Merge pull request #4305 from opendatalab/release-2.7.1

Release 2.7.1

README.md

@@ -45,6 +45,11 @@
 
 # Changelog
 
+- 2026/01/06 2.7.1 Release
+  - fix bug: #4300
+  - Updated pdfminer.six dependency version to resolve [CVE-2025-64512](https://github.com/advisories/GHSA-wf5f-4jwr-ppcp)
+  - Support automatic correction of input image exif orientation to improve OCR recognition accuracy  #4283
+
 - 2025/12/30 2.7.0 Release
   - Simplified installation process. No need to separately install `vlm` acceleration engine dependencies. Using `uv pip install mineru[all]` during installation will install all optional backend dependencies.
   - Added new `hybrid` backend, which combines the advantages of `pipeline` and `vlm` backends. Built on vlm, it integrates some capabilities of pipeline, adding extra extensibility on top of high accuracy:

README_zh-CN.md

@@ -45,6 +45,11 @@
 
 # 更新记录
 
+- 2026/01/06 2.7.1 发布
+  - fix bug: #4300
+  - 更新pdfminer.six的依赖版本以解决 [CVE-2025-64512](https://github.com/advisories/GHSA-wf5f-4jwr-ppcp)
+  - 支持输入图像的exif方向自动校正，提升OCR识别效果  #4283
+
 - 2025/12/30 2.7.0 发布
   - 简化安装流程，现在不再需要单独安装`vlm`加速引擎依赖包，安装时使用`uv pip install mineru[all]`即可安装所有可选后端的依赖包。
   - 增加全新后端`hybrid`，该后端结合了`pipeline`和`vlm`后端的优势，在vlm的基础上，融入了pipeline的部分能力，在高精度的基础上增加了额外的扩展性：

docs/zh/quick_start/index.md

@@ -1,4 +1,4 @@
-# 快速开始
+# 快速入门
 
 如果遇到任何安装问题，请先查询 [FAQ](../faq/index.md) 
 

docs/zh/usage/acceleration_cards/Ascend.md

@@ -105,65 +105,50 @@ docker run -u root --name mineru_docker --privileged=true \
   </thead>
   <tbody>
     <tr>
-      <td rowspan="4">命令行工具(mineru)</td>
+      <td rowspan="3">命令行工具(mineru)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-http-client</td>
-      <td>🟢</td>
-      <td>🟢</td>
-    </tr>
-    <tr>
-      <td rowspan="4">fastapi服务(mineru-api)</td>
+      <td rowspan="3">fastapi服务(mineru-api)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-http-client</td>
-      <td>🟢</td>
-      <td>🟢</td>
-    </tr>
-    <tr>
-      <td rowspan="4">gradio界面(mineru-gradio)</td>
+      <td rowspan="3">gradio界面(mineru-gradio)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
-      <td>🟢</td>
-      <td>🟢</td>
-    </tr>
-    <tr>
-      <td>vlm-http-client</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>

docs/zh/usage/acceleration_cards/METAX.md

@@ -82,65 +82,50 @@ docker run --ipc host \
   </thead>
   <tbody>
     <tr>
-      <td rowspan="4">命令行工具(mineru)</td>
+      <td rowspan="3">命令行工具(mineru)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
-      <td>🟡</td>
-      <td>🟡</td>
-    </tr>
-    <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-http-client</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td rowspan="4">fastapi服务(mineru-api)</td>
+      <td rowspan="3">fastapi服务(mineru-api)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
-      <td>🟡</td>
-      <td>🟡</td>
-    </tr>
-    <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-http-client</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td rowspan="4">gradio界面(mineru-gradio)</td>
+      <td rowspan="3">gradio界面(mineru-gradio)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
-      <td>🟡</td>
-      <td>🟡</td>
-    </tr>
-    <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-http-client</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>

