natural-pdf 0.1.33__py3-none-any.whl → 0.1.35__py3-none-any.whl

natural_pdf/ocr/ocr_options.py

@@ -1,12 +1,7 @@
 # ocr_options.py
-import logging
 from dataclasses import dataclass, field
 from typing import Any, Dict, List, Optional, Tuple, Union
 
-# Configure logging
-# logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
-# logger = logging.getLogger(__name__)
-# Assume logger is configured elsewhere or remove if not needed globally
 
 
 # --- Base Options ---
@@ -58,8 +53,6 @@ class EasyOCROptions(BaseOCROptions):
     add_margin: float = 0.1
     output_format: str = "standard"
 
-    # def __post_init__(self):
-    #     logger.debug(f"Initialized EasyOCROptions: {self}")
 
 
 # --- PaddleOCR Specific Options ---
@@ -95,8 +88,8 @@ class PaddleOCROptions(BaseOCROptions):
 
     # Detection parameters (can be overridden at predict time)
     # https://github.com/PaddlePaddle/PaddleOCR/issues/15424
-    text_det_limit_side_len: Optional[int] = 736 # WAITING FOR FIX
-    text_det_limit_type: Optional[str] = 'max' # WAITING FOR FIX
+    text_det_limit_side_len: Optional[int] = 736  # WAITING FOR FIX
+    text_det_limit_type: Optional[str] = "max"  # WAITING FOR FIX
     text_det_thresh: Optional[float] = None
     text_det_box_thresh: Optional[float] = None
     text_det_unclip_ratio: Optional[float] = None
@@ -113,7 +106,7 @@ class PaddleOCROptions(BaseOCROptions):
     enable_hpi: Optional[bool] = None
     use_tensorrt: Optional[bool] = None
     precision: Optional[str] = None
-    enable_mkldnn: Optional[bool] = False # https://github.com/PaddlePaddle/PaddleOCR/issues/15294
+    enable_mkldnn: Optional[bool] = False  # https://github.com/PaddlePaddle/PaddleOCR/issues/15294
     # mkldnn_cache_capacity: Optional[int] = None
     cpu_threads: Optional[int] = None
     paddlex_config: Optional[str] = None

natural_pdf/qa/document_qa.py

@@ -9,6 +9,7 @@ import numpy as np
 from PIL import Image, ImageDraw
 
 from natural_pdf.elements.collections import ElementCollection
+
 from .qa_result import QAResult
 
 logger = logging.getLogger("natural_pdf.qa.document_qa")
@@ -252,13 +253,17 @@ class DocumentQA:
             # Save per-question result in debug mode
             if debug:
                 # File names: debug_qa_result_0.json, …
-                result_path = os.path.join(debug_output_dir, f"debug_qa_result_{q[:30].replace(' ', '_')}.json")
+                result_path = os.path.join(
+                    debug_output_dir, f"debug_qa_result_{q[:30].replace(' ', '_')}.json"
+                )
                 try:
                     with open(result_path, "w") as f:
                         serializable = {
                             k: (
                                 str(v)
-                                if not isinstance(v, (str, int, float, bool, list, dict, type(None)))
+                                if not isinstance(
+                                    v, (str, int, float, bool, list, dict, type(None))
+                                )
                                 else v
                             )
                             for k, v in top_res.items()
@@ -317,9 +322,9 @@ class DocumentQA:
             warnings.warn(
                 f"No text elements found on page {page.index}. "
                 "Consider applying OCR first using page.apply_ocr() to extract text from images.",
-                UserWarning
+                UserWarning,
             )
-            
+
             # Return appropriate "not found" result(s)
             if isinstance(question, (list, tuple)):
                 return [
@@ -376,7 +381,11 @@ class DocumentQA:
                     start_idx = res.start
                     end_idx = res.end
 
-                    if elements and 0 <= start_idx < len(word_boxes) and 0 <= end_idx < len(word_boxes):
+                    if (
+                        elements
+                        and 0 <= start_idx < len(word_boxes)
+                        and 0 <= end_idx < len(word_boxes)
+                    ):
                         matched_texts = [wb[0] for wb in word_boxes[start_idx : end_idx + 1]]
 
                         source_elements = []
@@ -426,9 +435,9 @@ class DocumentQA:
             warnings.warn(
                 f"No text elements found in region on page {region.page.index}. "
                 "Consider applying OCR first using region.apply_ocr() to extract text from images.",
-                UserWarning
+                UserWarning,
             )
-            
+
             # Return appropriate "not found" result(s)
             if isinstance(question, (list, tuple)):
                 return [
@@ -488,7 +497,11 @@ class DocumentQA:
                     start_idx = res.start
                     end_idx = res.end
 
