vlm4ocr 0.0.1__py3-none-any.whl → 0.2.0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (15) hide show
  1. vlm4ocr/__init__.py +3 -1
  2. vlm4ocr/assets/default_prompt_templates/ocr_HTML_system_prompt.txt +1 -0
  3. vlm4ocr/assets/default_prompt_templates/ocr_HTML_user_prompt.txt +1 -0
  4. vlm4ocr/assets/default_prompt_templates/ocr_text_user_prompt.txt +1 -0
  5. vlm4ocr/cli.py +367 -0
  6. vlm4ocr/data_types.py +109 -0
  7. vlm4ocr/ocr_engines.py +359 -195
  8. vlm4ocr/utils.py +328 -18
  9. vlm4ocr/vlm_engines.py +317 -191
  10. {vlm4ocr-0.0.1.dist-info → vlm4ocr-0.2.0.dist-info}/METADATA +4 -2
  11. vlm4ocr-0.2.0.dist-info/RECORD +16 -0
  12. vlm4ocr-0.2.0.dist-info/entry_points.txt +3 -0
  13. vlm4ocr-0.0.1.dist-info/RECORD +0 -10
  14. /vlm4ocr/assets/default_prompt_templates/{ocr_user_prompt.txt → ocr_markdown_user_prompt.txt} +0 -0
  15. {vlm4ocr-0.0.1.dist-info → vlm4ocr-0.2.0.dist-info}/WHEEL +0 -0
{vlm4ocr-0.0.1.dist-info → vlm4ocr-0.2.0.dist-info}/METADATA RENAMED
@@ -1,7 +1,7 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: vlm4ocr
3
- Version: 0.0.1
4
- Summary: OCR with vision language models.
3
+ Version: 0.2.0
4
+ Summary: Python package and Web App for OCR with vision language models.
5
5
  License: MIT
6
6
  Author: Enshuo (David) Hsu
7
7
  Requires-Python: >=3.11,<4.0
@@ -9,8 +9,10 @@ Classifier: License :: OSI Approved :: MIT License
9
9
  Classifier: Programming Language :: Python :: 3
10
10
  Classifier: Programming Language :: Python :: 3.11
11
11
  Classifier: Programming Language :: Python :: 3.12
12
+ Provides-Extra: tesseract
12
13
  Requires-Dist: pdf2image (>=1.16.0)
13
14
  Requires-Dist: pillow (>=10.0.0)
15
+ Requires-Dist: pytesseract (>=0.3.13) ; extra == "tesseract"
14
16
  Description-Content-Type: text/markdown
15
17
 
16
18
  This is the readme for vlm4ocr Python package.
vlm4ocr-0.2.0.dist-info/RECORD ADDED
@@ -0,0 +1,16 @@
1
+ vlm4ocr/__init__.py,sha256=k5TZY0LmRnjGyjHD0H5AxJHJMw_cS2SzGxTJ0NQbQsc,315
2
+ vlm4ocr/assets/default_prompt_templates/ocr_HTML_system_prompt.txt,sha256=igPOntiLDZXTB71-QrTmMJveb6XC1TgArg1serPc9V8,547
3
+ vlm4ocr/assets/default_prompt_templates/ocr_HTML_user_prompt.txt,sha256=cVn538JojZfCtIhfrcOPWt0dO7dtDqgB9xdS_5VvAqo,41
4
+ vlm4ocr/assets/default_prompt_templates/ocr_markdown_system_prompt.txt,sha256=pIsYO2G3jkZ5EWg7MJixre3Itz1oPqJSduUZT34_RNY,436
5
+ vlm4ocr/assets/default_prompt_templates/ocr_markdown_user_prompt.txt,sha256=61EJv8POsQGIIUVwCjDU73lMXJE7F3qhPIYl6zSbl1Q,45
6
+ vlm4ocr/assets/default_prompt_templates/ocr_text_system_prompt.txt,sha256=WbLSOerqFjlYGaGWJ-w2enhky1WhnPl011s0fgRPgnQ,398
7
+ vlm4ocr/assets/default_prompt_templates/ocr_text_user_prompt.txt,sha256=ftgNAIPy_UlrcY6m7-IkH2ApHkCzRnymra1w2wg60Ks,47
8
+ vlm4ocr/cli.py,sha256=b13WswreFxTNLA7n2F2jPR7Wrb2Onb06zFnvf7MOLi0,20268
9
+ vlm4ocr/data_types.py,sha256=OnbI5IFonp5jPUgq0RIHHSzR9EypJ3jaGoPxtbBpS04,3919
10
+ vlm4ocr/ocr_engines.py,sha256=KDU70U-SWkFUZR5fV0vKlnP_9HE0pE5ghzU7ihdTMjc,23605
11
+ vlm4ocr/utils.py,sha256=mwp3YDQaoS7YObXtLeuinSEGQ0fJ5KeQMAZmONIKvcg,11907
12
+ vlm4ocr/vlm_engines.py,sha256=jQuRZ5HlJtTtJXESiFcoYQXwX-lYu0gc-KKOpRLuW6A,22331
13
+ vlm4ocr-0.2.0.dist-info/METADATA,sha256=bpANMPUizxAnWDgrUO3bqNFHnlK72FnCX-7Q7x2bSpA,638
14
+ vlm4ocr-0.2.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
15
+ vlm4ocr-0.2.0.dist-info/entry_points.txt,sha256=qzWUk_QTZ12cH4DLjjfqce89EAlOydD85dreRRZF3K4,44
16
+ vlm4ocr-0.2.0.dist-info/RECORD,,
vlm4ocr-0.2.0.dist-info/entry_points.txt ADDED
@@ -0,0 +1,3 @@
1
+ [console_scripts]
2
+ vlm4ocr=vlm4ocr.cli:main
3
+
vlm4ocr-0.0.1.dist-info/RECORD DELETED
@@ -1,10 +0,0 @@
1
- vlm4ocr/__init__.py,sha256=gD7WuuxH9_iUu8dgbfonRQD89xiveVP3e9YP4RAHeIQ,219
2
- vlm4ocr/assets/default_prompt_templates/ocr_markdown_system_prompt.txt,sha256=pIsYO2G3jkZ5EWg7MJixre3Itz1oPqJSduUZT34_RNY,436
3
- vlm4ocr/assets/default_prompt_templates/ocr_text_system_prompt.txt,sha256=WbLSOerqFjlYGaGWJ-w2enhky1WhnPl011s0fgRPgnQ,398
4
- vlm4ocr/assets/default_prompt_templates/ocr_user_prompt.txt,sha256=61EJv8POsQGIIUVwCjDU73lMXJE7F3qhPIYl6zSbl1Q,45
5
- vlm4ocr/ocr_engines.py,sha256=gNwN_itHqIQLF3cfMoGV51lAuY8ZJI2A78LWJivqVi4,13687
6
- vlm4ocr/utils.py,sha256=3o6TLY8YIEorYHluHrNoJesEM8th89uRO_KNEzfdDA8,1763
7
- vlm4ocr/vlm_engines.py,sha256=3Aymjo3xa-CTHbktOIHSeLSLAdbo7fk9qRw8d3uRmOM,18836
8
- vlm4ocr-0.0.1.dist-info/METADATA,sha256=_8E0__RQ6bJy_WLlkirJRvf3G8LqmG5_WtOF1ddBaZw,520
9
- vlm4ocr-0.0.1.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
10
- vlm4ocr-0.0.1.dist-info/RECORD,,