Merge pull request #4304 from opendatalab/dev

Dev

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：

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

@@ -232,13 +232,17 @@ def images_bytes_to_pdf_bytes(image_bytes):
     # 载入并转换所有图像为 RGB 模式
     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）
-    ImageOps.exif_transpose(image, in_place=True)
+    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()

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'",
 ]