Fix some issues in API (#2902)

### What problem does this PR solve?

Fix some issues in API

### Type of change

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

Co-authored-by: liuhua <[email protected]>

api/apps/sdk/chat.py

@@ -30,18 +30,17 @@ from api.utils.api_utils import get_result
 @token_required
 def create(tenant_id):
     req=request.json
-    if not req.get("knowledgebases"):
-        return get_error_data_result(retmsg="knowledgebases are required")
-    kb_list = []
-    for kb in req.get("knowledgebases"):
-        if not kb["id"]:
-            return get_error_data_result(retmsg="knowledgebase needs id")
-        if not KnowledgebaseService.query(id=kb["id"], tenant_id=tenant_id):
-            return get_error_data_result(retmsg="you do not own the knowledgebase")
-        # if not DocumentService.query(kb_id=kb["id"]):
-        #  return get_error_data_result(retmsg="There is a invalid knowledgebase")
-        kb_list.append(kb["id"])
-    req["kb_ids"] = kb_list
+    ids= req.get("knowledgebases")
+    if not ids:
+        return get_error_data_result(retmsg="`knowledgebases` is required")
+    for kb_id in ids:
+        kbs = KnowledgebaseService.query(id=kb_id,tenant_id=tenant_id)
+        if not kbs:
+            return get_error_data_result(f"You don't own the dataset {kb_id}")
+        kb=kbs[0]
+        if kb.chunk_num == 0:
+            return get_error_data_result(f"The dataset {kb_id} doesn't own parsed file")
+    req["kb_ids"] = ids
     # llm
     llm = req.get("llm")
     if llm:
@@ -81,24 +80,24 @@ def create(tenant_id):
     else:
         req["llm_id"] = tenant.llm_id
     if not req.get("name"):
-        return get_error_data_result(retmsg="name is required.")
+        return get_error_data_result(retmsg="`name` is required.")
     if DialogService.query(name=req["name"], tenant_id=tenant_id, status=StatusEnum.VALID.value):
-        return get_error_data_result(retmsg="Duplicated chat name in creating dataset.")
+        return get_error_data_result(retmsg="Duplicated chat name in creating chat.")
     # tenant_id
     if req.get("tenant_id"):
-        return get_error_data_result(retmsg="tenant_id must not be provided.")
+        return get_error_data_result(retmsg="`tenant_id` must not be provided.")
     req["tenant_id"] = tenant_id
     # prompt more parameter
     default_prompt = {
-        "system": """你是一个智能助手，请总结知识库的内容来回答问题，请列举知识库中的数据详细回答。当所有知识库内容都与问题无关时，你的回答必须包括“知识库中未找到您要的答案！”这句话。回答需要考虑聊天历史。
-            以下是知识库：
-            {knowledge}
-            以上是知识库。""",
-        "prologue": "您好，我是您的助手小樱，长得可爱又善良，can I help you?",
+        "system": """You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence "The answer you are looking for is not found in the knowledge base!" Answers need to consider chat history.
+      Here is the knowledge base:
+      {knowledge}
+      The above is the knowledge base.""",
+        "prologue": "Hi! I'm your assistant, what can I do for you?",
         "parameters": [
             {"key": "knowledge", "optional": False}
         ],
-        "empty_response": "Sorry! 知识库中未找到相关内容！"
+        "empty_response": "Sorry! No relevant content was found in the knowledge base!"
     }
     key_list_2 = ["system", "prologue", "parameters", "empty_response"]
     if "prompt_config" not in req:
@@ -149,7 +148,7 @@ def update(tenant_id,chat_id):
     req =request.json
     if "knowledgebases" in req:
         if not req.get("knowledgebases"):
-            return  get_error_data_result(retmsg="knowledgebases can't be empty value")
+            return  get_error_data_result(retmsg="`knowledgebases` can't be empty value")
         kb_list = []
         for kb in req.get("knowledgebases"):
             if not kb["id"]:
@@ -189,10 +188,10 @@ def update(tenant_id,chat_id):
     res = res.to_json()
     if "llm_id" in req:
         if not TenantLLMService.query(llm_name=req["llm_id"]):
-            return get_error_data_result(retmsg="the model_name does not exist.")
+            return get_error_data_result(retmsg="The `model_name` does not exist.")
     if "name" in req:
         if not req.get("name"):
-            return get_error_data_result(retmsg="name is not empty.")
+            return get_error_data_result(retmsg="`name` is not empty.")
         if req["name"].lower() != res["name"].lower() \
                 and len(
             DialogService.query(name=req["name"], tenant_id=tenant_id, status=StatusEnum.VALID.value)) > 0:
@@ -224,7 +223,7 @@ def delete(tenant_id):
     req = request.json
     ids = req.get("ids")
     if not ids:
-        return get_error_data_result(retmsg="ids are required")
+        return get_error_data_result(retmsg="`ids` are required")
     for id in ids:
         if not DialogService.query(tenant_id=tenant_id, id=id, status=StatusEnum.VALID.value):
             return get_error_data_result(retmsg=f"You don't own the chat {id}")
@@ -234,7 +233,7 @@ def delete(tenant_id):
 