-                    if elements and 0 <= start_idx < len(word_boxes) and 0 <= end_idx < len(word_boxes):
+                    if (
+                        elements
+                        and 0 <= start_idx < len(word_boxes)
+                        and 0 <= end_idx < len(word_boxes)
+                    ):
                         matched_texts = [wb[0] for wb in word_boxes[start_idx : end_idx + 1]]
 
                         source_elements = []

natural_pdf/qa/qa_result.py

@@ -24,13 +24,9 @@ class QAResult(dict):
         """
         source = self.get("source_elements")
         if source is None:
-            raise AttributeError(
-                "QAResult does not contain 'source_elements'; nothing to show()."
-            )
+            raise AttributeError("QAResult does not contain 'source_elements'; nothing to show().")
         if not hasattr(source, "show"):
-            raise AttributeError(
-                "'source_elements' object has no 'show' method; cannot visualise."
-            )
+            raise AttributeError("'source_elements' object has no 'show' method; cannot visualise.")
         return source.show(*args, **kwargs)
 
     # ------------------------------------------------------------------
@@ -52,4 +48,4 @@ class QAResult(dict):
 
     # Ensure ``copy`` keeps the subclass type
     def copy(self):
-        return QAResult(self) 
+        return QAResult(self)

natural_pdf/search/__init__.py

@@ -18,9 +18,9 @@ SEARCH_DEPENDENCIES_AVAILABLE = False
 
 try:
     import numpy as np
+
     # Lazy import for sentence_transformers to avoid heavy loading at module level
     # import sentence_transformers
-
     # Basic search dependencies are available
     SEARCH_DEPENDENCIES_AVAILABLE = True
 
@@ -51,6 +51,7 @@ def _check_sentence_transformers():
     """Lazy check for sentence_transformers availability."""
     try:
         import sentence_transformers
+
         return True
     except ImportError:
         return False
@@ -63,7 +64,7 @@ def check_search_availability():
             "Search functionality requires 'lancedb' and pyarrow. "
             "Install with: pip install natural-pdf[search] (or pip install lancedb pyarrow)"
         )
-    
+
     # Lazy check for sentence_transformers when actually needed
     if not _check_sentence_transformers():
         raise ImportError(

natural_pdf/search/lancedb_search_service.py

@@ -7,15 +7,13 @@ from typing import Any, Dict, Iterable, List, Optional, Union
 
 import lancedb
 import pyarrow as pa
+
+from .search_options import BaseSearchOptions
+from .search_service_protocol import Indexable, IndexConfigurationError, SearchServiceProtocol
+
 # Lazy import for SentenceTransformer to avoid heavy loading at module level
 # from sentence_transformers import SentenceTransformer
 
-from .search_options import BaseSearchOptions
-from .search_service_protocol import (
-    Indexable,
-    IndexConfigurationError,
-    SearchServiceProtocol,
-)
 
 logger = logging.getLogger(__name__)
 
@@ -26,6 +24,7 @@ DEFAULT_LANCEDB_PERSIST_PATH = "./lancedb_data"
 def _get_sentence_transformer(model_name: str):
     """Lazy import and instantiation of SentenceTransformer."""
     from sentence_transformers import SentenceTransformer
+
     return SentenceTransformer(model_name)
 
 

natural_pdf/search/numpy_search_service.py

@@ -6,8 +6,6 @@ from pathlib import Path
 from typing import Any, Dict, Iterable, List, Optional, Union
 
 import numpy as np
-# Lazy import for SentenceTransformer to avoid heavy loading at module level
-# from sentence_transformers import SentenceTransformer
 
 from .search_options import BaseSearchOptions
 from .search_service_protocol import (
@@ -17,6 +15,10 @@ from .search_service_protocol import (
     SearchServiceProtocol,
 )
 
+# Lazy import for SentenceTransformer to avoid heavy loading at module level
+# from sentence_transformers import SentenceTransformer
+
+
 logger = logging.getLogger(__name__)
 
 DEFAULT_EMBEDDING_MODEL = "all-MiniLM-L6-v2"
@@ -25,6 +27,7 @@ DEFAULT_EMBEDDING_MODEL = "all-MiniLM-L6-v2"
 def _get_sentence_transformer(model_name: str):
     """Lazy import and instantiation of SentenceTransformer."""
     from sentence_transformers import SentenceTransformer
+
     return SentenceTransformer(model_name)
 