docs/zh/usage/acceleration_cards/THead.md

@@ -73,65 +73,50 @@ docker run --privileged=true \
   </thead>
   <tbody>
     <tr>
-      <td rowspan="4">命令行工具(mineru)</td>
+      <td rowspan="3">命令行工具(mineru)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-http-client</td>
-      <td>🟢</td>
-      <td>🟢</td>
-    </tr>
-    <tr>
-      <td rowspan="4">fastapi服务(mineru-api)</td>
+      <td rowspan="3">fastapi服务(mineru-api)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-http-client</td>
-      <td>🟢</td>
-      <td>🟢</td>
-    </tr>
-    <tr>
-      <td rowspan="4">gradio界面(mineru-gradio)</td>
+      <td rowspan="3">gradio界面(mineru-gradio)</td>
       <td>pipeline</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>&lt;vlm/hybrid&gt;-auto-engine</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>
     <tr>
-      <td>vlm-&lt;engine_name&gt;-engine</td>
-      <td>🟢</td>
-      <td>🟢</td>
-    </tr>
-    <tr>
-      <td>vlm-http-client</td>
+      <td>&lt;vlm/hybrid&gt;-http-client</td>
       <td>🟢</td>
       <td>🟢</td>
     </tr>

docs/zh/usage/acceleration_cards/VastAI.md

@@ -14,21 +14,21 @@
     cpu: Hygon C86-4G
     gpu: VA16 / VA1L / VA10L
     torch: 2.8.0+cpu
-    torch-vacc: 1.3.3.626
+    torch-vacc: 1.3.3.777
     vllm: 0.11.1.dev0+gb8b302cde.d20251030.cpu
-    vllm-vacc: 0.11.0.626 
-    driver: 00.25.12.02 d3_3_v2_9_a3_1 3ef7cf3 20251202
+    vllm-vacc: 0.11.0.777
+    driver: 00.25.12.30 d3_3_v2_9_a3_1 a76bf37 20251230
     docker: 28.1.1
     ```
 
 ## 3. 环境准备
 
-- 获取Docker镜像
+- 获取vllm_vacc基础镜像
     ```bash
-    sudo docker pull harbor.vastaitech.com/ai_deliver/vllm_vacc:VVI-25.12.SP1
+    sudo docker pull harbor.vastaitech.com/ai_deliver/vllm_vacc:VVI-25.12.SP2
     ```
 
-- 启动Docker容器
+- 启动容器
     ```bash
     sudo docker run -it \
         --privileged=true \
@@ -36,23 +36,7 @@
         --name vllm_service \
         --ipc=host \
         --network=host \
-        harbor.vastaitech.com/ai_deliver/vllm_vacc:VVI-25.12.SP1 bash
-    ```
-
-
->[!TIP]
-> - 镜像内已包含`torch/vllm`等相关依赖
-> - 和`NVIDIA`硬件下`CUDA_VISIBLE_DEVICES`类似；在`VastAI`硬件中可以使用`VACC_VISIBLE_DEVICES`指定`可见计算卡ID`，如`-e VACC_VISIBLE_DEVICES=0,1,2,3`
-> - 需指定适当的`--shm-size`虚拟内存
-
-## 4. MinerU功能
-
->[!NOTE]
-> - `VastAI`加速卡仅支持使用`vlm-vllm-engine`和`vlm-http-client`形式进行`VLM`模型推理加速
-
-- 进入容器
-    ```bash
-    sudo docker exec -it vllm_service bash
+        harbor.vastaitech.com/ai_deliver/vllm_vacc:VVI-25.12.SP2 bash
     ```
 
 - 安装MinerU
@@ -60,6 +44,10 @@
    - 参考官方文档安装：[README_zh-CN.md#安装-mineru](https://github.com/opendatalab/MinerU/blob/master/README_zh-CN.md#安装-mineru)
 
         ```bash
+        # 启动容器
+        # sudo docker exec -it vllm_service bash
+        
+        # 可选pypi源
         # https://mirrors.163.com/pypi/simple/
         # https://mirrors.aliyun.com/pypi/simple/
         # https://pypi.mirrors.ustc.edu.cn/simple/
@@ -68,26 +56,42 @@
 