 @manager.route('/chat', methods=['GET'])
 @token_required
-def list(tenant_id):
+def list_chat(tenant_id):
     id = request.args.get("id")
     name = request.args.get("name")
     chat = DialogService.query(id=id,name=name,status=StatusEnum.VALID.value)

api/apps/sdk/dataset.py

@@ -25,28 +25,38 @@ from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.user_service import TenantService
 from api.settings import RetCode
 from api.utils import get_uuid
-from api.utils.api_utils import get_result, token_required,get_error_data_result
+from api.utils.api_utils import get_result, token_required, get_error_data_result, valid
+
 
 @manager.route('/dataset', methods=['POST'])
 @token_required
 def create(tenant_id):
     req = request.json
     e, t = TenantService.get_by_id(tenant_id)
+    permission = req.get("permission")
+    language = req.get("language")
+    chunk_method = req.get("chunk_method")
+    valid_permission = ("me", "team")
+    valid_language =("Chinese", "English")
+    valid_chunk_method = ("naive","manual","qa","table","paper","book","laws","presentation","picture","one","knowledge_graph","email")
+    check_validation=valid(permission,valid_permission,language,valid_language,chunk_method,valid_chunk_method)
+    if check_validation:
+        return check_validation
     if "tenant_id" in req or "embedding_model" in req:
         return get_error_data_result(
-            retmsg="Tenant_id or embedding_model must not be provided")
+            retmsg="`tenant_id` or `embedding_model` must not be provided")
     chunk_count=req.get("chunk_count")
     document_count=req.get("document_count")
     if chunk_count or document_count:
-        return get_error_data_result(retmsg="chunk_count or document_count must be 0 or not be provided")
+        return get_error_data_result(retmsg="`chunk_count` or `document_count` must be 0 or not be provided")
     if "name" not in req:
         return get_error_data_result(
-            retmsg="Name is not empty!")
+            retmsg="`name` is not empty!")
     req['id'] = get_uuid()
     req["name"] = req["name"].strip()
     if req["name"] == "":
         return get_error_data_result(
-            retmsg="Name is not empty string!")
+            retmsg="`name` is not empty string!")
     if KnowledgebaseService.query(name=req["name"], tenant_id=tenant_id, status=StatusEnum.VALID.value):
         return get_error_data_result(
             retmsg="Duplicated knowledgebase name in creating dataset.")
@@ -55,7 +65,7 @@ def create(tenant_id):
     key_mapping = {
         "chunk_num": "chunk_count",
         "doc_num": "document_count",
-        "parser_id": "parse_method",
+        "parser_id": "chunk_method",
         "embd_id": "embedding_model"
     }
     mapped_keys = {new_key: req[old_key] for new_key, old_key in key_mapping.items() if old_key in req}
@@ -90,7 +100,7 @@ def delete(tenant_id):
         File2DocumentService.delete_by_document_id(doc.id)
     if not KnowledgebaseService.delete_by_id(id):
         return get_error_data_result(
-            retmsg="Delete dataset error.(Database serror)")
+            retmsg="Delete dataset error.(Database error)")
     return get_result(retcode=RetCode.SUCCESS)
 
