feat: external document loader support

backend/open_webui/config.py

@@ -1838,6 +1838,18 @@ CONTENT_EXTRACTION_ENGINE = PersistentConfig(
     os.environ.get("CONTENT_EXTRACTION_ENGINE", "").lower(),
 )
 
+EXTERNAL_DOCUMENT_LOADER_URL = PersistentConfig(
+    "EXTERNAL_DOCUMENT_LOADER_URL",
+    "rag.external_document_loader_url",
+    os.environ.get("EXTERNAL_DOCUMENT_LOADER_URL", ""),
+)
+
+EXTERNAL_DOCUMENT_LOADER_API_KEY = PersistentConfig(
+    "EXTERNAL_DOCUMENT_LOADER_API_KEY",
+    "rag.external_document_loader_api_key",
+    os.environ.get("EXTERNAL_DOCUMENT_LOADER_API_KEY", ""),
+)
+
 TIKA_SERVER_URL = PersistentConfig(
     "TIKA_SERVER_URL",
     "rag.tika_server_url",

backend/open_webui/main.py

@@ -206,6 +206,8 @@ from open_webui.config import (
     CHUNK_OVERLAP,
     CHUNK_SIZE,
     CONTENT_EXTRACTION_ENGINE,
+    EXTERNAL_DOCUMENT_LOADER_URL,
+    EXTERNAL_DOCUMENT_LOADER_API_KEY,
     TIKA_SERVER_URL,
     DOCLING_SERVER_URL,
     DOCLING_OCR_ENGINE,
@@ -646,6 +648,8 @@ app.state.config.ENABLE_RAG_HYBRID_SEARCH = ENABLE_RAG_HYBRID_SEARCH
 app.state.config.ENABLE_WEB_LOADER_SSL_VERIFICATION = ENABLE_WEB_LOADER_SSL_VERIFICATION
 
 app.state.config.CONTENT_EXTRACTION_ENGINE = CONTENT_EXTRACTION_ENGINE
+app.state.config.EXTERNAL_DOCUMENT_LOADER_URL = EXTERNAL_DOCUMENT_LOADER_URL
+app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY = EXTERNAL_DOCUMENT_LOADER_API_KEY
 app.state.config.TIKA_SERVER_URL = TIKA_SERVER_URL
 app.state.config.DOCLING_SERVER_URL = DOCLING_SERVER_URL
 app.state.config.DOCLING_OCR_ENGINE = DOCLING_OCR_ENGINE

backend/open_webui/retrieval/loaders/external_document.py

@@ -0,0 +1,58 @@
+import requests
+import logging
+from typing import Iterator, List, Union
+
+from langchain_core.document_loaders import BaseLoader
+from langchain_core.documents import Document
+from open_webui.env import SRC_LOG_LEVELS
+
+log = logging.getLogger(__name__)
+log.setLevel(SRC_LOG_LEVELS["RAG"])
+
+
+class ExternalDocumentLoader(BaseLoader):
+    def __init__(
+        self,
+        file_path,
+        url: str,
+        api_key: str,
+        mime_type=None,
+        **kwargs,
+    ) -> None:
+        self.url = url
+        self.api_key = api_key
+
+        self.file_path = file_path
+        self.mime_type = mime_type
+
+    def load(self) -> list[Document]:
+        with open(self.file_path, "rb") as f:
+            data = f.read()
+
+        headers = {}
+        if self.mime_type is not None:
+            headers["Content-Type"] = self.mime_type
+
+        if self.api_key is not None:
+            headers["Authorization"] = f"Bearer {self.api_key}"
+
+        url = self.url
+        if url.endswith("/"):
+            url = url[:-1]
+
+        r = requests.put(f"{url}/process", data=data, headers=headers)
+
+        if r.ok:
+            res = r.json()
+
+            if res:
+                return [
+                    Document(
+                        page_content=res.get("page_content"),
+                        metadata=res.get("metadata"),
+                    )
+                ]
+            else:
+                raise Exception("Error loading document: No content returned")
+        else:
+            raise Exception(f"Error loading document: {r.status_code} {r.text}")

backend/open_webui/retrieval/loaders/external_web.py

@@ -10,7 +10,7 @@ log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
-class ExternalLoader(BaseLoader):
+class ExternalWebLoader(BaseLoader):
     def __init__(
         self,
         web_paths: Union[str, List[str]],
@@ -32,7 +32,7 @@ class ExternalLoader(BaseLoader):
                 response = requests.post(
                     self.external_url,
                     headers={
-                        "User-Agent": "Open WebUI (https://github.com/open-webui/open-webui) RAG Bot",
+                        "User-Agent": "Open WebUI (https://github.com/open-webui/open-webui) External Web Loader",
                         "Authorization": f"Bearer {self.external_api_key}",
                     },
                     json={

backend/open_webui/retrieval/loaders/main.py

@@ -21,6 +21,8 @@ from langchain_community.document_loaders import (
 )
 from langchain_core.documents import Document
 
