natural-pdf 0.2.10__py3-none-any.whl → 0.2.12__py3-none-any.whl

{natural_pdf-0.2.10.dist-info → natural_pdf-0.2.12.dist-info}/RECORD

@@ -2,7 +2,7 @@ natural_pdf/__init__.py,sha256=N4pR0LbuPEnUYFZqbdVqc_FGKldgwPQc1wjJhYKTBBM,3417
 natural_pdf/cli.py,sha256=0zO9ZoRiP8JmyGBaVavrMATnvbARWTl7WD2PEefu9BM,4061
 natural_pdf/text_mixin.py,sha256=eFCiHj6Okcw3aum4955BepcI2NPRalkf9UFFVTc_H30,4012
 natural_pdf/analyzers/__init__.py,sha256=3XGoNq3OgiVkZP7tOdeP5XVUl7fDgyztdA8DlOcMLXg,1138
-natural_pdf/analyzers/guides.py,sha256=O3MaeVLgH5l1qbj2fpJCUsDKF44350CiUb_T1J6HTiQ,178025
+natural_pdf/analyzers/guides.py,sha256=B2_Etb0o-lOku-FQw-T1Fo1qxbcAXT4FB0hdp-5kXRs,188171
 natural_pdf/analyzers/shape_detection_mixin.py,sha256=mgpyJ4jIulz9l9HCqThabJIsLSrXh9BB2AmLxUoHmw0,62584
 natural_pdf/analyzers/text_options.py,sha256=qEkDaYWla0rIM_gszEOsu52q7C_dAfV81P2HLJZM2sw,3333
 natural_pdf/analyzers/text_structure.py,sha256=3WWusi-BI0krUnJxB05DD6XmKj5qRNvQBqH7zOQGm1M,28451
@@ -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=KPuKM7SstfErTkRnGq4vrgE0Tv8iazN13Jp7yAXGKso,55575
 natural_pdf/core/highlighting_service.py,sha256=7on8nErhi50CEH2L4XzGIZ6tIqZtMzmmFlp-2lmwnYE,68856
-natural_pdf/core/page.py,sha256=XmXii652iM-JVKgzpbKQ8f59U0TvDLD5iAfdtx92gis,152675
+natural_pdf/core/page.py,sha256=Pid5hqVjcyX-gcCzxCJ62k6AQhNbUMNM_5QmEcylIjM,155264
 natural_pdf/core/page_collection.py,sha256=IjdFq9q0D0P6ZKWInf0H25rLzxfMb7RsUXucogkhNkU,63169
 natural_pdf/core/page_groupby.py,sha256=V2e_RNlHaasUzYm2h2vNJI7_aV_fl3_pg7kU3F2j0z8,8218
 natural_pdf/core/pdf.py,sha256=ovdeu9TRPnVYyMltD7QpcdcFYBLZFXh3LlfC5ifj6RY,104227
@@ -44,7 +44,7 @@ natural_pdf/elements/element_collection.py,sha256=idM_BUWEfbCJ5Sq0Ae_KfbVHy8TdkN
 natural_pdf/elements/image.py,sha256=zu-P2Y8fRoEXf6IeZU0EYRWsgZ6I_a5vy1FA3VXTGkQ,1424
 natural_pdf/elements/line.py,sha256=TFn7KXjPT_jUQyQyabU0F7XYU4dC-qadwodJMZF4DCU,3844
 natural_pdf/elements/rect.py,sha256=0lNkVkPkvbRbrFED856RXoUcTcDkeeOIs5xldKGAQT8,3324
-natural_pdf/elements/region.py,sha256=hCpbKg0R5TGfWEskZ6P-o_ZXPKhU4keaYjWIVX0Y7F4,165244
+natural_pdf/elements/region.py,sha256=HF6KzeuudO9upVLIrPsp3omcziLcILE3nnzl1a-LvK0,165400
 natural_pdf/elements/text.py,sha256=829uSJv9E-8cC6T6iR_Va7Xtv54pJoyRN78fq4NN1d4,20687
 natural_pdf/export/mixin.py,sha256=L1q3MIEFWuvie4j4_EmW7GT3NerbZ1as0XMUoqTS7gM,5083
 natural_pdf/exporters/__init__.py,sha256=QffoARekR6WzXEd05oxOytly4qPdBizuIF-SUkeFpig,643
@@ -107,12 +107,32 @@ natural_pdf/vision/results.py,sha256=F2zXG3MVZIpOUvPkJHotOq6-9rFz68BaO_8pnSndlOs
 natural_pdf/vision/similarity.py,sha256=YH8legN-t9uf1b_XULi4JLNDaRfPNKQwU1FZ4Qu08jY,11740
 natural_pdf/widgets/__init__.py,sha256=QTVaUmsw__FCweFYZebwPssQxxUFUMd0wpm_cUbGZJY,181
 natural_pdf/widgets/viewer.py,sha256=KW3JogdR2TMg2ECUMYp8hwd060hfg8EsYBWxb5IEzBY,24942
-natural_pdf-0.2.10.dist-info/licenses/LICENSE,sha256=9zfwINwJlarbDmdh6iJV4QUG54QSJlSAUcnC1YiC_Ns,1074
+natural_pdf-0.2.12.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
 optimization/test_cleanup_methods.py,sha256=PmLOL4MRgvV0j_DW9W1TS8MsGGgu57QCuq6_5y7zK3s,6209