 @manager.route('/dataset/<dataset_id>', methods=['PUT'])
@@ -103,30 +113,39 @@ def update(tenant_id,dataset_id):
     invalid_keys = {"id", "embd_id", "chunk_num", "doc_num", "parser_id"}
     if any(key in req for key in invalid_keys):
         return get_error_data_result(retmsg="The input parameters are invalid.")
+    permission = req.get("permission")
+    language = req.get("language")
+    chunk_method = req.get("chunk_method")
+    valid_permission = ("me", "team")
+    valid_language =("Chinese", "English")
+    valid_chunk_method = ("naive","manual","qa","table","paper","book","laws","presentation","picture","one","knowledge_graph","email")
+    check_validation=valid(permission,valid_permission,language,valid_language,chunk_method,valid_chunk_method)
+    if check_validation:
+        return check_validation
     if "tenant_id" in req:
         if req["tenant_id"] != tenant_id:
             return get_error_data_result(
-                retmsg="Can't change tenant_id.")
+                retmsg="Can't change `tenant_id`.")
     e, kb = KnowledgebaseService.get_by_id(dataset_id)
     if "chunk_count" in req:
         if req["chunk_count"] != kb.chunk_num:
             return get_error_data_result(
-                retmsg="Can't change chunk_count.")
+                retmsg="Can't change `chunk_count`.")
         req.pop("chunk_count")
     if "document_count" in req:
         if req['document_count'] != kb.doc_num:
             return get_error_data_result(
-                retmsg="Can't change document_count.")
+                retmsg="Can't change `document_count`.")
         req.pop("document_count")
-    if "parse_method" in req:
-        if kb.chunk_num != 0 and req['parse_method'] != kb.parser_id:
+    if "chunk_method" in req:
+        if kb.chunk_num != 0 and req['chunk_method'] != kb.parser_id:
             return get_error_data_result(
-                retmsg="If chunk count is not 0, parse method is not changable.")
-        req['parser_id'] = req.pop('parse_method')
+                retmsg="If `chunk_count` is not 0, `chunk_method` is not changeable.")
+        req['parser_id'] = req.pop('chunk_method')
     if "embedding_model" in req:
-        if kb.chunk_num != 0 and req['parse_method'] != kb.parser_id:
+        if kb.chunk_num != 0 and req['embedding_model'] != kb.embd_id:
             return get_error_data_result(
-                retmsg="If chunk count is not 0, parse method is not changable.")
+                retmsg="If `chunk_count` is not 0, `embedding_method` is not changeable.")
         req['embd_id'] = req.pop('embedding_model')
     if "name" in req:
         req["name"] = req["name"].strip()
@@ -162,7 +181,7 @@ def list(tenant_id):
         key_mapping = {
             "chunk_num": "chunk_count",
             "doc_num": "document_count",
-            "parser_id": "parse_method",
+            "parser_id": "chunk_method",
             "embd_id": "embedding_model"
         }
         renamed_data = {}

api/apps/sdk/doc.py

@@ -88,20 +88,20 @@ def upload(dataset_id, tenant_id):
 def update_doc(tenant_id, dataset_id, document_id):
     req = request.json
     if not KnowledgebaseService.query(id=dataset_id, tenant_id=tenant_id):
-        return get_error_data_result(retmsg='You do not own the dataset.')
+        return get_error_data_result(retmsg="You don't own the dataset.")
     doc = DocumentService.query(kb_id=dataset_id, id=document_id)
     if not doc:
-        return get_error_data_result(retmsg='The dataset not own the document.')
+        return get_error_data_result(retmsg="The dataset doesn't own the document.")
     doc = doc[0]
     if "chunk_count" in req:
         if req["chunk_count"] != doc.chunk_num:
-            return get_error_data_result(retmsg="Can't change chunk_count.")
+            return get_error_data_result(retmsg="Can't change `chunk_count`.")
     if "token_count" in req:
         if req["token_count"] != doc.token_num:
-            return get_error_data_result(retmsg="Can't change token_count.")
+            return get_error_data_result(retmsg="Can't change `token_count`.")
     if "progress" in req:
         if req['progress'] != doc.progress:
-            return get_error_data_result(retmsg="Can't change progress.")
+            return get_error_data_result(retmsg="Can't change `progress`.")
 