 

natural_pdf/selectors/parser.py

@@ -1,5 +1,29 @@
-"""
-CSS-like selector parser for natural-pdf.
+"""CSS-like selector parser for natural-pdf.
+
+This module implements a sophisticated selector parsing system that enables
+jQuery-style element selection in PDF documents. It supports complex CSS-like
+selectors with extensions for PDF-specific attributes and spatial relationships.
+
+The parser handles:
+- Basic element selectors (text, rect, line, image)
+- Attribute selectors with comparisons ([size>12], [color="red"])
+- Pseudo-selectors for text content (:contains(), :regex())
+- Spatial relationship selectors (:above(), :below(), :near())
+- Color matching with Delta E distance calculations
+- Logical operators (AND, OR) and grouping
+- Complex nested expressions with proper precedence
+
+Key features:
+- Safe value parsing without eval() for security
+- Color parsing from multiple formats (hex, RGB, names, CSS functions)
+- Font and style attribute matching
+- Coordinate and dimension-based selections
+- Performance-optimized filtering functions
+
+This enables powerful document navigation like:
+- page.find('text[size>12]:bold:contains("Summary")')
+- page.find_all('rect[color~="red"]:above(text:contains("Total"))')
+- page.find('text:regex("[0-9]{4}-[0-9]{2}-[0-9]{2}")')
 """
 
 import ast
@@ -16,14 +40,35 @@ logger = logging.getLogger(__name__)
 
 
 def safe_parse_value(value_str: str) -> Any:
-    """
-    Safely parse a value string without using eval().
+    """Safely parse a value string without using eval().
+
+    Parses various value formats commonly found in PDF attributes while maintaining
+    security by avoiding eval(). Supports numbers, tuples, lists, booleans, and
+    quoted strings with proper type conversion.
 
     Args:
-        value_str: String representation of a value (number, tuple, string, etc.)
+        value_str: String representation of a value. Can be a number ("12"),
+            tuple ("(1.0, 0.5, 0.2)"), list ("[1, 2, 3]"), quoted string
+            ('"Arial"'), boolean ("True"), or plain string ("Arial").
 
     Returns:
-        Parsed value
+        Parsed value with appropriate Python type. Numbers become int/float,
+        tuples/lists maintain structure, quoted strings are unquoted, and
+        unrecognized values are returned as strings.
+
+    Example:
+        ```python
+        safe_parse_value("12")          # -> 12
+        safe_parse_value("12.5")        # -> 12.5
+        safe_parse_value("(1,0,0)")     # -> (1, 0, 0)
+        safe_parse_value('"Arial"')     # -> "Arial"
+        safe_parse_value("True")        # -> True
+        safe_parse_value("plain_text")  # -> "plain_text"
+        ```
+
+    Note:
+        This function deliberately avoids eval() for security reasons and uses
+        ast.literal_eval() for safe parsing of Python literals.
     """
     # Strip quotes first if it's a quoted string
     value_str = value_str.strip()

natural_pdf/tables/__init__.py

@@ -1,5 +1,5 @@
 # new file
 # Re-export for convenient import
 from .result import TableResult
- 
-__all__ = ["TableResult"] 
+
+__all__ = ["TableResult"]

natural_pdf/tables/result.py

@@ -1,8 +1,9 @@
 """Sequence wrapper for table data with convenient DataFrame helpers."""
+
 from __future__ import annotations
 
 from collections.abc import Sequence
-from typing import Any, List, Iterator, Optional, Union
+from typing import Any, Iterator, List, Optional, Union
 
 
 class TableResult(Sequence):
@@ -12,9 +13,7 @@ class TableResult(Sequence):
     list of cell values) but offers an easy hand-off to *pandas*.
     """
 
-    _IMMUTABLE_MESSAGE = (
-        "TableResult is read-only; convert to list(result) if you need to mutate"
-    )
+    _IMMUTABLE_MESSAGE = "TableResult is read-only; convert to list(result) if you need to mutate"
 
     def __init__(self, rows: Optional[List[List[Any]]] = None) -> None:
         # Normalise to list of list so that Sequence operations work as expected
@@ -81,7 +80,9 @@ class TableResult(Sequence):
 
         df = pd.DataFrame(body, columns=hdr)
         if index_col is not None and not df.empty:
-            df.set_index(df.columns[index_col] if isinstance(index_col, int) else index_col, inplace=True)
+            df.set_index(
+                df.columns[index_col] if isinstance(index_col, int) else index_col, inplace=True
+            )
 
