refactor code (infiniflow#583)

### What problem does this PR solve?

### Type of change

- [x] Refactoring

api/apps/api_app.py

@@ -33,7 +33,7 @@
 from itsdangerous import URLSafeTimedSerializer
 
 from api.utils.file_utils import filename_type, thumbnail
-from rag.utils import MINIO
+from rag.utils.minio_conn import MINIO
 
 
 def generate_confirmation_token(tenent_id):

api/apps/chunk_app.py

@@ -21,7 +21,8 @@
 
 from rag.app.qa import rmPrefix, beAdoc
 from rag.nlp import search, huqie
-from rag.utils import ELASTICSEARCH, rmSpace
+from rag.utils.es_conn import ELASTICSEARCH
+from rag.utils import rmSpace
 from api.db import LLMType, ParserType
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.llm_service import TenantLLMService

api/apps/document_app.py

@@ -27,7 +27,7 @@
 from api.db.services.file2document_service import File2DocumentService
 from api.db.services.file_service import FileService
 from rag.nlp import search
-from rag.utils import ELASTICSEARCH
+from rag.utils.es_conn import ELASTICSEARCH
 from api.db.services import duplicate_name
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.utils.api_utils import server_error_response, get_data_error_result, validate_request

api/apps/file2document_app.py

@@ -29,7 +29,7 @@
 from api.settings import RetCode
 from api.utils.api_utils import get_json_result
 from rag.nlp import search
-from rag.utils import ELASTICSEARCH
+from rag.utils.es_conn import ELASTICSEARCH
 
 
 @manager.route('/convert', methods=['POST'])

api/apps/file_app.py

@@ -33,7 +33,7 @@
 from api.utils.api_utils import get_json_result
 from api.utils.file_utils import filename_type
 from rag.nlp import search
-from rag.utils import ELASTICSEARCH
+from rag.utils.es_conn import ELASTICSEARCH
 from rag.utils.minio_conn import MINIO
 
 

api/apps/kb_app.py

@@ -28,7 +28,7 @@
 from api.settings import stat_logger, RetCode
 from api.utils.api_utils import get_json_result
 from rag.nlp import search
-from rag.utils import ELASTICSEARCH
+from rag.utils.es_conn import ELASTICSEARCH
 
 
 @manager.route('/create', methods=['post'])

api/db/services/document_service.py

@@ -16,7 +16,7 @@
 from peewee import Expression
 
 from elasticsearch_dsl import Q
-from rag.utils import ELASTICSEARCH
+from rag.utils.es_conn import ELASTICSEARCH
 from rag.utils.minio_conn import MINIO
 from rag.nlp import search
 

api/settings.py

@@ -32,7 +32,7 @@
 database_logger = getLogger("database")
 chat_logger = getLogger("chat")
 
-from rag.utils import ELASTICSEARCH
+from rag.utils.es_conn import ELASTICSEARCH
 from rag.nlp import search
 from api.utils import get_base_config, decrypt_database_config
 

deepdoc/parser/__init__.py

@@ -1,6 +1,6 @@
 
 
-from .pdf_parser import HuParser as PdfParser, PlainParser
-from .docx_parser import HuDocxParser as DocxParser
-from .excel_parser import HuExcelParser as ExcelParser
-from .ppt_parser import HuPptParser as PptParser
+from .pdf_parser import RAGFlowPdfParser as PdfParser, PlainParser
+from .docx_parser import RAGFlowDocxParser as DocxParser
+from .excel_parser import RAGFlowExcelParser as ExcelParser
+from .ppt_parser import RAGFlowPptParser as PptParser

deepdoc/parser/docx_parser.py

@@ -7,7 +7,7 @@
 from io import BytesIO
 
 
-class HuDocxParser:
+class RAGFlowDocxParser:
 
     def __extract_table_content(self, tb):
         df = []

deepdoc/parser/excel_parser.py

@@ -6,7 +6,7 @@
 from rag.nlp import find_codec
 
 
-class HuExcelParser:
+class RAGFlowExcelParser:
     def html(self, fnm):
         if isinstance(fnm, str):
             wb = load_workbook(fnm)
@@ -74,5 +74,5 @@ def row_number(fnm, binary):
 
 
 if __name__ == "__main__":
-    psr = HuExcelParser()
+    psr = RAGFlowExcelParser()
     psr(sys.argv[1])

deepdoc/parser/pdf_parser.py

@@ -23,7 +23,7 @@
 logging.getLogger("pdfminer").setLevel(logging.WARNING)
 
 
-class HuParser:
+class RAGFlowPdfParser:
     def __init__(self):
         self.ocr = OCR()
         if hasattr(self, "model_speciess"):

deepdoc/parser/ppt_parser.py

@@ -14,7 +14,7 @@
 from pptx import Presentation
 
 
-class HuPptParser(object):
+class RAGFlowPptParser(object):
     def __init__(self):
         super().__init__()
 

deepdoc/vision/t_ocr.py

@@ -11,10 +11,6 @@
 #  limitations under the License.
 #
 
-from deepdoc.vision.seeit import draw_box
-from deepdoc.vision import OCR, init_in_out
-import argparse
-import numpy as np
 import os
 import sys
 sys.path.insert(
@@ -25,6 +21,11 @@
                 os.path.abspath(__file__)),
             '../../')))
 
+from deepdoc.vision.seeit import draw_box
+from deepdoc.vision import OCR, init_in_out
+import argparse
+import numpy as np
+
 
 def main(args):
     ocr = OCR()

deepdoc/vision/t_recognizer.py

@@ -10,17 +10,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-
-from deepdoc.vision.seeit import draw_box
-from deepdoc.vision import Recognizer, LayoutRecognizer, TableStructureRecognizer, OCR, init_in_out
-from api.utils.file_utils import get_project_base_directory
-import argparse
-import os
-import sys
-import re
-
-import numpy as np
-
+import os, sys
 sys.path.insert(
     0,
     os.path.abspath(
@@ -29,6 +19,13 @@
                 os.path.abspath(__file__)),
             '../../')))
 
+from deepdoc.vision.seeit import draw_box
+from deepdoc.vision import Recognizer, LayoutRecognizer, TableStructureRecognizer, OCR, init_in_out
+from api.utils.file_utils import get_project_base_directory
+import argparse
+import re
+import numpy as np
+
 
 def main(args):
     images, outputs = init_in_out(args)

rag/llm/__init__.py

@@ -22,7 +22,7 @@
     "Ollama": OllamaEmbed,
     "OpenAI": OpenAIEmbed,
     "Xinference": XinferenceEmbed,
-    "Tongyi-Qianwen": HuEmbedding, #QWenEmbed,
+    "Tongyi-Qianwen": DefaultEmbedding, #QWenEmbed,
     "ZHIPU-AI": ZhipuEmbed,
     "FastEmbed": FastEmbed,
     "Youdao": YoudaoEmbed

rag/llm/embedding_model.py

@@ -56,7 +56,7 @@ def encode_queries(self, text: str):
         raise NotImplementedError("Please implement encode method!")
 
 
-class HuEmbedding(Base):
+class DefaultEmbedding(Base):
     def __init__(self, *args, **kwargs):
         """
         If you have trouble downloading HuggingFace models, -_^ this might help!!