mistocr 0.3.1__tar.gz → 0.3.2__tar.gz

{mistocr-0.3.1/mistocr.egg-info → mistocr-0.3.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mistocr
-Version: 0.3.1
+Version: 0.3.2
 Summary: Batch OCR for PDFs with heading restoration and visual content integration
 Home-page: https://github.com/franckalbinet/mistocr
 Author: Solveit

mistocr-0.3.2/mistocr/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.3.2"

{mistocr-0.3.1 → mistocr-0.3.2}/mistocr/_modidx.py

@@ -21,7 +21,7 @@ d = { 'settings': { 'branch': 'main',
                               'mistocr.core.subset_pdf': ('core.html#subset_pdf', 'mistocr/core.py'),
                               'mistocr.core.upload_pdf': ('core.html#upload_pdf', 'mistocr/core.py'),
                               'mistocr.core.wait_for_job': ('core.html#wait_for_job', 'mistocr/core.py')},
-            'mistocr.pipeline': {},
+            'mistocr.pipeline': {'mistocr.pipeline.pdf_to_md': ('pipeline.html#pdf_to_md', 'mistocr/pipeline.py')},
             'mistocr.refine': { 'mistocr.refine.HeadingCorrection': ('refine.html#headingcorrection', 'mistocr/refine.py'),
                                 'mistocr.refine.HeadingCorrections': ('refine.html#headingcorrections', 'mistocr/refine.py'),
                                 'mistocr.refine.ImgDescription': ('refine.html#imgdescription', 'mistocr/refine.py'),

mistocr-0.3.2/mistocr/pipeline.py

@@ -0,0 +1,41 @@
+"""End-to-End Pipeline: PDF OCR, Markdown Heading Correction, and AI Image Descriptions"""
+
+# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/02_pipeline.ipynb.
+
+# %% auto 0
+__all__ = ['pdf_to_md']
+
+# %% ../nbs/02_pipeline.ipynb 3
+from fastcore.all import *
+from .core import read_pgs, ocr_pdf
+from .refine import add_img_descs, fix_hdgs
+from pathlib import Path
+from asyncio import Semaphore, gather, sleep
+import tempfile
+import os, json, shutil
+
+# %% ../nbs/02_pipeline.ipynb 4
+@delegates(add_img_descs)
+async def pdf_to_md(
+    pdf_path:str,                   # Path to input PDF file
+    dst:str,                        # Destination directory for output markdown
+    ocr_dst:str=None,               # Optional OCR output directory
+    model:str='claude-sonnet-4-5',  # Model to use for heading fixes and image descriptions
+    add_img_desc:bool=True,         # Whether to add image descriptions
+    progress:bool=True,             # Whether to show progress messages
+    **kwargs
+    ):
+    "Convert PDF to markdown with OCR, fixed heading hierarchy, and optional image descriptions"
+    cleanup = ocr_dst is None
+    if cleanup: ocr_dst = tempfile.mkdtemp()
+    n_steps = 3 if add_img_desc else 2
+    if progress: print(f"Step 1/{n_steps}: Running OCR on {pdf_path}...")
+    ocr_dir = ocr_pdf(pdf_path, ocr_dst)[0]
+    if progress: print(f"Step 2/{n_steps}: Fixing heading hierarchy...")
+    fix_hdgs(ocr_dir, model=model)
+    if add_img_desc:
+        if progress: print(f"Step 3/{n_steps}: Adding image descriptions...")
+        await add_img_descs(ocr_dir, dst=dst, model=model, progress=progress, **kwargs)
+    elif dst != str(ocr_dir): shutil.copytree(ocr_dir, dst, dirs_exist_ok=True)
+    if cleanup: shutil.rmtree(ocr_dst)
+    if progress: print("Done!")

{mistocr-0.3.1 → mistocr-0.3.2/mistocr.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mistocr
-Version: 0.3.1
+Version: 0.3.2
 Summary: Batch OCR for PDFs with heading restoration and visual content integration
 Home-page: https://github.com/franckalbinet/mistocr
 Author: Solveit

{mistocr-0.3.1 → mistocr-0.3.2}/settings.ini

@@ -1,7 +1,7 @@
 [DEFAULT]
 repo = mistocr
 lib_name = mistocr
-version = 0.3.1
+version = 0.3.2
 min_python = 3.9
 license = apache2
 black_formatting = False

mistocr-0.3.1/mistocr/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.3.1"

mistocr-0.3.1/mistocr/pipeline.py

@@ -1,15 +0,0 @@
-"""End-to-End Pipeline: PDF OCR, Markdown Heading Correction, and AI Image Descriptions"""
-
-# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/02_pipeline.ipynb.
-
-# %% auto 0
-__all__ = []
-
-# %% ../nbs/02_pipeline.ipynb 3
-from fastcore.all import *
-from .core import read_pgs, ocr_pdf
-from .refine import add_img_descs, fix_hdgs
-from pathlib import Path
-from asyncio import Semaphore, gather, sleep
-import tempfile
-import os, json, shutil