         if kwargs:
             df = pd.DataFrame(df, **kwargs)
@@ -98,4 +99,4 @@ class TableResult(Sequence):
     # Nice repr in notebooks
     def __repr__(self) -> str:  # noqa: D401 (simple)
         preview = "…" if len(self._rows) > 5 else ""
-        return f"TableResult(rows={len(self._rows)}{preview})" 
+        return f"TableResult(rows={len(self._rows)}{preview})"

natural_pdf/utils/bidi_mirror.py

@@ -6,6 +6,7 @@ replaces each bracket/parenthesis character with its Unicode-defined pair.
 For everyday PDFs the six ASCII pairs are enough, but the mapping can be
 extended easily from Unicode's BidiBrackets.txt.
 """
+
 from typing import Dict
 
 # Minimal mapping – ( ) [ ] { }
@@ -33,4 +34,4 @@ def mirror_brackets(text: str) -> str:  # pragma: no cover
     append = out_chars.append
     for ch in text:
         append(_ASCII_MIRROR.get(ord(ch), ch))
-    return "".join(out_chars) 
+    return "".join(out_chars)

natural_pdf/utils/reading_order.py

@@ -178,8 +178,9 @@ def _complex_reading_order(elements: List[Dict[str, Any]]) -> List[Dict[str, Any
     Returns:
         List of elements in reading order
     """
-    # TODO: Implement complex layout analysis
-    # For now, fall back to column-aware reading order
+    # TODO: Implement complex layout analysis for sophisticated document structures
+    # Would include: multi-column detection, figure/caption relationships, sidebars
+    # For now, fall back to column-aware reading order which handles most cases
     return _column_reading_order(elements)
 
 

natural_pdf/utils/visualization.py

@@ -237,11 +237,11 @@ def merge_images_with_legend(
 def render_plain_page(page, resolution):
     """
     Render a page to PIL Image using the specified resolution.
-    
+
     Args:
         page: Page object to render
         resolution: DPI resolution for rendering
-        
+
     Returns:
         PIL Image of the rendered page
     """
@@ -252,7 +252,7 @@ def render_plain_page(page, resolution):
     # Convert resolution (DPI) to scale factor for pypdfium2
     # PDF standard is 72 DPI, so scale = resolution / 72
     scale_factor = resolution / 72.0
-    
+
     bitmap = pdf_page.render(
         scale=scale_factor,
     )

natural_pdf/widgets/viewer.py

@@ -1,7 +1,6 @@
 # natural_pdf/widgets/viewer.py
 
 import logging
-import os
 
 from natural_pdf.utils.visualization import render_plain_page
 

{natural_pdf-0.1.33.dist-info → natural_pdf-0.1.35.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: natural-pdf
-Version: 0.1.33
+Version: 0.1.35
 Summary: A more intuitive interface for working with PDFs
 Author-email: Jonathan Soma <jonathan.soma@gmail.com>
 License-Expression: MIT

