Procházet zdrojové kódy

feat: 添加对B用户扫描流水的支持,更新数据源配置

zhch158_admin před 1 týdnem
rodič
revize
0562a5f7b1
1 změnil soubory, kde provedl 15 přidání a 3 odebrání
  1. 15 3
      table_line_generator/table_line_generator.yaml

+ 15 - 3
table_line_generator/table_line_generator.yaml

@@ -28,7 +28,22 @@ table_editor:
         rgb: [255, 0, 0]
 
   data_sources:
+    - name: "B用户_扫描流水"
+      tool: "mineru"  # 格式同 MinerU
+      base_dir: "/Users/zhch158/workspace/data/流水分析"
+      json_dir: "{{name}}/mineru_vllm_results_cell_bbox"
+      image_dir: "{{name}}/mineru_vllm_results/{{name}}"
+      # {{ name }}_page_(?P<page>\d{3})\.json 来匹配 JSON 文件名。模板变量 {{ name }} 会在运行时被替换成实际前缀;随后 _page_ 是字面固定部分。命名捕获组 (?P<page>\d{3}) 强制页面编号恰好为三位数字,并将其存入 page 组  
+      json_pattern: "{{ name }}_page_(?P<page>\\d{3})\\.json"
+      image_pattern: "{{ name }}_page_(?P<page>\\d{3})\\.png"
+      sort_key: "page"
+      output:
+        directory: "{{base_dir}}/{{ name }}.wiredtable"
+        structure_suffix: "_structure.json"
+        image_suffix: ".png"
+
     - name: "康强_北京农村商业银行"
+      tool: "ppstructv3"
       base_dir: "/Users/zhch158/workspace/data/流水分析"
       json_dir: "{{name}}/ppstructurev3_client_results"
       image_dir: "{{name}}/ppstructurev3_client_results/{{name}}"
@@ -40,6 +55,3 @@ table_editor:
         directory: "{{base_dir}}/{{ name }}.wiredtable"
         structure_suffix: "_structure.json"
         image_suffix: ".png"
-    # - name: "示例文档"
-    #   json_dir: "../demo/json"
-    #   image_dir: "../demo/img"