fix bug about fetching knowledge graph (#3394)

### What problem does this PR solve?


### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

api/apps/chunk_app.py

@@ -301,16 +301,13 @@ def retrieval_test():
 @login_required
 def knowledge_graph():
     doc_id = request.args["doc_id"]
-    e, doc = DocumentService.get_by_id(doc_id)
-    if not e:
-        return get_data_error_result(message="Document not found!")
     tenant_id = DocumentService.get_tenant_id(doc_id)
     kb_ids = KnowledgebaseService.get_kb_ids(tenant_id)
     req = {
         "doc_ids":[doc_id],
         "knowledge_graph_kwd": ["graph", "mind_map"]
     }
-    sres = retrievaler.search(req, search.index_name(tenant_id), kb_ids, doc.kb_id)
+    sres = retrievaler.search(req, search.index_name(tenant_id), kb_ids)
     obj = {"graph": {}, "mind_map": {}}
     for id in sres.ids[:2]:
         ty = sres.field[id]["knowledge_graph_kwd"]

api/apps/document_app.py

@@ -524,7 +524,7 @@ def upload_and_parse():
 @manager.route('/parse', methods=['POST'])
 @login_required
 def parse():
-    url = request.json.get("url")
+    url = request.json.get("url") if request.json else ""
     if url:
         if not is_valid_url(url):
             return get_json_result(
@@ -537,7 +537,7 @@ def parse():
         options.add_argument('--disable-dev-shm-usage')
         driver = Chrome(options=options)
         driver.get(url)
-        sections = RAGFlowHtmlParser()("", binary=driver.page_source)
+        sections = RAGFlowHtmlParser().parser_txt(driver.page_source)
         return get_json_result(data="\n".join(sections))
 
     if 'file' not in request.files:

api/db/services/file_service.py

@@ -15,6 +15,8 @@
 #
 import re
 import os
+from concurrent.futures import ThreadPoolExecutor
+
 from flask_login import current_user
 from peewee import fn
 
@@ -385,6 +387,41 @@ class FileService(CommonService):
 
         return err, files
 
+    @staticmethod
+    def parse_docs(file_objs, user_id):
+        from rag.app import presentation, picture, naive, audio, email
+
+        def dummy(prog=None, msg=""):
+            pass
+
+        FACTORY = {
+            ParserType.PRESENTATION.value: presentation,
+            ParserType.PICTURE.value: picture,
+            ParserType.AUDIO.value: audio,
+            ParserType.EMAIL.value: email
+        }
+        parser_config = {"chunk_token_num": 16096, "delimiter": "\n!?;。；！？", "layout_recognize": False}
+        exe = ThreadPoolExecutor(max_workers=12)
+        threads = []
+        for file in file_objs:
+            kwargs = {
+                "lang": "English",
+                "callback": dummy,
+                "parser_config": parser_config,
+                "from_page": 0,
+                "to_page": 100000,
+                "tenant_id": user_id
+            }
+            filetype = filename_type(file.filename)
+            blob = file.read()
+            threads.append(exe.submit(FACTORY.get(FileService.get_parser(filetype, file.filename, ""), naive).chunk, file.filename, blob, **kwargs))
+
+        res = []
+        for th in threads:
+            res.append("\n".join([ck["content_with_weight"] for ck in th.result()]))
+
+        return "\n\n".join(res)
+
     @staticmethod
     def get_parser(doc_type, filename, default):
         if doc_type == FileType.VISUAL:

api/db/services/knowledgebase_service.py

@@ -73,7 +73,7 @@ class KnowledgebaseService(CommonService):
             cls.model.id,
         ]
         kbs = cls.model.select(*fields).where(cls.model.tenant_id == tenant_id)
-        kb_ids = [kb["id"] for kb in kbs]
+        kb_ids = [kb.id for kb in kbs]
         return kb_ids
 
     @classmethod