+
+from open_webui.retrieval.loaders.external_document import ExternalDocumentLoader
 from open_webui.retrieval.loaders.mistral import MistralLoader
 
 from open_webui.env import SRC_LOG_LEVELS, GLOBAL_LOG_LEVEL
@@ -213,6 +215,17 @@ class Loader:
     def _get_loader(self, filename: str, file_content_type: str, file_path: str):
         file_ext = filename.split(".")[-1].lower()
 
+        if (
+            self.engine == "external"
+            and self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_URL")
+            and self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_API_KEY")
+        ):
+            loader = ExternalDocumentLoader(
+                file_path=file_path,
+                url=self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_URL"),
+                api_key=self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_API_KEY"),
+                mime_type=file_content_type,
+            )
         if self.engine == "tika" and self.kwargs.get("TIKA_SERVER_URL"):
             if self._is_text_file(file_ext, file_content_type):
                 loader = TextLoader(file_path, autodetect_encoding=True)
@@ -269,6 +282,15 @@ class Loader:
             loader = MistralLoader(
                 api_key=self.kwargs.get("MISTRAL_OCR_API_KEY"), file_path=file_path
             )
+        elif (
+            self.engine == "external"
+            and self.kwargs.get("MISTRAL_OCR_API_KEY") != ""
+            and file_ext
+            in ["pdf"]  # Mistral OCR currently only supports PDF and images
+        ):
+            loader = MistralLoader(
+                api_key=self.kwargs.get("MISTRAL_OCR_API_KEY"), file_path=file_path
+            )
         else:
             if file_ext == "pdf":
                 loader = PyPDFLoader(

backend/open_webui/retrieval/web/utils.py

@@ -25,7 +25,7 @@ from langchain_community.document_loaders.firecrawl import FireCrawlLoader
 from langchain_community.document_loaders.base import BaseLoader
 from langchain_core.documents import Document
 from open_webui.retrieval.loaders.tavily import TavilyLoader
-from open_webui.retrieval.loaders.external import ExternalLoader
+from backend.open_webui.retrieval.loaders.external_web import ExternalWebLoader
 from open_webui.constants import ERROR_MESSAGES
 from open_webui.config import (
     ENABLE_RAG_LOCAL_WEB_FETCH,
@@ -628,7 +628,7 @@ def get_web_loader(
         web_loader_args["extract_depth"] = TAVILY_EXTRACT_DEPTH.value
 
     if WEB_LOADER_ENGINE.value == "external":
-        WebLoaderClass = ExternalLoader
+        WebLoaderClass = ExternalWebLoader
         web_loader_args["external_url"] = EXTERNAL_WEB_LOADER_URL.value
         web_loader_args["external_api_key"] = EXTERNAL_WEB_LOADER_API_KEY.value
 

