natural-pdf 0.2.17__py3-none-any.whl → 0.2.18__py3-none-any.whl

natural_pdf/core/page.py

@@ -1,5 +1,6 @@
 import base64
 import concurrent.futures  # Added import
+import contextlib
 import hashlib
 import io
 import json
@@ -275,6 +276,9 @@ class Page(
         self._load_elements()
         self._to_image_cache: Dict[tuple, Optional["Image.Image"]] = {}
 
+        # Flag to prevent infinite recursion when computing exclusions
+        self._computing_exclusions = False
+
     def _get_render_specs(
         self,
         mode: Literal["show", "render"] = "show",
@@ -412,6 +416,35 @@ class Page(
         self._exclusions = []
         return self
 
+    @contextlib.contextmanager
+    def without_exclusions(self):
+        """
+        Context manager that temporarily disables exclusion processing.
+
+        This prevents infinite recursion when exclusion callables themselves
+        use find() operations. While in this context, all find operations
+        will skip exclusion filtering.
+
+        Example:
+            ```python
+            # This exclusion would normally cause infinite recursion:
+            page.add_exclusion(lambda p: p.find("text:contains('Header')").expand())
+
+            # But internally, it's safe because we use:
+            with page.without_exclusions():
+                region = exclusion_callable(page)
+            ```
+
+        Yields:
+            The page object with exclusions temporarily disabled.
+        """
+        old_value = self._computing_exclusions
+        self._computing_exclusions = True
+        try:
+            yield self
+        finally:
+            self._computing_exclusions = old_value
+
     def add_exclusion(
         self,
         exclusion_func_or_region: Union[
@@ -759,15 +792,10 @@ class Page(
                     if debug:
                         print(f"  - Evaluating callable '{exclusion_label}'...")
 
-                    # Temporarily clear exclusions (consider if really needed)
-                    temp_original_exclusions = self._exclusions
-                    self._exclusions = []
-
-                    # Call the function - Expects it to return a Region or None
-                    region_result = exclusion_item(self)
-
-                    # Restore exclusions
-                    self._exclusions = temp_original_exclusions
+                    # Use context manager to prevent infinite recursion
+                    with self.without_exclusions():
+                        # Call the function - Expects it to return a Region or None
+                        region_result = exclusion_item(self)
 
                     if isinstance(region_result, Region):
                         # Assign the label to the returned region
@@ -947,6 +975,11 @@ class Page(
         Returns:
             A new list containing only the elements not excluded.
         """
+        # Skip exclusion filtering if we're currently computing exclusions
+        # This prevents infinite recursion when exclusion callables use find operations
+        if self._computing_exclusions:
+            return elements
+
         # Check both page-level and PDF-level exclusions
         has_page_exclusions = bool(self._exclusions)
         has_pdf_exclusions = (

natural_pdf/utils/spatial.py

@@ -10,11 +10,11 @@ with include_boundaries='none'.
 
 Example:
     from natural_pdf.utils.spatial import is_element_in_region
-    
+
     # Check if element is in region using center-based logic (default)
     if is_element_in_region(element, region):
         print("Element is in region")
-    
+
     # Use different strategies
     if is_element_in_region(element, region, strategy="intersects"):
         print("Element overlaps with region")
@@ -35,16 +35,16 @@ InclusionStrategy = Literal["center", "intersects", "contains"]
 
 def is_element_in_region(
     element: "Element",
-    region: "Region", 
+    region: "Region",
     strategy: InclusionStrategy = "center",
-    check_page: bool = True
+    check_page: bool = True,
 ) -> bool:
     """
     Unified function to check if an element is inside a region.
-    
+
     This centralizes the logic used across Region, Page, and Flow to ensure
     consistent behavior throughout the library.
-    
+
     Args:
         element: The element to check
         region: The region to check against
@@ -53,7 +53,7 @@ def is_element_in_region(
             - "intersects": Element belongs if any part overlaps
             - "contains": Element belongs only if fully contained
         check_page: Whether to verify element and region are on the same page
-        
+
     Returns:
         bool: True if element is in region according to the strategy
     """
@@ -61,18 +61,18 @@ def is_element_in_region(
     if not hasattr(element, "bbox") or not element.bbox:
         logger.debug(f"Element lacks bbox attributes: {element}")
         return False
-        
+
     if not hasattr(region, "bbox") or not region.bbox:
         logger.debug(f"Region lacks bbox attributes: {region}")
         return False
-    
+
     # Check page membership if requested
     if check_page:
         if not hasattr(element, "page") or not hasattr(region, "page"):
             return False
         if element.page != region.page:
             return False
-    
+
     # Apply the appropriate strategy
     if strategy == "center":
         # Use existing region method if available
@@ -82,37 +82,43 @@ def is_element_in_region(
             # Fallback calculation
             elem_center_x = (element.x0 + element.x1) / 2
             elem_center_y = (element.top + element.bottom) / 2
-            
+
             # Use region's is_point_inside if available
             if hasattr(region, "is_point_inside"):
                 return region.is_point_inside(elem_center_x, elem_center_y)
             else:
                 # Simple bounds check
-                return (region.x0 <= elem_center_x <= region.x1 and 
-                       region.top <= elem_center_y <= region.bottom)
-    
+                return (
+                    region.x0 <= elem_center_x <= region.x1
+                    and region.top <= elem_center_y <= region.bottom
+                )
+
     elif strategy == "intersects":
         # Use existing region method if available
         if hasattr(region, "intersects"):
             return region.intersects(element)
         else:
             # Simple bbox overlap check
-            return not (element.x1 < region.x0 or 
-                       element.x0 > region.x1 or 
-                       element.bottom < region.top or 
-                       element.top > region.bottom)
-    
+            return not (
+                element.x1 < region.x0
+                or element.x0 > region.x1
+                or element.bottom < region.top
+                or element.top > region.bottom
+            )
+
     elif strategy == "contains":
         # Use existing region method if available
         if hasattr(region, "contains"):
             return region.contains(element)
         else:
             # Simple full containment check
-            return (region.x0 <= element.x0 and 
-                   element.x1 <= region.x1 and 
-                   region.top <= element.top and 
-                   element.bottom <= region.bottom)
-    
+            return (
+                region.x0 <= element.x0
+                and element.x1 <= region.x1
+                and region.top <= element.top
+                and element.bottom <= region.bottom
+            )
+
     else:
         raise ValueError(f"Unknown inclusion strategy: {strategy}")
 
