Merge pull request #164 from breezedeus/dev

fix: model downloads on Windows

.gitignore

@@ -8,12 +8,22 @@ __pycache__/
 
 # Distribution / packaging
 .Python
+.DS_Store
+data/
+models/
+output-*/
+outputs-*/
+outputs/
+*.jpg
+*.jpeg
+*.png
 build/
 develop-eggs/
 dist/
 downloads/
 eggs/
 .eggs/
+.idea/
 lib/
 lib64/
 parts/

docs/RELEASE.md

@@ -1,5 +1,16 @@
 # Release Notes
 
+# Update 2024.12.17: **V1.1.2.3** Released
+
+Major Changes:
+
+- Bugfix: Fixed issues related to downloading models on Windows.
+
+主要变更：
+
+- 修复了在 Windows 环境下下载模型的问题。
+
+
 # Update 2024.12.11: **V1.1.2.2** Released
 
 Major Changes:

pix2text/__version__.py

@@ -2,4 +2,4 @@
 # [Pix2Text](https://github.com/breezedeus/pix2text): an Open-Source Alternative to Mathpix.
 # Copyright (C) 2022-2024, [Breezedeus](https://www.breezedeus.com).
 
-__version__ = '1.1.2.2'
+__version__ = '1.1.2.3'

pix2text/doc_xl_layout/doc_xl_layout_parser.py

@@ -28,6 +28,7 @@
     box2list,
     x_overlap,
     merge_boxes,
+    prepare_model_files2,
 )
 
 logger = logging.getLogger(__name__)
@@ -124,15 +125,11 @@ def _prepare_model_files(self, root, model_info):
         model_fp = model_dir / 'DocXLayout_231012.pth'
         if model_fp.exists():
             return model_fp
-        if model_dir.exists():
-            shutil.rmtree(str(model_dir))
-        model_dir.mkdir(parents=True)
-        download_cmd = f'huggingface-cli download --repo-type model --resume-download --local-dir-use-symlinks False breezedeus/pix2text-layout --local-dir {model_dir}'
-        os.system(download_cmd)
-        if not model_fp.exists():  # download failed above
-            if model_dir.exists():
-                shutil.rmtree(str(model_dir))
-            os.system('HF_ENDPOINT=https://hf-mirror.com ' + download_cmd)
+        model_fp = prepare_model_files2(
+            model_fp_or_dir=model_fp,
+            remote_repo="breezedeus/pix2text-layout",
+            file_or_dir="file",
+        )
         return model_fp
 
     def convert_eval_format(self, all_bboxes, opt):