     if "name" in req and req["name"] != doc.name:
         if pathlib.Path(req["name"].lower()).suffix != pathlib.Path(doc.name.lower()).suffix:
@@ -121,8 +121,8 @@ def update_doc(tenant_id, dataset_id, document_id):
             FileService.update_by_id(file.id, {"name": req["name"]})
     if "parser_config" in req:
         DocumentService.update_parser_config(doc.id, req["parser_config"])
-    if "parser_method" in req:
-        if doc.parser_id.lower() == req["parser_method"].lower():
+    if "chunk_method" in req:
+        if doc.parser_id.lower() == req["chunk_method"].lower():
                 return get_result()
 
         if doc.type == FileType.VISUAL or re.search(
@@ -130,7 +130,7 @@ def update_doc(tenant_id, dataset_id, document_id):
             return get_error_data_result(retmsg="Not supported yet!")
 
         e = DocumentService.update_by_id(doc.id,
-                                         {"parser_id": req["parser_method"], "progress": 0, "progress_msg": "",
+                                         {"parser_id": req["chunk_method"], "progress": 0, "progress_msg": "",
                                           "run": TaskStatus.UNSTART.value})
         if not e:
             return get_error_data_result(retmsg="Document not found!")
@@ -196,7 +196,7 @@ def list_docs(dataset_id, tenant_id):
             "chunk_num": "chunk_count",
             "kb_id": "knowledgebase_id",
             "token_num": "token_count",
-            "parser_id": "parser_method"
+            "parser_id": "chunk_method"
         }
         renamed_doc = {}
         for key, value in doc.items():
@@ -213,7 +213,7 @@ def delete(tenant_id,dataset_id):
         return get_error_data_result(retmsg=f"You don't own the dataset {dataset_id}. ")
     req = request.json
     if not req.get("ids"):
-        return get_error_data_result(retmsg="ids is required")
+        return get_error_data_result(retmsg="`ids` is required")
     doc_ids = req["ids"]
     root_folder = FileService.get_root_folder(tenant_id)
     pf_id = root_folder["id"]
@@ -457,7 +457,7 @@ def rm_chunk(tenant_id,dataset_id,document_id):
 
 @manager.route('/dataset/<dataset_id>/document/<document_id>/chunk/<chunk_id>', methods=['PUT'])
 @token_required
-def set(tenant_id,dataset_id,document_id,chunk_id):
+def update_chunk(tenant_id,dataset_id,document_id,chunk_id):
     try:
         res = ELASTICSEARCH.get(
         chunk_id, search.index_name(
@@ -519,9 +519,15 @@ def retrieval_test(tenant_id):
     req = request.json
     if not req.get("datasets"):
         return get_error_data_result("`datasets` is required.")
-    kb_id = req["datasets"]
-    if isinstance(kb_id, str): kb_id = [kb_id]
-    for id in kb_id:
+    kb_ids = req["datasets"]
+    kbs = KnowledgebaseService.get_by_ids(kb_ids)
+    embd_nms = list(set([kb.embd_id for kb in kbs]))
+    if len(embd_nms) != 1:
+        return get_result(
+            retmsg='Knowledge bases use different embedding models or does not exist."',
+            retcode=RetCode.AUTHENTICATION_ERROR)
+    if isinstance(kb_ids, str): kb_ids = [kb_ids]
+    for id in kb_ids:
         if not KnowledgebaseService.query(id=id,tenant_id=tenant_id):
             return get_error_data_result(f"You don't own the dataset {id}.")
     if "question" not in req:
@@ -538,7 +544,7 @@ def retrieval_test(tenant_id):
     else:
         highlight = True
     try:
-        e, kb = KnowledgebaseService.get_by_id(kb_id[0])
+        e, kb = KnowledgebaseService.get_by_id(kb_ids[0])
         if not e:
             return get_error_data_result(retmsg="Knowledgebase not found!")
         embd_mdl = TenantLLMService.model_instance(
@@ -554,7 +560,7 @@ def retrieval_test(tenant_id):
             question += keyword_extraction(chat_mdl, question)
 