@@ -120,10 +126,10 @@ def is_element_in_region(
 def get_inclusion_strategy() -> InclusionStrategy:
     """
     Get the current global inclusion strategy.
-    
+
     This could be made configurable via environment variable or settings.
     For now, returns the default strategy.
-    
+
     Returns:
         The current inclusion strategy (default: "center")
     """
@@ -132,38 +138,35 @@ def get_inclusion_strategy() -> InclusionStrategy:
     return "center"
 
 
-def calculate_element_overlap_percentage(
-    element: "Element",
-    region: "Region"
-) -> float:
+def calculate_element_overlap_percentage(element: "Element", region: "Region") -> float:
     """
     Calculate what percentage of an element overlaps with a region.
-    
+
     Args:
         element: The element to check
         region: The region to check against
-        
+
     Returns:
         float: Percentage of element area that overlaps with region (0.0 to 1.0)
     """
     if not hasattr(element, "bbox") or not hasattr(region, "bbox"):
         return 0.0
-    
+
     # Calculate intersection bounds
     intersect_x0 = max(element.x0, region.x0)
-    intersect_y0 = max(element.top, region.top) 
+    intersect_y0 = max(element.top, region.top)
     intersect_x1 = min(element.x1, region.x1)
     intersect_y1 = min(element.bottom, region.bottom)
-    
+
     # Check if there's an intersection
     if intersect_x1 <= intersect_x0 or intersect_y1 <= intersect_y0:
         return 0.0
-    
+
     # Calculate areas
     element_area = (element.x1 - element.x0) * (element.bottom - element.top)
     if element_area == 0:
         return 0.0
-        
+
     intersect_area = (intersect_x1 - intersect_x0) * (intersect_y1 - intersect_y0)
-    
-    return intersect_area / element_area
+
+    return intersect_area / element_area

{natural_pdf-0.2.17.dist-info → natural_pdf-0.2.18.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: natural-pdf
-Version: 0.2.17
+Version: 0.2.18
 Summary: A more intuitive interface for working with PDFs
 Author-email: Jonathan Soma <jonathan.soma@gmail.com>
 License-Expression: MIT

