magic-pdf 1.3.6__py3-none-any.whl → 1.3.7__py3-none-any.whl

magic_pdf/libs/version.py

@@ -1 +1 @@
-__version__ = "1.3.6"
+__version__ = "1.3.7"

magic_pdf/model/batch_analyze.py

@@ -161,20 +161,13 @@ class BatchAnalyze:
             for table_res_dict in tqdm(table_res_list_all_page, desc="Table Predict"):
                 _lang = table_res_dict['lang']
                 atom_model_manager = AtomModelSingleton()
-                ocr_engine = atom_model_manager.get_atom_model(
-                    atom_model_name='ocr',
-                    ocr_show_log=False,
-                    det_db_box_thresh=0.5,
-                    det_db_unclip_ratio=1.6,
-                    lang=_lang
-                )
                 table_model = atom_model_manager.get_atom_model(
                     atom_model_name='table',
                     table_model_name='rapid_table',
                     table_model_path='',
                     table_max_time=400,
                     device='cpu',
-                    ocr_engine=ocr_engine,
+                    lang=_lang,
                     table_sub_model_name='slanet_plus'
                 )
                 html_code, table_cell_bboxes, logic_points, elapse = table_model.predict(table_res_dict['table_img'])

magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorch_paddle.py

@@ -53,6 +53,11 @@ class PytorchPaddleOCR(TextSystem):
         args = parser.parse_args(args)
 
         self.lang = kwargs.get('lang', 'ch')
+
+        device = get_device()
+        if device == 'cpu' and self.lang == 'ch':
+            self.lang = 'ch_lite'
+
         if self.lang in latin_lang:
             self.lang = 'latin'
         elif self.lang in arabic_lang:

{magic_pdf-1.3.6.dist-info → magic_pdf-1.3.7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: magic-pdf
-Version: 1.3.6
+Version: 1.3.7
 Summary: A practical tool for converting PDF to Markdown
 License: AGPL-3.0
 Project-URL: Home, https://mineru.net/
@@ -107,6 +107,9 @@ Easier to use: Just grab MinerU Desktop. No coding, no login, just a simple inte
 </div>
 
 # Changelog
+- 2025/04/22 1.3.7 Released
+  - Fixed the issue where the `lang` parameter was ineffective during table parsing model initialization.
+  - Fixed the significant slowdown in OCR and table parsing speed in `cpu` mode.
 - 2025/04/16 1.3.4 Released
   - Slightly improved the speed of OCR detection by removing some unused blocks.
   - Fixed page-level sorting errors caused by footnotes in certain cases.