         retr = retrievaler if kb.parser_id != ParserType.KG else kg_retrievaler
-        ranks = retr.retrieval(question, embd_mdl, kb.tenant_id, kb_id, page, size,
+        ranks = retr.retrieval(question, embd_mdl, kb.tenant_id, kb_ids, page, size,
                                similarity_threshold, vector_similarity_weight, top,
                                doc_ids, rerank_mdl=rerank_mdl, highlight=highlight)
         for c in ranks["chunks"]:
@@ -580,6 +586,6 @@ def retrieval_test(tenant_id):
         return get_result(data=ranks)
     except Exception as e:
         if str(e).find("not_found") > 0:
-            return get_result(retmsg=f'No chunk found! Check the chunk statu    s please!',
+            return get_result(retmsg=f'No chunk found! Check the chunk status please!',
                                    retcode=RetCode.DATA_ERROR)
         return server_error_response(e)

api/apps/sdk/session.py

@@ -39,7 +39,7 @@ def create(tenant_id,chat_id):
         "message": [{"role": "assistant", "content": "Hi! I am your assistant，can I help you?"}]
     }
     if not conv.get("name"):
-        return get_error_data_result(retmsg="Name can not be empty.")
+        return get_error_data_result(retmsg="`name` can not be empty.")
     ConversationService.save(**conv)
     e, conv = ConversationService.get_by_id(conv["id"])
     if not e:
@@ -62,11 +62,11 @@ def update(tenant_id,chat_id,session_id):
     if not DialogService.query(id=chat_id, tenant_id=tenant_id, status=StatusEnum.VALID.value):
         return get_error_data_result(retmsg="You do not own the session")
     if "message" in req or "messages" in req:
-        return get_error_data_result(retmsg="Message can not be change")
+        return get_error_data_result(retmsg="`message` can not be change")
     if "reference" in req:
-        return get_error_data_result(retmsg="Reference can not be change")
+        return get_error_data_result(retmsg="`reference` can not be change")
     if "name" in req and not req.get("name"):
-        return get_error_data_result(retmsg="Name can not be empty.")
+        return get_error_data_result(retmsg="`name` can not be empty.")
     if not ConversationService.update_by_id(conv_id, req):
         return get_error_data_result(retmsg="Session updates error")
     return get_result()
@@ -87,7 +87,7 @@ def completion(tenant_id,chat_id):
             "message": [{"role": "assistant", "content": "Hi! I am your assistant，can I help you?"}]
         }
         if not conv.get("name"):
-            return get_error_data_result(retmsg="Name can not be empty.")
+            return get_error_data_result(retmsg="`name` can not be empty.")
         ConversationService.save(**conv)
         e, conv = ConversationService.get_by_id(conv["id"])
         session_id=conv.id

api/db/db_models.py

@@ -879,8 +879,8 @@ class Dialog(DataBaseModel):
         default="simple",
         help_text="simple|advanced",
         index=True)
-    prompt_config = JSONField(null=False, default={"system": "", "prologue": "您好，我是您的助手小樱，长得可爱又善良，can I help you?",
-                                                   "parameters": [], "empty_response": "Sorry! 知识库中未找到相关内容！"})
+    prompt_config = JSONField(null=False, default={"system": "", "prologue": "Hi! I'm your assistant, what can I do for you?",
+                                                   "parameters": [], "empty_response": "Sorry! No relevant content was found in the knowledge base!"})
 