{natural_pdf-0.2.17.dist-info → natural_pdf-0.2.18.dist-info}/RECORD

@@ -27,7 +27,7 @@ natural_pdf/collections/mixins.py,sha256=Se2C5AcpP9B5E0d0pIrey6-f_P32tAXTK4M7666
 natural_pdf/core/__init__.py,sha256=QC8H4M3KbXwMFiQORZ0pdPlzx1Ix6oKKQSS7Ib2KEaA,38
 natural_pdf/core/element_manager.py,sha256=619R97OtMd7uhaax7fZNJmhy9GxSs9HCNP4OzGgP828,55882
 natural_pdf/core/highlighting_service.py,sha256=wEV-koqHoHf7S3wZ3j8D2L-ucGp3Nd0YhhStz9yqeLc,70406
-natural_pdf/core/page.py,sha256=-0OaIoXz0zjT_jnPjjI2jpb8vvNKh-1W56auA5UBhTA,158791
+natural_pdf/core/page.py,sha256=-78LuCbU9AEd4MGMm7_yoBl9rMAvOvrbPWcVsrMoe0s,159986
 natural_pdf/core/page_collection.py,sha256=bLZ3TqTQbmP3oYrbfEi7HUoPMbcGplEtUMZ3Z1y7fuw,66728
 natural_pdf/core/page_groupby.py,sha256=V2e_RNlHaasUzYm2h2vNJI7_aV_fl3_pg7kU3F2j0z8,8218
 natural_pdf/core/pdf.py,sha256=i8dYCimL_k5FV6BmPI1a2Dk7XZfwLP8TziXr2n3O_fI,105639
@@ -101,7 +101,7 @@ natural_pdf/utils/packaging.py,sha256=TM0jafwS5yVbTGC-RMi4TyWunf9cUUo9h5J6rMzkT-
 natural_pdf/utils/pdfminer_patches.py,sha256=Ob81OMoNUGMUIy9nMw3deSQ_Z6cQmhbRlHUC3EHw2jk,4201
 natural_pdf/utils/reading_order.py,sha256=u7XyVZdKMPMK0CL1C7xFogKnZ92b0JKT068KFjQWe18,7437
 natural_pdf/utils/sections.py,sha256=HZX7829-fquKgIF7vUN2tL10-aXckEaM25g_2VcgWU4,12941
-natural_pdf/utils/spatial.py,sha256=JOH2LHnF5WBDcjNQsHQdj458zwUgKtSWW7Tj0motn70,5968
+natural_pdf/utils/spatial.py,sha256=YjzGO4A013ZDYGYbs7hl4RbJOaKrg8-x__Dl_BamwUA,5908
 natural_pdf/utils/text_extraction.py,sha256=CCwPTmMoTgtQt2P00X_ADIf6ZGNfxvjCO9FO0_HqG40,13900
 natural_pdf/utils/visualization.py,sha256=zhZEHgYnZFuX7YxTHXF8Y3D97uHp2beTKMaC-JkCFwk,22364
 natural_pdf/vision/__init__.py,sha256=TkoQtdODlh0n_99dsjLIWKE9dgK0m4jfrui_cQ3gTwU,221
@@ -111,7 +111,7 @@ natural_pdf/vision/similarity.py,sha256=HWmXDBNLSOlRWH-_1K3FVR7tSsRuMFqXZwrVhhg2
 natural_pdf/vision/template_matching.py,sha256=91XQt5tp-vmcMX_4b2Bz-YwIAlb-hc8E5ih_qAHQuCk,7145
 natural_pdf/widgets/__init__.py,sha256=QTVaUmsw__FCweFYZebwPssQxxUFUMd0wpm_cUbGZJY,181
 natural_pdf/widgets/viewer.py,sha256=KW3JogdR2TMg2ECUMYp8hwd060hfg8EsYBWxb5IEzBY,24942
-natural_pdf-0.2.17.dist-info/licenses/LICENSE,sha256=9zfwINwJlarbDmdh6iJV4QUG54QSJlSAUcnC1YiC_Ns,1074
+natural_pdf-0.2.18.dist-info/licenses/LICENSE,sha256=9zfwINwJlarbDmdh6iJV4QUG54QSJlSAUcnC1YiC_Ns,1074
 optimization/memory_comparison.py,sha256=0i_foFSRmppj-fY069qjwH36s_zkx-1L2ASAAlepWzA,6541
 optimization/pdf_analyzer.py,sha256=HjrmTgu2qchxPeDckc5kjgxppGwd40UESrYS9Myj7pY,19352
 optimization/performance_analysis.py,sha256=JBXnR9hc7Ix7YCnt3EJPSpsyqIUgKsc7GEffQ_TDCBk,13033
@@ -148,8 +148,8 @@ tools/bad_pdf_eval/llm_enrich.py,sha256=mCh4KGi1HmIkzGjj5rrHz1Osd7sEX1IZ_FW08H1t
 tools/bad_pdf_eval/llm_enrich_with_retry.py,sha256=XUtPF1hUvqd3frDXT0wDTXoonuAivhjM5vgFdZ-tm0A,9373
 tools/bad_pdf_eval/reporter.py,sha256=e1g__mkSB4q02p3mGWOwMhvFs7F2HJosNBxup0-LkyU,400
 tools/bad_pdf_eval/utils.py,sha256=hR95XQ7qf7Cu6BdyX0L7ggGVx-ah5sK0jHWblTJUUic,4896
-natural_pdf-0.2.17.dist-info/METADATA,sha256=8K5PCwh_OuI8vkWRLChHeT-LuEd0sRmigkRm55ZNeDo,6960
-natural_pdf-0.2.17.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-natural_pdf-0.2.17.dist-info/entry_points.txt,sha256=1R_KMv7g60UBBpRqGfw7bppsMNGdayR-iJlb9ohEk_8,81
-natural_pdf-0.2.17.dist-info/top_level.txt,sha256=ZDKhxE_tg508o9BpagsjCGcI8GY4cF_8bg0e0IaLsPI,41
-natural_pdf-0.2.17.dist-info/RECORD,,
+natural_pdf-0.2.18.dist-info/METADATA,sha256=_hwRZyYPDD_bl-dRHE2KLo8oeo2TPxVxGi66grA-ZIs,6960
+natural_pdf-0.2.18.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+natural_pdf-0.2.18.dist-info/entry_points.txt,sha256=1R_KMv7g60UBBpRqGfw7bppsMNGdayR-iJlb9ohEk_8,81
+natural_pdf-0.2.18.dist-info/top_level.txt,sha256=ZDKhxE_tg508o9BpagsjCGcI8GY4cF_8bg0e0IaLsPI,41
+natural_pdf-0.2.18.dist-info/RECORD,,