         # 通过源码安装MinerU
         git clone https://github.com/opendatalab/MinerU.git
-        git checkout eed479eb56bba93ee99c1a8c255d509bd2f837e5
+        git checkout 8c4b3ef3a20b11ddac9903f25124d24ea82639b5
         pip install -e .[core] -i https://mirrors.aliyun.com/pypi/simple
 
-        # 使用pip安装MinerU
-        pip install -U "mineru[core]" -i https://mirrors.aliyun.com/pypi/simple
+        # 或使用pip安装MinerU
+        pip install -U "mineru[core]==2.7.0" -i https://mirrors.aliyun.com/pypi/simple
         ```
 
+> [!NOTE]
+> - `vllm_vacc`基础镜像内已包含`torch/vllm`等相关依赖
+> - 截至`2025/12/31`，`VastAI`已支持`MinerU`至最新版本`2.7.0`，`master分支8c4b3ef3`
+> - 和`NVIDIA`硬件下`CUDA_VISIBLE_DEVICES`类似；在`VastAI`硬件中可以使用`VACC_VISIBLE_DEVICES`指定`可见计算卡ID`，如`-e VACC_VISIBLE_DEVICES=0,1,2,3`
+> - 需指定适当的`--shm-size`虚拟内存
+
+## 4. MinerU功能
+
+> [!NOTE]
+> - `VastAI`加速卡仅支持使用`vlm-auto-engine`和`vlm-http-client`形式进行`VLM`模型推理加速
+
+- 进入容器
+    ```bash
+    sudo docker exec -it vllm_service bash
+    ```
+
 - 使用MinerU
 
     - 模型准备，参考官方介绍：[model_source.md](https://github.com/opendatalab/MinerU/blob/master/docs/zh/usage/model_source.md)
 
-    - 方式一：`vlm-vllm-engine`
+    - 方式一：`vlm-auto-engine`
 
         ```bash
         export MINERU_MODEL_SOURCE=modelscope
 
-        # step1, 以`vlm-vllm-engine`方式启动MinerU解析任务
-        mineru -p /path/to/demo/pdfs/demo1.pdf \
+        # step1, 以`vlm-auto-engine`方式启动MinerU解析任务
+        mineru -p image.png \
         -o ./output \
-        -b vlm-vllm-engine \
+        -b vlm-auto-engine \
         --http-timeout 1200 \
         --tensor-parallel-size 2 \
         --enforce_eager \
@@ -108,7 +112,7 @@
         --served-model-name MinerU2.5-2509-1.2B
 
         # step2，以`vlm-http-client`方式启动MinerU解析任务
-        mineru -p /path/to/demo/pdfs/demo1.pdf \
+        mineru -p demo/pdfs/demo1.pdf \
         -o ./output \
         -b vlm-http-client \
         -u http://127.0.0.1:8090 \
@@ -116,8 +120,7 @@
         ```
 
 
->[!NOTE]
-> - 截至`2025/12/23`，`VastAI`已支持`MinerU`至最新版本`2.6.8`，`master分支eed479eb`
+> [!NOTE]
 > - 注意在执行任意与`vllm`相关命令需追加`--enforce_eager`参数
 