     similarity_threshold = FloatField(default=0.2)
     vector_similarity_weight = FloatField(default=0.3)

api/utils/api_utils.py

@@ -324,4 +324,17 @@ def get_error_data_result(retmsg='Sorry! Data missing!', retcode=RetCode.DATA_ER
 
 def generate_confirmation_token(tenent_id):
     serializer = URLSafeTimedSerializer(tenent_id)
-    return "ragflow-" + serializer.dumps(get_uuid(), salt=tenent_id)[2:34]
+    return "ragflow-" + serializer.dumps(get_uuid(), salt=tenent_id)[2:34]
+
+
+def valid(permission,valid_permission,language,valid_language,chunk_method,valid_chunk_method):
+    if valid_parameter(permission,valid_permission):
+        return valid_parameter(permission,valid_permission)
+    if valid_parameter(language,valid_language):
+        return valid_parameter(language,valid_language)
+    if valid_parameter(chunk_method,valid_chunk_method):
+        return valid_parameter(chunk_method,valid_chunk_method)
+
+def valid_parameter(parameter,valid_values):
+    if parameter and parameter not in valid_values:
+       return get_error_data_result(f"{parameter}  not in {valid_values}")

sdk/python/ragflow/modules/base.py

@@ -22,7 +22,7 @@ class Base(object):
         res = self.rag.post(path, json, stream=stream,files=files)
         return res
 
-    def get(self, path, params):
+    def get(self, path, params=None):
         res = self.rag.get(path, params)
         return res
 

sdk/python/ragflow/modules/chat.py

@@ -73,6 +73,3 @@ class Chat(Base):
         res = res.json()
         if res.get("code") != 0:
             raise Exception(res.get("message"))
-
-    def get_prologue(self):
-        return self.prompt.opener