{magic_pdf-1.3.6.dist-info → magic_pdf-1.3.7.dist-info}/RECORD

@@ -52,9 +52,9 @@ magic_pdf/libs/pdf_check.py,sha256=7GWWvDR6g_rj_fE6XJlbTq5AFVX11ngRIzT0N18F214,3
 magic_pdf/libs/pdf_image_tools.py,sha256=_au7plmKKctpPKozBumSKgP8689q4vH1mU8VMLO0IbM,2260
 magic_pdf/libs/performance_stats.py,sha256=DW-c6nUTUnWKGTONRKfpucsYZm1ake016F9K7jJwbik,2136
 magic_pdf/libs/safe_filename.py,sha256=ckwcM_eqoysTb5id8czp-tXq2G9da0-l3pshZDCHQtE,236
-magic_pdf/libs/version.py,sha256=5ZbAQtod5QalTI1C2N07edlxplzG_Q2XvGOSyOok4uA,22
+magic_pdf/libs/version.py,sha256=9peaXOar2qezOPJEKG6cD_A0aaXrzdVN8h-v6fBoBEk,22
 magic_pdf/model/__init__.py,sha256=sa-dO2k-TLy25I2gRrzjm_cQeYfzMf-pLwBJHkIxGo0,51
-magic_pdf/model/batch_analyze.py,sha256=yKhKQuZTh9GG83p61bw2BRqKMbnsjsmX73gfuTRk8xE,11272
+magic_pdf/model/batch_analyze.py,sha256=F0WsjbQ6z9txdiUiVy6n6zhyJWJ-4moljNx8fe8HFws,10977
 magic_pdf/model/doc_analyze_by_custom_model.py,sha256=-cjn7DQi6kZCqVZ0IxbXuL2kmeGhSVLzLaezIHPFzMU,10317
 magic_pdf/model/magic_model.py,sha256=yZKWo_wRck_-YLyFGRiUHGar8sV1Y6458BFLbyBAt74,30682
 magic_pdf/model/model_list.py,sha256=aqfEJlEfbib3D3ISrxc0Coh6SbffYh8Yq2FlQN35_zA,213
@@ -109,7 +109,7 @@ magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_swin/modeling_unim
 magic_pdf/model/sub_modules/ocr/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/__init__.py,sha256=8CRrCQVuExa0BttRFh3Z40lFy2K5jN0sp67KWjOlj5c,50
 magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/ocr_utils.py,sha256=3qxu0lAjqzZQ2Ci-C_wz_YSakyq_5-KnckA3-5bICTM,12589
-magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorch_paddle.py,sha256=FaPo02L1IJKybGYfydsohOiHstJIL8d5UKzGck2tYvk,7283
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorch_paddle.py,sha256=zCQ9cv8yBwTNELakqYXb3KJ0buj7tocByOY4xjXOG4U,7399
 magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/base_ocr_v20.py,sha256=5bI7MAu65r-vn28krwdJ6pjZMkEvWjspE7EQaTsRERw,1319
 magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/data/__init__.py,sha256=YYu3c-W4fgEErxxDM98uQ3oWwPEh-6w75LY4zcj4VtM,199
@@ -195,9 +195,9 @@ magic_pdf/tools/common.py,sha256=-x0RSFr7SNbdYq7DntaLYmQmaxyF-xKSf4xMpSUTzA0,126
 magic_pdf/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 magic_pdf/utils/annotations.py,sha256=82ou3uELNbQWa9hOFFkVt0gsIskAKf5msCv5J2IJ5V0,211
 magic_pdf/utils/office_to_pdf.py,sha256=bFRYe6v3-pfx5R8-bV8cmf12jPnOrYZsleKoECTXzbM,3958
-magic_pdf-1.3.6.dist-info/LICENSE.md,sha256=jVa0BUaKrRH4erV2P5AeJ24I2WRv9chIGxditreJ6e0,34524
-magic_pdf-1.3.6.dist-info/METADATA,sha256=WciKDOUh2xH6OdTqDPJlMlLrUhaHBseO5dycx0_7RLo,45798
-magic_pdf-1.3.6.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
-magic_pdf-1.3.6.dist-info/entry_points.txt,sha256=wXwYke3j8fqDQTocUspL-CqDUEv3Tfcwp09fM8dZAhA,98
-magic_pdf-1.3.6.dist-info/top_level.txt,sha256=J9I0AzmHWGkp9c6DL8Oe4mEx3yYphLzkRn4H25Lg1rE,10
-magic_pdf-1.3.6.dist-info/RECORD,,
+magic_pdf-1.3.7.dist-info/LICENSE.md,sha256=jVa0BUaKrRH4erV2P5AeJ24I2WRv9chIGxditreJ6e0,34524
+magic_pdf-1.3.7.dist-info/METADATA,sha256=9JPxH4h9dqV4n6TU7L__FCFCDEGf495v5X4uN4rvXMs,46013
+magic_pdf-1.3.7.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
+magic_pdf-1.3.7.dist-info/entry_points.txt,sha256=wXwYke3j8fqDQTocUspL-CqDUEv3Tfcwp09fM8dZAhA,98
+magic_pdf-1.3.7.dist-info/top_level.txt,sha256=J9I0AzmHWGkp9c6DL8Oe4mEx3yYphLzkRn4H25Lg1rE,10
+magic_pdf-1.3.7.dist-info/RECORD,,