 
@@ -140,17 +143,17 @@
       <td>🔴</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>hybrid-http-client</td>
       <td>🔴</td>
     </tr>
     <tr>
-      <td>vlm-vllm-engine</td>
+      <td>hybrid-auto-engine</td>
+      <td>🔴</td>
+    </tr>
+    <tr>
+      <td>vlm-auto-engine</td>
       <td>🟢</td>
     </tr>
-    <tr>
-      <td>vlm-lmdeploy-client</td>
-      <td>🔴</td>
-    </tr>
     <tr>
       <td>vlm-http-client</td>
       <td>🟢</td>
@@ -161,17 +164,17 @@
       <td>🔴</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>hybrid-http-client</td>
       <td>🔴</td>
     </tr>
     <tr>
-      <td>vlm-vllm-engine</td>
+      <td>hybrid-auto-engine</td>
+      <td>🔴</td>
+    </tr>
+    <tr>
+      <td>vlm-auto-engine</td>
       <td>🟢</td>
     </tr>
-    <tr>
-      <td>vlm-lmdeploy-engine</td>
-      <td>🔴</td>
-    </tr>
     <tr>
       <td>vlm-http-client</td>
       <td>🟢</td>
@@ -182,17 +185,17 @@
       <td>🔴</td>
     </tr>
     <tr>
-      <td>vlm-transformers</td>
+      <td>hybrid-http-client</td>
       <td>🔴</td>
     </tr>
     <tr>
-      <td>vlm-vllm-engine</td>
+      <td>hybrid-auto-engine</td>
+      <td>🔴</td>
+    </tr>
+    <tr>
+      <td>vlm-auto-engine</td>
       <td>🟢</td>
     </tr>
-    <tr>
-      <td>vlm-lmdeploy-engine</td>
-      <td>🔴</td>
-    </tr>
     <tr>
       <td>vlm-http-client</td>
       <td>🟢</td>
@@ -202,18 +205,19 @@
       <td>🟢</td>
     </tr>
     <tr>
-      <td colspan="2">Tensor并行 (--tensor-parallel-size/--tp)</td>
-      <td>🟢 仅支持tp1/tp2</td>
+      <td colspan="2">Tensor并行 (--tensor-parallel-size)</td>
+      <td>🟢</td>
     </tr>
     <tr>
-      <td colspan="2">数据并行 (--data-parallel-size/--dp)</td>
+      <td colspan="2">数据并行 (--data-parallel-size)</td>
       <td>🔴</td>
     </tr>
   </tbody>
 </table>
 
 
-注：  
-🟢: 支持，运行较稳定，精度与Nvidia GPU基本一致  
-🟡: 支持但较不稳定，在某些场景下可能出现异常，或精度存在一定差异  
-🔴: 不支持，无法运行，或精度存在较大差异
+> [!NOTE]
+> - 🟢: 支持，运行较稳定，精度与NVIDIA GPU基本一致  
+> - 🟡: 支持但较不稳定，在某些场景下可能出现异常，或精度存在一定差异  
+> - 🔴: 不支持，无法运行，或精度存在较大差异
+> - `vlm-auto-engine`：VastAI仅支持vLLM后端

docs/zh/usage/index.md

@@ -4,10 +4,10 @@
 
 ## 目录
 - 本地部署
-    * [快速使用](./quick_usage.md) - 快速上手和基本使用
+    * [基础使用](./quick_usage.md) - 快速上手和基本使用
     * [模型源配置](./model_source.md) - 模型源的详细配置说明  
     * [命令行工具](./cli_tools.md) - 命令行工具的详细参数说明
-    * [进阶优化参数](./advanced_cli_parameters.md) - 一些适配命令行工具的进阶参数说明
+    * [命令行进阶参数](./advanced_cli_parameters.md) - 一些适配命令行工具的进阶参数说明
 - 其他加速卡适配（🚀官方支持/❤️社区贡献）
     * [昇腾 Ascend](acceleration_cards/Ascend.md) 🚀
     * [平头哥 T-Head](acceleration_cards/THead.md) 🚀

mineru/cli/common.py

@@ -17,8 +17,6 @@ from mineru.utils.pdf_image_tools import images_bytes_to_pdf_bytes
 from mineru.backend.vlm.vlm_middle_json_mkcontent import union_make as vlm_union_make
 from mineru.backend.vlm.vlm_analyze import doc_analyze as vlm_doc_analyze
 from mineru.backend.vlm.vlm_analyze import aio_doc_analyze as aio_vlm_doc_analyze
-from mineru.backend.hybrid.hybrid_analyze import doc_analyze as hybrid_doc_analyze
-from mineru.backend.hybrid.hybrid_analyze import aio_doc_analyze as aio_hybrid_doc_analyze
 from mineru.utils.pdf_page_id import get_end_page_id
 