natural_pdf-0.1.35.dist-info/RECORD

@@ -0,0 +1,121 @@
+natural_pdf/__init__.py,sha256=qDFJNF8sbEDO-2WSFAxoWEM8updOUP6dB-ckya0kxfs,3275
+natural_pdf/cli.py,sha256=SkPwhhMM-GhLsj3O1n1Agxz4KOxcZ08sj8hVQSFJB5c,4064
+natural_pdf/analyzers/__init__.py,sha256=IPu_PMKFviDeEIeiC8_2KdeqH7z8OQ6q2v980hkByFY,672
+natural_pdf/analyzers/guides.py,sha256=5Lqc51trtqmLvjxLjDS__mgeyviRrjV-CIIT69RmEt4,92327
+natural_pdf/analyzers/shape_detection_mixin.py,sha256=Ef1o73QYVXQ2QcQMM_W9XRwY6vaIQHgxzD7etJ6LbiM,62820
+natural_pdf/analyzers/text_options.py,sha256=qEkDaYWla0rIM_gszEOsu52q7C_dAfV81P2HLJZM2sw,3333
+natural_pdf/analyzers/text_structure.py,sha256=VfKTsTFrK877sC0grsis9jK3rrgp0Mbp13VWEbukTcs,28437
+natural_pdf/analyzers/utils.py,sha256=PYbzJzSAHZ7JsMes84WIrSbA0zkjJGs0CLvIeINsf_k,2100
+natural_pdf/analyzers/layout/__init__.py,sha256=oq1uJ5UkGGMbBKGirV1aRKK3hxAUyjTLywYkPCQH1f0,33
+natural_pdf/analyzers/layout/base.py,sha256=F5xPOJcI65N4nxwm0szvhtbDD6lVMqWDut8PSkTCobU,8349
+natural_pdf/analyzers/layout/docling.py,sha256=4BJYyNVR6VegZGxyisvNIBBRvVk6YKPyDVs7ZdVfzEU,12676
+natural_pdf/analyzers/layout/gemini.py,sha256=ldECVCQ5HNQA3Omjg2NOsTrJXslyYb0vErDncmLIiuE,10510
+natural_pdf/analyzers/layout/layout_analyzer.py,sha256=Ff7OfjMyGIWkfPKl6dHgkFyb-iru83_VDk0gmvyHbbg,15549
+natural_pdf/analyzers/layout/layout_manager.py,sha256=ivMfTPjS14dyISu2o2Q6K48jkftvAOD04aCOtInkZGo,10267
+natural_pdf/analyzers/layout/layout_options.py,sha256=2JENtBMHhP3hP0zpFI5-UP3-t1y49E7oLZnjd9d1eB0,7704
+natural_pdf/analyzers/layout/paddle.py,sha256=44GG1sbaYTgvmtnrckNaCbDaNyw_D7FLLiSvzKP2cbk,23048
+natural_pdf/analyzers/layout/pdfplumber_table_finder.py,sha256=Tk0Q7wv7nGYPo69lh6RoezjdepTnMl90SaNIrP29Pwc,5902
+natural_pdf/analyzers/layout/surya.py,sha256=ugRXPIHiLoh65lfbbiXO317TbgdtQ-5kVN1nonEf4ws,9778
+natural_pdf/analyzers/layout/table_structure_utils.py,sha256=_sugFWvVpRK3EimOCrikTDAalGnSaWqiqFbtJw8t-lg,2770
+natural_pdf/analyzers/layout/tatr.py,sha256=cVr0ZyhY2mNLAKZ4DGMm-b7XNJpILKh8x8ZpyDeUhLk,15032
+natural_pdf/analyzers/layout/yolo.py,sha256=2Iz2-WsMy--ftkZQ8j5PGqp_1fTD7Mskl2kNnMUuwCU,8286
+natural_pdf/classification/manager.py,sha256=wyENltPSeWpJNjqzU91-ydJTnACZ_LC1q-ox_tRhMIM,22172
+natural_pdf/classification/mixin.py,sha256=CXygXXhe_qx1563SmIjiu4uSnZkxCkuRR4fGvLokS2w,9416
+natural_pdf/classification/results.py,sha256=5ha77CxK0GYwkBMJbvUBZkBjsL5GpOveIZDK9nO4j8I,3239
+natural_pdf/collections/mixins.py,sha256=sj76Cn6EdBtb5f-bdAV-1qpdixX8tI4BzPccPiYLI1w,5117
+natural_pdf/collections/pdf_collection.py,sha256=sDVEbFMNME_2OaHIsCoR_W7V1cAATNw4ZRqKWa6nbqA,30131
+natural_pdf/core/__init__.py,sha256=QC8H4M3KbXwMFiQORZ0pdPlzx1Ix6oKKQSS7Ib2KEaA,38
+natural_pdf/core/element_manager.py,sha256=DRZvntd99wjXy6KeDjCq5uRhjMftZop9QklOZqlUH8M,55349
