mistocr 0.2.12__tar.gz → 0.3.1__tar.gz

{mistocr-0.2.12/mistocr.egg-info → mistocr-0.3.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mistocr
-Version: 0.2.12
+Version: 0.3.1
 Summary: Batch OCR for PDFs with heading restoration and visual content integration
 Home-page: https://github.com/franckalbinet/mistocr
 Author: Solveit
@@ -23,6 +23,7 @@ Requires-Dist: mistralai
 Requires-Dist: pillow
 Requires-Dist: dotenv
 Requires-Dist: lisette
+Requires-Dist: PyPDF2
 Provides-Extra: dev
 Dynamic: author
 Dynamic: author-email

mistocr-0.3.1/mistocr/__init__.py

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

{mistocr-0.2.12 → mistocr-0.3.1}/mistocr/_modidx.py

@@ -18,9 +18,10 @@ d = { 'settings': { 'branch': 'main',
                               'mistocr.core.save_page': ('core.html#save_page', 'mistocr/core.py'),
                               'mistocr.core.save_pages': ('core.html#save_pages', 'mistocr/core.py'),
                               'mistocr.core.submit_batch': ('core.html#submit_batch', 'mistocr/core.py'),
+                              '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.pdf_to_md': ('pipeline.html#pdf_to_md', 'mistocr/pipeline.py')},
+            'mistocr.pipeline': {},
             '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-0.2.12 → mistocr-0.3.1}/mistocr/core.py

@@ -4,7 +4,8 @@
 
 # %% auto 0
 __all__ = ['ocr_model', 'ocr_endpoint', 'get_api_key', 'upload_pdf', 'create_batch_entry', 'prep_pdf_batch', 'submit_batch',
-           'wait_for_job', 'download_results', 'save_images', 'save_page', 'save_pages', 'ocr_pdf', 'read_pgs']
+           'wait_for_job', 'download_results', 'save_images', 'save_page', 'save_pages', 'ocr_pdf', 'read_pgs',
+           'subset_pdf']
 
 # %% ../nbs/00_core.ipynb 3
 from fastcore.all import *
@@ -13,6 +14,7 @@ from io import BytesIO
 from pathlib import Path
 from PIL import Image
 from mistralai import Mistral
+import PyPDF2
 
 # %% ../nbs/00_core.ipynb 6
 def get_api_key(
@@ -181,7 +183,7 @@ def ocr_pdf(
     results = _run_batch(entries, c, poll_interval)
     return L([save_pages(r['response']['body'], dst, r['custom_id']) for r in results])
 
-# %% ../nbs/00_core.ipynb 47
+# %% ../nbs/00_core.ipynb 50
 def read_pgs(
     path:str, # OCR output directory, 
     join:bool=True # Join pages into single string
@@ -191,3 +193,24 @@ def read_pgs(
     pgs = sorted(path.glob('page_*.md'), key=lambda p: int(p.stem.split('_')[1]))
     contents = L([p.read_text() for p in pgs])
     return '\n\n'.join(contents) if join else contents
+
+# %% ../nbs/00_core.ipynb 57
+def subset_pdf(
+    path:str, # Path to PDF file
+    start:int=1, # Start page (1-based)
+    end:int=None, # End page (1-based, inclusive)
+    dst:str='.' # Output directory
+    ) -> Path: # Path to subset PDF
+    "Extract page range from PDF and save with range suffix"
+    path = Path(path)
+    writer = PyPDF2.PdfWriter()
+    with open(path, 'rb') as f:
+        reader = PyPDF2.PdfReader(f)
+        n = len(reader.pages)
+        end = end or n
+        s, e = max(0, start-1), min(n, end) - 1
+        for i in range(s, e+1): writer.add_page(reader.pages[i])
+    suffix = f"_p{s+1}-{e+1}" if s>0 or e<n-1 else ""
+    out = Path(dst) / f"{path.stem}{suffix}.pdf"
+    with open(out, 'wb') as f: writer.write(f)
+    return out

mistocr-0.3.1/mistocr/pipeline.py

@@ -0,0 +1,15 @@
+"""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

{mistocr-0.2.12 → mistocr-0.3.1/mistocr.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mistocr
-Version: 0.2.12
+Version: 0.3.1
 Summary: Batch OCR for PDFs with heading restoration and visual content integration
 Home-page: https://github.com/franckalbinet/mistocr
 Author: Solveit
@@ -23,6 +23,7 @@ Requires-Dist: mistralai
 Requires-Dist: pillow
 Requires-Dist: dotenv
 Requires-Dist: lisette
+Requires-Dist: PyPDF2
 Provides-Extra: dev
 Dynamic: author
 Dynamic: author-email

{mistocr-0.2.12 → mistocr-0.3.1}/mistocr.egg-info/requires.txt

@@ -3,5 +3,6 @@ mistralai
 pillow
 dotenv
 lisette
+PyPDF2
 
 [dev]

{mistocr-0.2.12 → mistocr-0.3.1}/settings.ini

@@ -1,7 +1,7 @@
 [DEFAULT]
 repo = mistocr
 lib_name = mistocr
-version = 0.2.12
+version = 0.3.1
 min_python = 3.9
 license = apache2
 black_formatting = False
@@ -27,7 +27,7 @@ keywords = nbdev jupyter notebook python
 language = English
 status = 3
 user = franckalbinet
-requirements = fastcore mistralai pillow dotenv lisette
+requirements = fastcore mistralai pillow dotenv lisette PyPDF2
 readme_nb = index.ipynb
 allowed_metadata_keys = 
 allowed_cell_metadata_keys = 

mistocr-0.2.12/mistocr/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.2.12"

mistocr-0.2.12/mistocr/pipeline.py

@@ -1,37 +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__ = ['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 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_output:str=None, # Optional OCR output directory (defaults to pdf_path stem)
-    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"
-    n_steps = 3 if add_img_desc else 2
-    if progress: print(f"Step 1/{n_steps}: Running OCR on {pdf_path}...")
-    ocr_dirs = ocr_pdf(pdf_path, ocr_output or 'ocr_temp')
-    ocr_dir = ocr_dirs[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 and Path(dst) != ocr_dir: shutil.copytree(ocr_dir, dst, dirs_exist_ok=True)
-    if progress: print("Done!")