sdk/python/ragflow/modules/dataset.py

@@ -1,78 +1,78 @@
-from typing import Optional, List
-
-from transformers.models.bloom.modeling_bloom import bloom_gelu_back
-
-from .document import Document
-
-from .base import Base
-
-
-class DataSet(Base):
-    class ParserConfig(Base):
-        def __init__(self, rag, res_dict):
-            self.chunk_token_count = 128
-            self.layout_recognize = True
-            self.delimiter = '\n!?。；！？'
-            self.task_page_size = 12
-            super().__init__(rag, res_dict)
-
-    def __init__(self, rag, res_dict):
-        self.id = ""
-        self.name = ""
-        self.avatar = ""
-        self.tenant_id = None
-        self.description = ""
-        self.language = "English"
-        self.embedding_model = ""
-        self.permission = "me"
-        self.document_count = 0
-        self.chunk_count = 0
-        self.parse_method = "naive"
-        self.parser_config = None
-        for k in list(res_dict.keys()):
-            if k not in self.__dict__:
-                res_dict.pop(k)
-        super().__init__(rag, res_dict)
-
-    def update(self, update_message: dict):
-        res = self.put(f'/dataset/{self.id}',
-                        update_message)
-        res = res.json()
-        if res.get("code") != 0:
-            raise Exception(res["message"])
-
-    def upload_documents(self,document_list: List[dict]):
-        url = f"/dataset/{self.id}/document"
-        files = [("file",(ele["name"],ele["blob"])) for ele in document_list]
-        res = self.post(path=url,json=None,files=files)
-        res = res.json()
-        if res.get("code") != 0:
-            raise Exception(res.get("message"))
-
-    def list_documents(self, id: str = None, keywords: str = None, offset: int =1, limit: int = 1024, orderby: str = "create_time", desc: bool = True):
-        res = self.get(f"/dataset/{self.id}/info",params={"id": id,"keywords": keywords,"offset": offset,"limit": limit,"orderby": orderby,"desc": desc})
-        res = res.json()
-        documents = []
-        if res.get("code") == 0:
-            for document in res["data"].get("docs"):
-                documents.append(Document(self.rag,document))
-            return documents
-        raise Exception(res["message"])
-
-    def delete_documents(self,ids: List[str] = None):
-        res = self.rm(f"/dataset/{self.id}/document",{"ids":ids})
-        res = res.json()
-        if res.get("code") != 0:
-            raise Exception(res["message"])
-
-    def async_parse_documents(self,document_ids):
-        res = self.post(f"/dataset/{self.id}/chunk",{"document_ids":document_ids})
-        res = res.json()
-        if res.get("code") != 0:
-            raise Exception(res.get("message"))
-
-    def async_cancel_parse_documents(self,document_ids):
-        res = self.rm(f"/dataset/{self.id}/chunk",{"document_ids":document_ids})
-        res = res.json()
-        if res.get("code") != 0:
-            raise Exception(res.get("message"))
+from typing import Optional, List
+
+from transformers.models.bloom.modeling_bloom import bloom_gelu_back
+
+from .document import Document
+
+from .base import Base
+
+
+class DataSet(Base):
+    class ParserConfig(Base):
+        def __init__(self, rag, res_dict):
+            self.chunk_token_count = 128
+            self.layout_recognize = True
+            self.delimiter = '\n!?。；！？'
+            self.task_page_size = 12
+            super().__init__(rag, res_dict)
+
+    def __init__(self, rag, res_dict):
+        self.id = ""
+        self.name = ""
+        self.avatar = ""
+        self.tenant_id = None
+        self.description = ""
+        self.language = "English"
+        self.embedding_model = ""
+        self.permission = "me"
+        self.document_count = 0
+        self.chunk_count = 0
+        self.chunk_method = "naive"
+        self.parser_config = None
+        for k in list(res_dict.keys()):
+            if k not in self.__dict__:
+                res_dict.pop(k)
+        super().__init__(rag, res_dict)
+
+    def update(self, update_message: dict):
+        res = self.put(f'/dataset/{self.id}',
+                        update_message)
+        res = res.json()
+        if res.get("code") != 0:
+            raise Exception(res["message"])
+
+    def upload_documents(self,document_list: List[dict]):
+        url = f"/dataset/{self.id}/document"
+        files = [("file",(ele["name"],ele["blob"])) for ele in document_list]
+        res = self.post(path=url,json=None,files=files)
+        res = res.json()
+        if res.get("code") != 0:
+            raise Exception(res.get("message"))
+
+    def list_documents(self, id: str = None, keywords: str = None, offset: int =1, limit: int = 1024, orderby: str = "create_time", desc: bool = True):
+        res = self.get(f"/dataset/{self.id}/info",params={"id": id,"keywords": keywords,"offset": offset,"limit": limit,"orderby": orderby,"desc": desc})
+        res = res.json()
+        documents = []
+        if res.get("code") == 0:
+            for document in res["data"].get("docs"):
+                documents.append(Document(self.rag,document))
+            return documents
+        raise Exception(res["message"])
+
+    def delete_documents(self,ids: List[str] = None):
+        res = self.rm(f"/dataset/{self.id}/document",{"ids":ids})
+        res = res.json()
+        if res.get("code") != 0:
+            raise Exception(res["message"])
+
+    def async_parse_documents(self,document_ids):
+        res = self.post(f"/dataset/{self.id}/chunk",{"document_ids":document_ids})
+        res = res.json()
+        if res.get("code") != 0:
+            raise Exception(res.get("message"))
+
+    def async_cancel_parse_documents(self,document_ids):
+        res = self.rm(f"/dataset/{self.id}/chunk",{"document_ids":document_ids})
+        res = res.json()
+        if res.get("code") != 0:
+            raise Exception(res.get("message"))

sdk/python/ragflow/modules/document.py

@@ -1,7 +1,4 @@
-import time
-
-from PIL.ImageFile import raise_oserror
-
+import json
 from .base import Base
 from .chunk import Chunk
 from typing import List
@@ -13,7 +10,7 @@ class Document(Base):
         self.name = ""
         self.thumbnail = None
         self.knowledgebase_id = None
-        self.parser_method = ""
+        self.chunk_method = ""
         self.parser_config = {"pages": [[1, 1000000]]}
         self.source_type = "local"
         self.type = ""
@@ -32,6 +29,23 @@ class Document(Base):
                 res_dict.pop(k)
         super().__init__(rag, res_dict)
 