+natural_pdf/core/highlighting_service.py,sha256=2tBrrEq6d6hz5f6Yf7z5TysJdlTyuHTURBnQxokJnDM,40645
+natural_pdf/core/page.py,sha256=Jw5SDshnHesqoC4yhtKEokeV08wMHuWZyWs5kDMOAjo,133204
+natural_pdf/core/pdf.py,sha256=9t8Ks-AZp3yjH_lRkFZAyIkjUQoCTRbmXK7vSi1e4UE,92415
+natural_pdf/describe/__init__.py,sha256=kIV7ORmWWB1SAur7nK2aAwR-wHqSedhKfUsaUl4hG0A,586
+natural_pdf/describe/base.py,sha256=CLhZXYQO6SOPUVWLt6VwZ7MK48t_6wgPMyFMLtTCKRc,18166
+natural_pdf/describe/elements.py,sha256=JicXC9SJmmasqxalpCXA47-kVwv-6JnR3Xiu778aNHM,12634
+natural_pdf/describe/mixin.py,sha256=rkX14aGrSz7Jvxx8Rbxv3eSfbO-_29DipwpstrV2pDQ,3109
+natural_pdf/describe/summary.py,sha256=cfT4ZQkeatCDAOwWPwhtEVXisNgk6E57fAXAnoRysSU,7645
+natural_pdf/elements/__init__.py,sha256=S8XeiNWJ1WcgnyYKdYV1yxQlAxCCO3FfITT8MQwNbyk,41
+natural_pdf/elements/base.py,sha256=-ZAcc8lb2aSWTKcprwKTvnR6hsDGDm7T8a1Y9V38E_A,52042
+natural_pdf/elements/collections.py,sha256=7i279l8kpgzRyvjRr13n1BeqbC5ufwYx7lu_WmfXWTE,131199
+natural_pdf/elements/image.py,sha256=zu-P2Y8fRoEXf6IeZU0EYRWsgZ6I_a5vy1FA3VXTGkQ,1424
+natural_pdf/elements/line.py,sha256=mHSeV-ZABY-Cc_K_NpFL53OGtTWlexYDlMvZc8_Vrx8,3845
+natural_pdf/elements/rect.py,sha256=QuQg0Qo7XYQKBac-3Ss0n0ELV6icdPcrygWM2VWzeX8,3325
+natural_pdf/elements/region.py,sha256=EqwtZJ2qgMyykuLVv2zO51oKJoSU4Hl7UA_mqTqRzmQ,143419
+natural_pdf/elements/text.py,sha256=409RqADe0FYG_i99n6Dy0hl_fWTtBHRCzCq7BP0eAL8,18854
+natural_pdf/export/mixin.py,sha256=L1q3MIEFWuvie4j4_EmW7GT3NerbZ1as0XMUoqTS7gM,5083
+natural_pdf/exporters/__init__.py,sha256=QffoARekR6WzXEd05oxOytly4qPdBizuIF-SUkeFpig,643
+natural_pdf/exporters/base.py,sha256=XhR1xlkHOh7suOuX7mWbsj1h2o1pZNet-OAS5YCJyeI,2115
+natural_pdf/exporters/hocr.py,sha256=wksvJvWLSxuAfhYzg_0T2_W8eqDoMgAVC-gwZ9FoO_k,19969
+natural_pdf/exporters/hocr_font.py,sha256=1wsGOMj6zoaRN2rxCwrv4MMLGawpNz984WgXpmWekgw,4574
+natural_pdf/exporters/original_pdf.py,sha256=KYW0f9_zdouZq_ZwNGvYnu6WHqv7JWrrEAdPCVmhRV4,6782
+natural_pdf/exporters/paddleocr.py,sha256=RBP03GCk0mLeC7tWtuti8AIUHlpOrtvbWkE2n7Ja7k8,19484
+natural_pdf/exporters/searchable_pdf.py,sha256=G2Tc4tpDXSYIufXJlkA8ppW_3DuzHAaweYKae33pI_c,16290
+natural_pdf/exporters/data/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+natural_pdf/exporters/data/pdf.ttf,sha256=x4RUIJJaI9iO2DCmOVe4r4Wmao2vjZ_JDoQ2c7LvGlk,572
+natural_pdf/exporters/data/sRGB.icc,sha256=KpLUuuRQt22LCqQhk9-XTXX2Jzjs6_dPAcXnWxKpV5Y,6922
+natural_pdf/extraction/manager.py,sha256=sASPJZ5cWFsl8A4PyTjg2yqkyC00tRl6glfoFA6HcsM,4979
+natural_pdf/extraction/mixin.py,sha256=z0HNRs4x4RoioNjzg3slDeqoHbiPug0HB37bUHehqMY,25066
+natural_pdf/extraction/result.py,sha256=PDaCCN2LQBbHsZy0_lrQ0ROeMsnmH1WRoXWOjk9M2o4,1825