 optimization/test_memory_fix.py,sha256=A3knK74fNhvHknDbLhbTmA276x1ifl-3ivJ_7BhVSTI,6170
+temp/debug_cell_extraction.py,sha256=nE0Z470P40v8xZfWO1V3qgNaejs_pernEQaUOFeOJ1U,1527
+temp/debug_exclusion_overlap.py,sha256=RptJXwqBXy5gsvMF037KEx1o2QgjwEDkMB6TD5aJdqA,1644
+temp/debug_exclusions_guides.py,sha256=s8siep9te1KRJ2j0vH1tvDQnBlz7PKbHeCiYMrZL8jE,2096
+temp/debug_extra_guide.py,sha256=95Tim-YnmAR4kICw2XDKVDvlW5WsjK_51cv5-EV11rc,1236
+temp/debug_outer_boundaries.py,sha256=uJUJwojTxOU4VtbGUouuhV65IYzS6NDIVKxnS7o64nU,1456
+temp/debug_st_search.py,sha256=F4c_mUVi_d5AKaKIpQ0AnW1amDqAwALoQQj7wZj--J0,1021
+temp/fix_page_exclusions.py,sha256=YIj62zF38TdoBARAuSIvEbetl_JfXG-mp4v9p355qmo,1358
+temp/test_exclusion_with_debug.py,sha256=CScxHvb43KrB5dzXuTOhuzjcBXZBdfYB5ygiKkEW26g,1393
+temp/test_find_exclusions_fix.py,sha256=1l5aEqnElcl3kiykdtmJFlVxQ1xMKGm1UckGYEQg--c,2103
+temp/test_find_exclusions_fix_no_recursion.py,sha256=qZspTBwxunRM93N_-fZ2fR5Lodj0ArQX3h10HlTXhfc,3592
+temp/test_fix_real_pdf.py,sha256=uuylxmpeAEbIix9wjl0Gri1sZlN61dBWTq6ZCyfvzF8,1454
+temp/test_fix_working.py,sha256=-Ryre1rXYA2EG_lmPZGYEGi8yz0slhHEXPJMYexZW84,1750
+temp/test_fixed_pdf_exclusions.py,sha256=Q5zxooKDvtTXo-dDsx3nsQw1ZVHX3TW47iZ_dXpFdrY,2168
+temp/test_horizontal_top_bottom.py,sha256=Mb3tjt9Z3wOTpzFOgK7i0K-j-_ynNh4vDu2x1L3nu-s,2163
+temp/test_marker_order.py,sha256=TFZkMxRiNoZGVcdDivYnkIDNvwHaiyKUdYoy2rTTIiI,1417
+temp/test_original_exclusions_now_work.py,sha256=G6LmaF-P9Qhj0j4lT_4ncfCddllfP6L8F_x2prUBr9w,1904
+temp/test_pdf_exclusions_with_guides.py,sha256=QaMl0frgKC8kCPQ2BUI8kqyvqsIjQPXKV_St1rK3zxg,2754
+temp/test_region_exclusions_detailed.py,sha256=EftdW3JY3JH_LX5QlWKt-4drM-joPggK2fKUZRXVTMA,814
+temp/test_stripes_real_pdf.py,sha256=FIvDoJrnuioOMw1A0aTCCfZLeg99lusfe0Fb0MiqnhQ,2618
+temp/test_vertical_stripes.py,sha256=Yf3TJfb_faqAFzlgb7i5u6dDHjF4UMSHIGM99vangRk,1877
 tools/bad_pdf_eval/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tools/bad_pdf_eval/analyser.py,sha256=oqSTo3NLyignp_XdCO9_SRCUUXMU8lfgDavKYZYNxws,13690
 tools/bad_pdf_eval/collate_summaries.py,sha256=L_YsdiqmwGIHYWTVJqo6gyazyn3GIQgpfGGKk8uwckk,5159
@@ -124,8 +144,8 @@ 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-0.2.10.dist-info/METADATA,sha256=DnjO1O3cHjhT-sXzLjIHQ286xqBAzeSJdPR50yol4D4,6960
-natural_pdf-0.2.10.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-natural_pdf-0.2.10.dist-info/entry_points.txt,sha256=1R_KMv7g60UBBpRqGfw7bppsMNGdayR-iJlb9ohEk_8,81
-natural_pdf-0.2.10.dist-info/top_level.txt,sha256=80t0F2ZeX4vN4Ke5iTflcOk_PN_0USn33ha3X6X86Ik,36
-natural_pdf-0.2.10.dist-info/RECORD,,
+natural_pdf-0.2.12.dist-info/METADATA,sha256=jRNM0JxYvPDuqzD63earjbaUwQgXCjPYPLC5pLl49Uk,6960
+natural_pdf-0.2.12.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+natural_pdf-0.2.12.dist-info/entry_points.txt,sha256=1R_KMv7g60UBBpRqGfw7bppsMNGdayR-iJlb9ohEk_8,81
+natural_pdf-0.2.12.dist-info/top_level.txt,sha256=ZDKhxE_tg508o9BpagsjCGcI8GY4cF_8bg0e0IaLsPI,41
+natural_pdf-0.2.12.dist-info/RECORD,,

