ws-bom-robot-app 0.0.47__tar.gz → 0.0.48__tar.gz

{ws_bom_robot_app-0.0.47/ws_bom_robot_app.egg-info → ws_bom_robot_app-0.0.48}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ws_bom_robot_app
-Version: 0.0.47
+Version: 0.0.48
 Summary: A FastAPI application serving ws bom/robot/llm platform ai.
 Home-page: https://github.com/websolutespa/bom
 Author: Websolute Spa

{ws_bom_robot_app-0.0.47 → ws_bom_robot_app-0.0.48}/setup.py

@@ -4,7 +4,7 @@ _requirements = [line.split('#')[0].strip() for line in open("requirements.txt")
 
 setup(
     name="ws_bom_robot_app",
-    version="0.0.47",
+    version="0.0.48",
     description="A FastAPI application serving ws bom/robot/llm platform ai.",
     long_description=open("README.md", encoding='utf-8').read(),
     long_description_content_type="text/markdown",

{ws_bom_robot_app-0.0.47 → ws_bom_robot_app-0.0.48}/ws_bom_robot_app/llm/vector_store/integration/sitemap.py

@@ -20,6 +20,7 @@ class Sitemap(IntegrationStrategy):
         data["excludeTag"] (list[str]): default to ["script", "noscript", "style", "head", "header","nav","footer", "iframe"]
         data["excludeClass"] (list[str]): ["class1", "class2"]
         data["excludeId"] (list[str]): ["id1", "id2"]
+        data["restrictDomain"] (bool): if True, only urls from the same domain will be loaded, default to True
     """
     def __init__(self, knowledgebase_path: str, data: dict[str, Any]):
         super().__init__(knowledgebase_path, data)
@@ -30,6 +31,7 @@ class Sitemap(IntegrationStrategy):
         self.__exclude_tag: list[str] = self.data.get("excludeTag",[]) # type: ignore
         self.__exclude_class: list[str] = self.data.get("excludeClass",[]) # type: ignore
         self.__exclude_id: list[str] = self.data.get("excludeId",[]) # type: ignore
+        self.__restrict_to_same_domain: bool = self.data.get("restrictDomain", True) # type: ignore
     def working_subdirectory(self) -> str:
         return ""
     def _extract(self, tag: Tag) -> str:
@@ -75,7 +77,8 @@ class Sitemap(IntegrationStrategy):
                 web_path=self._remap_if_local(self.__sitemap_url),
                 filter_urls=self.__filter_urls,
                 parsing_function=self._parse,
-                is_local=self._is_local(self.__sitemap_url)
+                is_local=self._is_local(self.__sitemap_url),
+                restrict_to_same_domain=self.__restrict_to_same_domain
             )
             _docs = self._output([document async for document in self.alazy_load(_loader)])
             if self._is_local(self.__sitemap_url):

{ws_bom_robot_app-0.0.47 → ws_bom_robot_app-0.0.48/ws_bom_robot_app.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ws_bom_robot_app
-Version: 0.0.47
+Version: 0.0.48
 Summary: A FastAPI application serving ws bom/robot/llm platform ai.
 Home-page: https://github.com/websolutespa/bom
 Author: Websolute Spa