 if os.getenv("MINERU_LMDEPLOY_DEVICE", "") == "maca":
@@ -326,6 +324,7 @@ def _process_hybrid(
         server_url=None,
         **kwargs,
 ):
+    from mineru.backend.hybrid.hybrid_analyze import doc_analyze as hybrid_doc_analyze
     """同步处理hybrid后端逻辑"""
     if not backend.endswith("client"):
         server_url = None
@@ -378,8 +377,8 @@ async def _async_process_hybrid(
         server_url=None,
         **kwargs,
 ):
+    from mineru.backend.hybrid.hybrid_analyze import aio_doc_analyze as aio_hybrid_doc_analyze
     """异步处理hybrid后端逻辑"""
-
     if not backend.endswith("client"):
         server_url = None
 

mineru/utils/pdf_image_tools.py

@@ -5,7 +5,7 @@ from io import BytesIO
 import numpy as np
 import pypdfium2 as pdfium
 from loguru import logger
-from PIL import Image
+from PIL import Image, ImageOps
 
 from mineru.data.data_reader_writer import FileBasedDataWriter
 from mineru.utils.check_sys_env import is_windows_environment
@@ -41,19 +41,23 @@ def pdf_page_to_image(page: pdfium.PdfPage, dpi=200, image_type=ImageType.PIL) -
     return image_dict
 
 
-def _load_images_from_pdf_worker(pdf_bytes, dpi, start_page_id, end_page_id, image_type):
+def _load_images_from_pdf_worker(
+    pdf_bytes, dpi, start_page_id, end_page_id, image_type
+):
     """用于进程池的包装函数"""
-    return load_images_from_pdf_core(pdf_bytes, dpi, start_page_id, end_page_id, image_type)
+    return load_images_from_pdf_core(
+        pdf_bytes, dpi, start_page_id, end_page_id, image_type
+    )
 
 
 def load_images_from_pdf(
-        pdf_bytes: bytes,
-        dpi=200,
-        start_page_id=0,
-        end_page_id=None,
-        image_type=ImageType.PIL,
-        timeout=None,
-        threads=4,
+    pdf_bytes: bytes,
+    dpi=200,
+    start_page_id=0,
+    end_page_id=None,
+    image_type=ImageType.PIL,
+    timeout=None,
+    threads=4,
 ):
     """带超时控制的 PDF 转图片函数,支持多进程加速
 
@@ -77,7 +81,7 @@ def load_images_from_pdf(
             dpi,
             start_page_id,
             get_end_page_id(end_page_id, len(pdf_doc)),
-            image_type
+            image_type,
         ), pdf_doc
     else:
         if timeout is None:
@@ -116,7 +120,7 @@ def load_images_from_pdf(
                     dpi,
                     range_start,
                     range_end,
-                    image_type
+                    image_type,
                 )
                 futures.append((range_start, future))
 
@@ -163,7 +167,14 @@ def load_images_from_pdf_core(
     return images_list
 
 
-def cut_image(bbox: tuple, page_num: int, page_pil_img, return_path, image_writer: FileBasedDataWriter, scale=2):
+def cut_image(
+    bbox: tuple,
+    page_num: int,
+    page_pil_img,
+    return_path,
+    image_writer: FileBasedDataWriter,
+    scale=2,
+):
     """从第page_num页的page中，根据bbox进行裁剪出一张jpg图片，返回图片路径 save_path：需要同时支持s3和本地,
     图片存放在save_path下，文件名是:
     {page_num}_{bbox[0]}_{bbox[1]}_{bbox[2]}_{bbox[3]}.jpg , bbox内数字取整。"""
@@ -197,7 +208,6 @@ def get_crop_img(bbox: tuple, pil_img, scale=2):
 
 
 def get_crop_np_img(bbox: tuple, input_img, scale=2):
-
     if isinstance(input_img, Image.Image):
         np_img = np.asarray(input_img)
     elif isinstance(input_img, np.ndarray):
@@ -212,17 +222,27 @@ def get_crop_np_img(bbox: tuple, input_img, scale=2):
         int(bbox[3] * scale),
     )
 
