mistocr 0.1.0__py3-none-any.whl → 0.1.1__py3-none-any.whl

mistocr/__init__.py

@@ -1 +1 @@
-__version__ = "0.1.0"
+__version__ = "0.1.1"

mistocr/refine.py

@@ -63,9 +63,10 @@ Headings to analyze:
 {headings_list}
 """
 
-# %% ../nbs/01_refine.ipynb 16
+# %% ../nbs/01_refine.ipynb 18
 def fix_hdg_hierarchy(
     hdgs: list[str], # List of markdown headings
+    prompt: str=prompt_fix_hdgs, # Prompt to use
     model: str='claude-sonnet-4-5', # Model to use
     api_key: str=os.getenv('ANTHROPIC_API_KEY') # API key
     ) -> dict[int, str]: # Dictionary of index → corrected heading
@@ -78,7 +79,7 @@ def fix_hdg_hierarchy(
         )
     return json.loads(r.choices[0].message.content)['corrections']
 
-# %% ../nbs/01_refine.ipynb 19
+# %% ../nbs/01_refine.ipynb 21
 def mk_fixes_lut(
     hdgs: list[str], # List of markdown headings
     model: str='claude-sonnet-4-5', # Model to use
@@ -88,7 +89,7 @@ def mk_fixes_lut(
     fixes = fix_hdg_hierarchy(hdgs, model, api_key)
     return {hdgs[int(k)]:v for k,v in fixes.items()}
 
-# %% ../nbs/01_refine.ipynb 22
+# %% ../nbs/01_refine.ipynb 24
 def apply_hdg_fixes(
     p:str, # Page to fix
     lut_fixes: dict[str, str], # Lookup table of fixes
@@ -98,7 +99,7 @@ def apply_hdg_fixes(
     for old in get_hdgs(p): p = p.replace(old, lut_fixes.get(old, old) + (f' .... page {pg}' if pg else ''))
     return p
 
-# %% ../nbs/01_refine.ipynb 25
+# %% ../nbs/01_refine.ipynb 27
 def fix_md_hdgs(
     src:str, # Source directory with markdown pages
     model:str='claude-sonnet-4-5', # Model

{mistocr-0.1.0.dist-info → mistocr-0.1.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mistocr
-Version: 0.1.0
+Version: 0.1.1
 Summary: Simple batch OCR for PDFs using Mistral's state-of-the-art vision model
 Home-page: https://github.com/franckalbinet/mistocr
 Author: Solveit

mistocr-0.1.1.dist-info/RECORD

@@ -0,0 +1,10 @@
+mistocr/__init__.py,sha256=rnObPjuBcEStqSO0S6gsdS_ot8ITOQjVj_-P1LUUYpg,22
+mistocr/_modidx.py,sha256=R9zVMv4dKz2sLStoB5wBoKRqjza216_z8xPXszoplU4,2660
+mistocr/core.py,sha256=wtaYZ_Fz0dXIb1cYLAtymwR9Z7_KBI4ULy-UpM3YTX4,7154
+mistocr/refine.py,sha256=0N0omMZvPqydArPaiTdkX8tts4eS4AgfAJQP98WnvwY,4307
+mistocr-0.1.1.dist-info/licenses/LICENSE,sha256=xV8xoN4VOL0uw9X8RSs2IMuD_Ss_a9yAbtGNeBWZwnw,11337
+mistocr-0.1.1.dist-info/METADATA,sha256=ZPvywkk_QS7UTfPtup5IppiLzf5aZ1tCSPkZLV9g4WE,4848
+mistocr-0.1.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+mistocr-0.1.1.dist-info/entry_points.txt,sha256=HjAiHozobM-alm_6bTF-ehRr2DD3KYE9PgRngelONOY,36
+mistocr-0.1.1.dist-info/top_level.txt,sha256=LelTYnSpSXLy1Hb1m2YP3gt8luwP-I8KV0NjP_ucdSs,8
+mistocr-0.1.1.dist-info/RECORD,,

mistocr-0.1.0.dist-info/RECORD

@@ -1,10 +0,0 @@
-mistocr/__init__.py,sha256=kUR5RAFc7HCeiqdlX36dZOHkUI5wI6V_43RpEcD8b-0,22
-mistocr/_modidx.py,sha256=R9zVMv4dKz2sLStoB5wBoKRqjza216_z8xPXszoplU4,2660
-mistocr/core.py,sha256=wtaYZ_Fz0dXIb1cYLAtymwR9Z7_KBI4ULy-UpM3YTX4,7154
-mistocr/refine.py,sha256=gWup79LGjmvKW5RyY1dRKUeAEt94mUJIeTZB3V4D-JE,4258
-mistocr-0.1.0.dist-info/licenses/LICENSE,sha256=xV8xoN4VOL0uw9X8RSs2IMuD_Ss_a9yAbtGNeBWZwnw,11337
-mistocr-0.1.0.dist-info/METADATA,sha256=JOyUQONpYUmmGk2kFzMkxaIBrHwjC9CfmI7fc9qa6ms,4848
-mistocr-0.1.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-mistocr-0.1.0.dist-info/entry_points.txt,sha256=HjAiHozobM-alm_6bTF-ehRr2DD3KYE9PgRngelONOY,36
-mistocr-0.1.0.dist-info/top_level.txt,sha256=LelTYnSpSXLy1Hb1m2YP3gt8luwP-I8KV0NjP_ucdSs,8
-mistocr-0.1.0.dist-info/RECORD,,