backend/open_webui/routers/retrieval.py

@@ -352,6 +352,8 @@ async def get_rag_config(request: Request, user=Depends(get_admin_user)):
         # Content extraction settings
         "CONTENT_EXTRACTION_ENGINE": request.app.state.config.CONTENT_EXTRACTION_ENGINE,
         "PDF_EXTRACT_IMAGES": request.app.state.config.PDF_EXTRACT_IMAGES,
+        "EXTERNAL_DOCUMENT_LOADER_URL": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL,
+        "EXTERNAL_DOCUMENT_LOADER_API_KEY": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY,
         "TIKA_SERVER_URL": request.app.state.config.TIKA_SERVER_URL,
         "DOCLING_SERVER_URL": request.app.state.config.DOCLING_SERVER_URL,
         "DOCLING_OCR_ENGINE": request.app.state.config.DOCLING_OCR_ENGINE,
@@ -493,6 +495,9 @@ class ConfigForm(BaseModel):
     # Content extraction settings
     CONTENT_EXTRACTION_ENGINE: Optional[str] = None
     PDF_EXTRACT_IMAGES: Optional[bool] = None
+    EXTERNAL_DOCUMENT_LOADER_URL: Optional[str] = None
+    EXTERNAL_DOCUMENT_LOADER_API_KEY: Optional[str] = None
+
     TIKA_SERVER_URL: Optional[str] = None
     DOCLING_SERVER_URL: Optional[str] = None
     DOCLING_OCR_ENGINE: Optional[str] = None
@@ -583,6 +588,16 @@ async def update_rag_config(
         if form_data.PDF_EXTRACT_IMAGES is not None
         else request.app.state.config.PDF_EXTRACT_IMAGES
     )
+    request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL = (
+        form_data.EXTERNAL_DOCUMENT_LOADER_URL
+        if form_data.EXTERNAL_DOCUMENT_LOADER_URL is not None
+        else request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL
+    )
+    request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY = (
+        form_data.EXTERNAL_DOCUMENT_LOADER_API_KEY
+        if form_data.EXTERNAL_DOCUMENT_LOADER_API_KEY is not None
+        else request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY
+    )
     request.app.state.config.TIKA_SERVER_URL = (
         form_data.TIKA_SERVER_URL
         if form_data.TIKA_SERVER_URL is not None
@@ -818,6 +833,8 @@ async def update_rag_config(
         # Content extraction settings
         "CONTENT_EXTRACTION_ENGINE": request.app.state.config.CONTENT_EXTRACTION_ENGINE,
         "PDF_EXTRACT_IMAGES": request.app.state.config.PDF_EXTRACT_IMAGES,
+        "EXTERNAL_DOCUMENT_LOADER_URL": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL,
+        "EXTERNAL_DOCUMENT_LOADER_API_KEY": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY,
         "TIKA_SERVER_URL": request.app.state.config.TIKA_SERVER_URL,
         "DOCLING_SERVER_URL": request.app.state.config.DOCLING_SERVER_URL,
         "DOCLING_OCR_ENGINE": request.app.state.config.DOCLING_OCR_ENGINE,
@@ -1139,6 +1156,8 @@ def process_file(
                 file_path = Storage.get_file(file_path)
                 loader = Loader(
                     engine=request.app.state.config.CONTENT_EXTRACTION_ENGINE,
+                    EXTERNAL_DOCUMENT_LOADER_URL=request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL,
+                    EXTERNAL_DOCUMENT_LOADER_API_KEY=request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY,
                     TIKA_SERVER_URL=request.app.state.config.TIKA_SERVER_URL,
                     DOCLING_SERVER_URL=request.app.state.config.DOCLING_SERVER_URL,
                     DOCLING_OCR_ENGINE=request.app.state.config.DOCLING_OCR_ENGINE,

src/lib/components/admin/Settings/Documents.svelte

@@ -124,6 +124,13 @@
 	};
 
 	const submitHandler = async () => {
+		if (
+			RAGConfig.CONTENT_EXTRACTION_ENGINE === 'external' &&
+			RAGConfig.EXTERNAL_DOCUMENT_LOADER_URL === ''
+		) {
+			toast.error($i18n.t('External Document Loader URL required.'));
+			return;
+		}
 		if (RAGConfig.CONTENT_EXTRACTION_ENGINE === 'tika' && RAGConfig.TIKA_SERVER_URL === '') {
 			toast.error($i18n.t('Tika Server URL required.'));
 			return;
@@ -256,6 +263,7 @@
 									bind:value={RAGConfig.CONTENT_EXTRACTION_ENGINE}
 								>
 									<option value="">{$i18n.t('Default')}</option>
+									<option value="external">{$i18n.t('External')}</option>
 									<option value="tika">{$i18n.t('Tika')}</option>
 									<option value="docling">{$i18n.t('Docling')}</option>
 									<option value="document_intelligence">{$i18n.t('Document Intelligence')}</option>
@@ -275,6 +283,19 @@
 									</div>
 								</div>
 							</div>
+						{:else if RAGConfig.CONTENT_EXTRACTION_ENGINE === 'external'}
+							<div class="my-0.5 flex gap-2 pr-2">
+								<input
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
+									placeholder={$i18n.t('Enter External Document Loader URL')}
+									bind:value={RAGConfig.EXTERNAL_DOCUMENT_LOADER_URL}
+								/>
+								<SensitiveInput
+									placeholder={$i18n.t('Enter External Document Loader API Key')}
+									required={false}
+									bind:value={RAGConfig.EXTERNAL_DOCUMENT_LOADER_API_KEY}
+								/>
+							</div>
 						{:else if RAGConfig.CONTENT_EXTRACTION_ENGINE === 'tika'}
 							<div class="flex w-full mt-1">
 								<div class="flex-1 mr-2">