+natural_pdf/flows/__init__.py,sha256=cUN4A8hTDLZSRr4PO2W_lR4z6hWpbNG8Seox-IIcrLU,277
+natural_pdf/flows/collections.py,sha256=iF8SsfKKb-YVIGi3m-yMRnfKgo_0n_EGhojnYK24h-Q,28493
+natural_pdf/flows/element.py,sha256=mKzk3B7A7sWNvu4CDvAjLr3_ZFLt--ktrSNoLfLpFxU,23940
+natural_pdf/flows/flow.py,sha256=ukkUqXsZmEw-QJEiVqEBLC8ktfBG2Bw56_RR1OEsd24,12802
+natural_pdf/flows/region.py,sha256=nB634NCuC2BzBHuXAn8Ynf5lwZnR5mWb3RD36iEaPYY,27659
+natural_pdf/ocr/__init__.py,sha256=VY8hhvDPf7Gh2lB-d2QRmghLLyTy6ydxlgo1cS4dOSk,2482
+natural_pdf/ocr/engine.py,sha256=SwNlWydtHbrIghV5JD_j5B4-rnjCMYIWUIEARag-zHw,11839
+natural_pdf/ocr/engine_doctr.py,sha256=ptKrupMWoulZb-R93zr9btoe94JPWU7vlJuN7OBJEIM,17740
+natural_pdf/ocr/engine_easyocr.py,sha256=bWz6kHUgAJfe3rqdnZBAF-IPvw3B35DlvX5KDdFUtzo,9888
+natural_pdf/ocr/engine_paddle.py,sha256=OmZlXVh2SSgNePqb6sMo2Mg5boX7REA4MUY25O7hKgU,16144
+natural_pdf/ocr/engine_surya.py,sha256=lOvSbZk53VKFVxRmqcQzM_0dHVdwTkRGiDZ9AWCgL1Q,5951
+natural_pdf/ocr/ocr_factory.py,sha256=Ix-p1SrV6dchq6YcbbCTf2BPBHSGwu9KBnwnZ_ohOuw,5282
+natural_pdf/ocr/ocr_manager.py,sha256=U8EVzNgeRQxxAbMpCEZhkF7nr_R8Fcvtp28oeV_D-Ms,16229
+natural_pdf/ocr/ocr_options.py,sha256=_BgLjIih6mY3k-AgkdXu9UDD8bykmQX2fpf37tAOhYQ,5146
+natural_pdf/ocr/utils.py,sha256=OxuHwDbHWj6setvnC0QYwMHrAjxGkhmLzWHpMqqGupA,4397
+natural_pdf/qa/__init__.py,sha256=2u2KJcA71g1I0HnLD-j6yvDw1moAjo9kkLhhfoYRURM,166
+natural_pdf/qa/document_qa.py,sha256=EduwpmUs8Oz35GrCfLw3312F_ngxIpWZLM8KNvasdrM,19887
+natural_pdf/qa/qa_result.py,sha256=8_jL5MJAHR4LcjGVe5lVsFizxWieF6VI86DWaqetYxs,2167
+natural_pdf/search/__init__.py,sha256=araouqM-l_m0VlluKf6i9BybAsHnfCuh39M0-xEI3jA,4273
+natural_pdf/search/lancedb_search_service.py,sha256=dfz5IiMIcAc3KFzkBDF6Ab_JDLpLHqW6DO1JDkPPu1k,14458
+natural_pdf/search/numpy_search_service.py,sha256=GwPwnX_wxBPFHe-bKS5upMRZLHj8PjLQ2d84lZygzHg,10331
+natural_pdf/search/search_options.py,sha256=sq_e8_jSROicD94b_xtDtLnjEr_Zsy4icjzPkK0a8QA,3566
+natural_pdf/search/search_service_protocol.py,sha256=Dl-Q-CrutkhZwI69scbW9EWPeYM63qxB60_EA7YqIYo,6699
+natural_pdf/search/searchable_mixin.py,sha256=hqQ_AuID5eTGRCtKYdFLZ1zF35y73uk3x1M1VW9Il8U,23514
+natural_pdf/selectors/__init__.py,sha256=oZGeqSv53EqmIZOhcnawuaGGlRg1h79vArXuZCWKm4A,123
+natural_pdf/selectors/parser.py,sha256=Flxjo_ZODBLQM8DQlQGqZTTQDyea3zUTzO9L2dtVabM,36402
+natural_pdf/tables/__init__.py,sha256=sCvCGbGsL6BiqlNxAYfVv003bIDLI11FmjHhaWfcU6w,104
+natural_pdf/tables/result.py,sha256=hrGIWDkImpdxsGzugcQKU-qrTgHwwfOigJDFdYl8aUc,3994
+natural_pdf/templates/__init__.py,sha256=jYBxzfi73vew0f6yhIh1MlRxw4F_TVN2hKQR0YXOFe0,20
+natural_pdf/utils/__init__.py,sha256=s3M8FggaK1P3EBYn6R_-HgSDjNc9C73gyKe1hihtNWg,43
