {
+      const response = await this.postRequest(
+        `${this.apiUrl}/v1/llmstxt`,
+        { url, ...params },
+        headers
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else {
+        this.handleError(response, "start LLMs.txt generation");
+      }
+    } catch (error) {
+      if (error.response?.data?.error) {
+        throw new FirecrawlError(`Request failed with status code ${error.response.status}. Error: ${error.response.data.error} ${error.response.data.details ? ` - ${JSON.stringify(error.response.data.details)}` : ""}`, error.response.status);
+      } else {
+        throw new FirecrawlError(error.message, 500);
+      }
+    }
+    return { success: false, error: "Internal server error." };
+  }
+  /**
+   * Checks the status of a LLMs.txt generation operation.
+   * @param id - The ID of the LLMs.txt generation operation.
+   * @returns The current status and results of the generation operation.
+   */
+  async checkGenerateLLMsTextStatus(id) {
+    const headers = this.prepareHeaders();
+    try {
+      const response = await this.getRequest(
+        `${this.apiUrl}/v1/llmstxt/${id}`,
+        headers
+      );
+      if (response.status === 200) {
+        return response.data;
+      } else if (response.status === 404) {
+        throw new FirecrawlError("LLMs.txt generation job not found", 404);
+      } else {
+        this.handleError(response, "check LLMs.txt generation status");
+      }
+    } catch (error) {
+      if (error.response?.data?.error) {
+        throw new FirecrawlError(`Request failed with status code ${error.response.status}. Error: ${error.response.data.error} ${error.response.data.details ? ` - ${JSON.stringify(error.response.data.details)}` : ""}`, error.response.status);
+      } else {
+        throw new FirecrawlError(error.message, 500);
+      }
+    }
+    return { success: false, error: "Internal server error." };
+  }
 };
 var CrawlWatcher = class extends import_typescript_event_target.TypedEventTarget {
   ws;
@@ -692,7 +1052,8 @@ var CrawlWatcher = class extends import_typescript_event_target.TypedEventTarget
   constructor(id, app) {
     super();
     this.id = id;
-    this.ws = new import_isows.WebSocket(`${app.apiUrl}/v1/crawl/${id}`, app.apiKey);
+    const wsUrl = app.apiUrl.replace(/^http/, "ws");
+    this.ws = new import_isows.WebSocket(`${wsUrl}/v1/crawl/${id}`, app.apiKey);
     this.status = "scraping";
     this.data = [];
     const messageHandler = (msg) => {