firecrawl 1.5.0__py3-none-any.whl → 1.6.1__py3-none-any.whl

firecrawl/__init__.py

@@ -13,7 +13,7 @@ import os
 
 from .firecrawl import FirecrawlApp # noqa
 
-__version__ = "1.5.0"
+__version__ = "1.6.1"
 
 # Define the logger for the Firecrawl project
 logger: logging.Logger = logging.getLogger("firecrawl")

firecrawl/firecrawl.py

@@ -12,15 +12,33 @@ Classes:
 import logging
 import os
 import time
-from typing import Any, Dict, Optional, List
+from typing import Any, Dict, Optional, List, Union
 import json
 
 import requests
+import pydantic
 import websockets
 
 logger : logging.Logger = logging.getLogger("firecrawl")
 
 class FirecrawlApp:
+    class ExtractParams(pydantic.BaseModel):
+        """
+        Parameters for the extract operation.
+        """
+        prompt: str
+        schema: Optional[Any] = None
+        system_prompt: Optional[str] = None
+        allow_external_links: Optional[bool] = False
+
+    class ExtractResponse(pydantic.BaseModel):
+        """
+        Response from the extract operation.
+        """
+        success: bool
+        data: Optional[Any] = None
+        error: Optional[str] = None
+
     def __init__(self, api_key: Optional[str] = None, api_url: Optional[str] = None) -> None:
       """
       Initialize the FirecrawlApp instance with API key, API URL.
@@ -434,6 +452,52 @@ class FirecrawlApp:
         else:
             self._handle_error(response, 'check batch scrape status')
 
+
+    def extract(self, urls: List[str], params: Optional[ExtractParams] = None) -> Any:
+        """
+        Extracts information from a URL using the Firecrawl API.
+
+        Args:
+            urls (List[str]): The URLs to extract information from.
+            params (Optional[ExtractParams]): Additional parameters for the extract request.
+
+        Returns:
+            Union[ExtractResponse, ErrorResponse]: The response from the extract operation.
+        """
+        headers = self._prepare_headers()
+
+        if not params or not params.get('prompt'):
+            raise ValueError("Prompt is required")
+
+        if not params.get('schema'):
+            raise ValueError("Schema is required for extraction")
+
+        jsonData = {'urls': urls, **params}
+        jsonSchema = params['schema'].schema() if hasattr(params['schema'], 'schema') else None
+
+        try:
+            response = self._post_request(
+                f'{self.api_url}/v1/extract',
+                {
+                    **jsonData,
+                    'allowExternalLinks': params.get('allow_external_links', False),
+                    'schema': jsonSchema
+                },
+                headers
+            )
+            if response.status_code == 200:
+                data = response.json()
+                if data['success']:
+                    return data
+                else:
+                    raise Exception(f'Failed to extract. Error: {data["error"]}')
+            else:
+                self._handle_error(response, "extract")
+        except Exception as e:
+            raise ValueError(str(e), 500)
+
+        return {'success': False, 'error': "Internal server error."}
+
     def _prepare_headers(self, idempotency_key: Optional[str] = None) -> Dict[str, str]:
         """
         Prepare the headers for API requests.

{firecrawl-1.5.0.dist-info → firecrawl-1.6.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: firecrawl
-Version: 1.5.0
+Version: 1.6.1
 Summary: Python SDK for Firecrawl API
 Home-page: https://github.com/mendableai/firecrawl
 Author: Mendable.ai

firecrawl-1.6.1.dist-info/RECORD

@@ -0,0 +1,11 @@
+firecrawl/__init__.py,sha256=TYVUAqBdGLanvONLqDTalnRvVgT6bsdusGx-tN-KCEM,2543
+firecrawl/firecrawl.py,sha256=szpOBWjFhNGtQ65xda__VWTRiF_77kg37NRz-XU_rp8,30210
+firecrawl/__tests__/e2e_withAuth/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+firecrawl/__tests__/e2e_withAuth/test.py,sha256=L-umFR3WyrJso1EwqkxjbTMr5AEI4t5zDfhQcCzitOI,7911
+firecrawl/__tests__/v1/e2e_withAuth/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+firecrawl/__tests__/v1/e2e_withAuth/test.py,sha256=KQMmGAtJAIafja6EGtJ-W9162w2Hm6PNjqKl3_RQXLA,16456
+firecrawl-1.6.1.dist-info/LICENSE,sha256=nPCunEDwjRGHlmjvsiDUyIWbkqqyj3Ej84ntnh0g0zA,1084
+firecrawl-1.6.1.dist-info/METADATA,sha256=2kVO2LrhCtoxoQ9K9oTdX_AWf9ForvPtyAZ7iguxghg,10596
+firecrawl-1.6.1.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+firecrawl-1.6.1.dist-info/top_level.txt,sha256=jTvz79zWhiyAezfmmHe4FQ-hR60C59UU5FrjMjijLu8,10
+firecrawl-1.6.1.dist-info/RECORD,,

firecrawl-1.5.0.dist-info/RECORD

@@ -1,11 +0,0 @@
-firecrawl/__init__.py,sha256=6UEPRP09ZHMJ99KG4N8qiTyWW7RNaGzY18bYkrXRAqw,2543
-firecrawl/firecrawl.py,sha256=09QENx-ME8455WiRpKV53-2cFh7T4MX0fX5vRP7It0M,28045
-firecrawl/__tests__/e2e_withAuth/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-firecrawl/__tests__/e2e_withAuth/test.py,sha256=L-umFR3WyrJso1EwqkxjbTMr5AEI4t5zDfhQcCzitOI,7911
-firecrawl/__tests__/v1/e2e_withAuth/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-firecrawl/__tests__/v1/e2e_withAuth/test.py,sha256=KQMmGAtJAIafja6EGtJ-W9162w2Hm6PNjqKl3_RQXLA,16456
-firecrawl-1.5.0.dist-info/LICENSE,sha256=nPCunEDwjRGHlmjvsiDUyIWbkqqyj3Ej84ntnh0g0zA,1084
-firecrawl-1.5.0.dist-info/METADATA,sha256=Wwo7Do4R_42Fbkq0jhOMxhgCTC4iZxRJ1aILD2ijFzs,10596
-firecrawl-1.5.0.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-firecrawl-1.5.0.dist-info/top_level.txt,sha256=jTvz79zWhiyAezfmmHe4FQ-hR60C59UU5FrjMjijLu8,10
-firecrawl-1.5.0.dist-info/RECORD,,