+
+    def update(self, update_message: dict):
+        res = self.put(f'/dataset/{self.knowledgebase_id}/info/{self.id}',
+                       update_message)
+        res = res.json()
+        if res.get("code") != 0:
+            raise Exception(res["message"])
+
+    def download(self):
+        res = self.get(f"/dataset/{self.knowledgebase_id}/document/{self.id}")
+        try:
+            res = res.json()
+            raise Exception(res.get("message"))
+        except json.JSONDecodeError:
+            return res.content
+
+
     def list_chunks(self,offset=0, limit=30, keywords="", id:str=None):
         data={"document_id": self.id,"keywords": keywords,"offset":offset,"limit":limit,"id":id}
         res = self.get(f'/dataset/{self.knowledgebase_id}/document/{self.id}/chunk', data)

sdk/python/ragflow/ragflow.py

@@ -24,11 +24,11 @@ from .modules.document import Document
 
 
 class RAGFlow:
-    def __init__(self, user_key, base_url, version='v1'):
+    def __init__(self, api_key, base_url, version='v1'):
         """
         api_url: http://<host_address>/api/v1
         """
-        self.user_key = user_key
+        self.user_key = api_key
         self.api_url = f"{base_url}/api/{version}"
         self.authorization_header = {"Authorization": "{} {}".format("Bearer", self.user_key)}
 
@@ -50,7 +50,7 @@ class RAGFlow:
 
     def create_dataset(self, name: str, avatar: str = "", description: str = "", language: str = "English",
                        permission: str = "me",
-                       document_count: int = 0, chunk_count: int = 0, parse_method: str = "naive",
+                       document_count: int = 0, chunk_count: int = 0, chunk_method: str = "naive",
                        parser_config: DataSet.ParserConfig = None) -> DataSet:
         if parser_config is None:
             parser_config = DataSet.ParserConfig(self, {"chunk_token_count": 128, "layout_recognize": True,
@@ -59,7 +59,7 @@ class RAGFlow:
         res = self.post("/dataset",
                         {"name": name, "avatar": avatar, "description": description, "language": language,
                          "permission": permission,
-                         "document_count": document_count, "chunk_count": chunk_count, "parse_method": parse_method,
+                         "document_count": document_count, "chunk_count": chunk_count, "chunk_method": chunk_method,
                          "parser_config": parser_config
                          }
                         )
@@ -93,7 +93,7 @@ class RAGFlow:
             return result_list
         raise Exception(res["message"])
 
-    def create_chat(self, name: str = "assistant", avatar: str = "path", knowledgebases: List[DataSet] = [],
+    def create_chat(self, name: str, avatar: str = "", knowledgebases: List[DataSet] = [],
                          llm: Chat.LLM = None, prompt: Chat.Prompt = None) -> Chat:
         datasets = []
         for dataset in knowledgebases:

sdk/python/test/t_document.py

@@ -35,7 +35,7 @@ class TestDocument(TestSdk):
     def test_update_document_with_success(self):
         """
         Test updating a document with success.
-        Update name or parser_method are supported
+        Update name or chunk_method are supported
         """
         rag = RAGFlow(API_KEY, HOST_ADDRESS)
         ds = rag.list_datasets(name="God")
@@ -43,7 +43,7 @@ class TestDocument(TestSdk):
         doc = ds.list_documents()
         doc = doc[0]
         if isinstance(doc, Document):
-            res = doc.update({"parser_method":"manual","name":"manual.txt"})
+            res = doc.update({"chunk_method":"manual","name":"manual.txt"})
             assert res is None, f"Failed to update document, error: {res}"
         else:
             assert False, f"Failed to get document, error: {doc}"