+natural_pdf/utils/bidi_mirror.py,sha256=jJEES0xDrMfo5Me8kHMxHv4COS51PitnYi2EvKv3HCE,1151
+natural_pdf/utils/debug.py,sha256=RN7H3E6ph-GtxubCW6psW7TO8o2BxcNLiEzByTVR9fk,995
+natural_pdf/utils/highlighting.py,sha256=EIY6ihVGtUTS_DjWyxpnr_UXpcR4btC1KhSGQ9VUfKg,698
+natural_pdf/utils/identifiers.py,sha256=P7n6owcubnF8oAMa_UfYtENmIaJQdH_AMC9Jbs2bWXo,1117
+natural_pdf/utils/locks.py,sha256=7HJqV0VsNcOfISnbw8goCKWP5ck11uSJo6T_x9XIPKI,215
+natural_pdf/utils/packaging.py,sha256=e7U2wWvpunlAWpPFexNkD_c4dYbPp5LcKo7og4bNGvk,22411
+natural_pdf/utils/reading_order.py,sha256=u7XyVZdKMPMK0CL1C7xFogKnZ92b0JKT068KFjQWe18,7437
+natural_pdf/utils/text_extraction.py,sha256=mDeN1_VevNi3RwvFe48PM5vBh-A5WeBlYgP6lSjBaOk,10854
+natural_pdf/utils/visualization.py,sha256=olDkWtuVzP0NxRg0CP0DL-eXNCY7Bs-SH-2Xn-cjbo0,9370
+natural_pdf/widgets/__init__.py,sha256=QTVaUmsw__FCweFYZebwPssQxxUFUMd0wpm_cUbGZJY,181
+natural_pdf/widgets/viewer.py,sha256=KW3JogdR2TMg2ECUMYp8hwd060hfg8EsYBWxb5IEzBY,24942
+natural_pdf-0.1.35.dist-info/licenses/LICENSE,sha256=9zfwINwJlarbDmdh6iJV4QUG54QSJlSAUcnC1YiC_Ns,1074
+optimization/memory_comparison.py,sha256=F90D_5WhliSGAct_lyx93xd4q4F-jeo8QpGyDr8tmNw,6543
+optimization/pdf_analyzer.py,sha256=xf6h-FNlqCpsm8NriXcs_bQZOB8eQkxgGGKVRL_jgCM,19347
+optimization/performance_analysis.py,sha256=RjAqeE3YS1r_7qTWkY6Ng5YMbb6MXJXfXX6LoVjg_xQ,13035
+optimization/test_cleanup_methods.py,sha256=PmLOL4MRgvV0j_DW9W1TS8MsGGgu57QCuq6_5y7zK3s,6209
+optimization/test_memory_fix.py,sha256=A3knK74fNhvHknDbLhbTmA276x1ifl-3ivJ_7BhVSTI,6170
+tools/bad_pdf_eval/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+tools/bad_pdf_eval/analyser.py,sha256=bKUT3muP3ESE5i1D8sGyAS5tMzFMcq-i-xD_ZeUxYhY,13692
+tools/bad_pdf_eval/collate_summaries.py,sha256=L_YsdiqmwGIHYWTVJqo6gyazyn3GIQgpfGGKk8uwckk,5159
+tools/bad_pdf_eval/compile_attempts_markdown.py,sha256=ArFDZaSa9dz0ez0lsNlbUSK4hbvB3___DlfwqPEAZpY,4359
+tools/bad_pdf_eval/eval_suite.py,sha256=zcapsGwO-VJ2OupJnPYKbrkzvzdGdoh2DZPK19bfkQg,4450
+tools/bad_pdf_eval/evaluate_quality.py,sha256=-LR_shgxPVbaEZyWSVYKXTp2LNNVSdIwrlN5rllqntg,7149
+tools/bad_pdf_eval/export_enrichment_csv.py,sha256=1hd1iaTinmT8K1rlaHFV_ZvvbyuLEAnIbmKZUtRWv8o,1958
+tools/bad_pdf_eval/llm_enrich.py,sha256=mCh4KGi1HmIkzGjj5rrHz1Osd7sEX1IZ_FW08H1tP5Q,13313
+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-0.1.35.dist-info/METADATA,sha256=SVdCwYrjweXrrmU8m2korCIMJENbN9zDasRCi2pkb8E,6711
+natural_pdf-0.1.35.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+natural_pdf-0.1.35.dist-info/entry_points.txt,sha256=1R_KMv7g60UBBpRqGfw7bppsMNGdayR-iJlb9ohEk_8,81
+natural_pdf-0.1.35.dist-info/top_level.txt,sha256=oZlRzSc3nZ9sV3L6kD_Di734Pp62ANrm46imFVa51qQ,58
+natural_pdf-0.1.35.dist-info/RECORD,,