-    return np_img[scale_bbox[1]:scale_bbox[3], scale_bbox[0]:scale_bbox[2]]
+    return np_img[scale_bbox[1] : scale_bbox[3], scale_bbox[0] : scale_bbox[2]]
+
 
 def images_bytes_to_pdf_bytes(image_bytes):
     # 内存缓冲区
     pdf_buffer = BytesIO()
 
     # 载入并转换所有图像为 RGB 模式
-    image = Image.open(BytesIO(image_bytes)).convert("RGB")
+    image = Image.open(BytesIO(image_bytes))
+    # 根据 EXIF 信息自动转正（处理手机拍摄的带 Orientation 标记的图片）
+    image = ImageOps.exif_transpose(image) or image
+    # 只在必要时转换
+    if image.mode != "RGB":
+        image = image.convert("RGB")
 
     # 第一张图保存为 PDF，其余追加
-    image.save(pdf_buffer, format="PDF", save_all=True)
+    image.save(
+        pdf_buffer,
+        format="PDF",
+        # save_all=True
+    )
 
     # 获取 PDF bytes 并重置指针（可选）
     pdf_bytes = pdf_buffer.getvalue()

mineru/utils/table_merge.py

@@ -9,7 +9,14 @@ from mineru.utils.char_utils import full_to_half
 from mineru.utils.enum_class import BlockType, SplitFlag
 
 
-CONTINUATION_MARKERS = ["(续)", "(续表)", "(continued)", "(cont.)"]
+CONTINUATION_MARKERS = [
+    "(续)",
+    "(续表)",
+    "(续上表)",
+    "(continued)",
+    "(cont.)",
+    "(cont’d)",
+]
 
 
 def calculate_table_total_columns(soup):

mineru/version.py

@@ -1 +1 @@
-__version__ = "2.6.8"
+__version__ = "2.7.0"

mkdocs.yml

@@ -93,7 +93,18 @@ nav:
       - FAQ: faq/index.md
     - Demo:
       - Demo: demo/index.md
+  - Quick Start:
+    - Quick Start: quick_start/index.md
+    - Extension Modules: quick_start/extension_modules.md
+    - Docker Deployment: quick_start/docker_deployment.md
+  - Usage:
+    - Usage: usage/index.md
+    - Quick Usage: usage/quick_usage.md
+    - Model Source: usage/model_source.md
+    - CLI Tools: usage/cli_tools.md
+    - Advanced CLI Parameters: usage/advanced_cli_parameters.md
   - Reference:
+    - Reference: reference/index.md
     - Output File Format: reference/output_files.md
     - Changelog: reference/changelog.md
   - FAQ:
@@ -119,13 +130,13 @@ plugins:
           build: true
           nav_translations:
             Home: 主页
-            Quick Start: 快速开始
+            Quick Start: 快速入门
             Extension Modules: 扩展模块安装
             Docker Deployment: Docker部署
-            Usage: 使用方法
-            Quick Usage: 快速使用
+            Usage: 使用指南
+            Quick Usage: 基础使用
             CLI Tools: 命令行工具
-            Model Source: 模型源
+            Model Source: 模型源配置
             Advanced CLI Parameters: 命令行进阶参数
             FAQ: 常见问题解答
             Reference: 参考资料

pyproject.toml

@@ -21,7 +21,7 @@ dependencies = [
     "click>=8.1.7",
     "loguru>=0.7.2",
     "numpy>=1.21.6",
-    "pdfminer.six==20250506",
+    "pdfminer.six>=20251230",
     "tqdm>=4.67.1",
     "requests",
     "httpx",
@@ -94,10 +94,10 @@ core = [
     "mineru[pipeline]",
     "mineru[api]",
     "mineru[gradio]",
-    "mineru[mlx] ; sys_platform == 'darwin'",
 ]
 all = [
     "mineru[core]",
+    "mineru[mlx] ; sys_platform == 'darwin'",
     "mineru[vllm] ; sys_platform == 'linux'",
     "mineru[lmdeploy] ; sys_platform == 'windows'",
 ]