v0.7.0

Author: hzkitty

demo.py

@@ -43,8 +43,8 @@ def do_parse(
     f_dump_model_output=True,  # Whether to dump model output files
     f_dump_orig_pdf=True,  # Whether to dump original PDF files
     f_dump_content_list=True,  # Whether to dump content list files
-    f_dump_md_html=True,  # Whether to convert markdown to HTML
-    f_dump_md_docx=True,  # Whether to convert markdown to docx (via Pandoc)
+    f_dump_md_html=False,  # Whether to convert markdown to HTML
+    f_dump_md_docx=False,  # Whether to convert markdown to docx (via Pandoc)
     f_make_md_mode=MakeMode.MM_MD,  # The mode for making markdown content, default is MM_MD
     start_page_id=0,  # Start page ID for parsing, default is 0
     end_page_id=None,  # End page ID for parsing, default is None (parse all pages until the end of the document)

docker/Dockerfile

@@ -21,8 +21,8 @@ RUN apt-get update && \
 WORKDIR /app
 
 # 安装 Python 依赖
-RUN python3 -m pip install 'rapid-doc[cpu]==0.6.1' -i https://pypi.org/simple --break-system-packages && \
-    python3 -m pip install 'rapid-doc[api]==0.6.1' -i https://pypi.org/simple --break-system-packages && \
+RUN python3 -m pip install 'rapid-doc[cpu]==0.7.0' -i https://pypi.org/simple --break-system-packages && \
+    python3 -m pip install 'rapid-doc[api]==0.7.0' -i https://pypi.org/simple --break-system-packages && \
     python3 -m pip cache purge
 
 # 复制配置文件和脚本（优先复制，利用Docker缓存）

docker/DockerfileGPU

@@ -29,7 +29,7 @@ WORKDIR /app
 
 # 安装 Python 依赖
 RUN python3 -m pip install --upgrade pip setuptools wheel && \
-    python3 -m pip install 'rapid-doc[gpu]==0.6.1' 'rapid-doc[api]==0.6.1' -i https://pypi.org/simple && \
+    python3 -m pip install 'rapid-doc[gpu]==0.7.0' 'rapid-doc[api]==0.7.0' -i https://pypi.org/simple && \
     python3 -m pip install 'onnxruntime-gpu==1.23.0' -i https://pypi.org/simple && \
     python3 -m pip cache purge
 

docker/README.md

@@ -12,10 +12,10 @@
 cd docker
 
 # 1. CPU 模式
-docker build -f Dockerfile -t hzkitty/rapid-doc:0.6.1 .
+docker build -f Dockerfile -t hzkitty/rapid-doc:0.7.0 .
 
 # 2. GPU 模式
-docker build -f DockerfileGPU -t hzkitty/rapid-doc:0.6.1-gpu .
+docker build -f DockerfileGPU -t hzkitty/rapid-doc:0.7.0-gpu .
 ```
 
 

docker/docker-compose-gpu.yml

@@ -1,7 +1,7 @@
 services:
   rapid-doc-server:
     container_name: rapid-doc-server
-    image: hzkitty/rapid-doc:0.6.1-gpu
+    image: hzkitty/rapid-doc:0.7.0-gpu
     deploy:
       resources:
         reservations:

docker/docker-compose.yml

@@ -1,7 +1,7 @@
 services:
   rapid-doc-server:
     container_name: rapid-doc-server
-    image: hzkitty/rapid-doc:0.6.1
+    image: hzkitty/rapid-doc:0.7.0
     ports:
       - "8888:8888"
     environment:

rapid_doc/backend/pipeline/model_init.py

@@ -86,13 +86,13 @@ def atom_model_init(model_name: str, **kwargs):
             kwargs.get('layout_config'),
         )
     elif model_name == AtomicModel.FORMULA:
-        atom_model = kwargs.get('formula_config').get('custom_model')
+        atom_model = (kwargs.get('formula_config') or {}).get('custom_model')
         if not isinstance(atom_model, CustomBaseModel):
             atom_model = formula_model_init(
                 kwargs.get('formula_config'),
             )
     elif model_name == AtomicModel.OCR:
-        atom_model = kwargs.get('ocr_config').get('custom_model')
+        atom_model = (kwargs.get('ocr_config') or {}).get('custom_model')
         if not isinstance(atom_model, CustomBaseModel):
             atom_model = ocr_model_init(
                 kwargs.get('det_db_box_thresh', 0.3),
@@ -102,7 +102,7 @@ def atom_model_init(model_name: str, **kwargs):
                 kwargs.get('enable_merge_det_boxes', True)
             )
     elif model_name == AtomicModel.Table:
-        atom_model = kwargs.get('table_config').get('custom_model')
+        atom_model = (kwargs.get('table_config') or {}).get('custom_model')
         if not isinstance(atom_model, CustomBaseModel):
             atom_model = table_model_init(
                 kwargs.get('lang'),

rapid_doc/model/table/rapid_table_self/table_structure/unet/utils/utils_table_line_rec.py

@@ -237,7 +237,9 @@ def min_area_rect_box(
     """
     boxes = []
     for region in regions:
-        bbox_area = getattr(region, "area_bbox", region.bbox_area)
+        bbox_area = getattr(region, "area_bbox", None)
+        if bbox_area is None:
+            bbox_area = region.bbox_area
         if bbox_area  > H * W * 3 / 4:  # 过滤大的单元格
             continue
         rect = cv2.minAreaRect(region.coords[:, ::-1])

rapid_doc/version.py

@@ -1,2 +1,2 @@
-__version__ = "0.6.1"
+__version__ = "0.7.0"
 __mineru_version__ = "2.6.4"