{natural_pdf-0.2.10.dist-info → natural_pdf-0.2.12.dist-info}/top_level.txt

@@ -1,4 +1,5 @@
 natural_pdf
 optimization
+temp
 todo
 tools

temp/debug_cell_extraction.py

@@ -0,0 +1,42 @@
+"""Debug cell text extraction with exclusions"""
+from natural_pdf import PDF
+from natural_pdf.analyzers.guides import Guides
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Add exclusions
+pdf.add_exclusion(lambda page: page.find(text="PREMISE").above(), label="header")
+
+# Check exclusions are registered
+print("Exclusions on page:")
+exclusions = page._get_exclusion_regions(debug=True)
+
+# Create guides and build grid
+headers = page.find(text="NUMBER").right(include_source=True).expand(top=3, bottom=3).find_all('text')
+guides = Guides(page)
+guides.vertical.from_content(headers, align='left')
+guides.horizontal.from_stripes()
+
+# Build grid and get cells
+grid_result = guides.build_grid(include_outer_boundaries=True)
+cells = grid_result["regions"]["cells"]
+
+print(f"\nTotal cells: {len(cells)}")
+
+# Check first row cells (these should be in excluded area)
+first_row_cells = [c for c in cells if c.bbox[1] < 90]  # y < 90
+print(f"\nFirst row cells: {len(first_row_cells)}")
+
+for i, cell in enumerate(first_row_cells[:3]):
+    print(f"\nCell {i}:")
+    print(f"  Bbox: {cell.bbox}")
+    print(f"  Raw text: {repr(cell.extract_text(apply_exclusions=False))}")
+    print(f"  With exclusions: {repr(cell.extract_text(apply_exclusions=True))}")
+
+# Now test the full table extraction
+print("\n\nFull table extraction:")
+result = guides.extract_table(include_outer_boundaries=True, apply_exclusions=True, header=False)
+df = result.to_df()
+print("\nFirst row of dataframe:")
+print(df.iloc[0].to_dict() if not df.empty else "Empty")

temp/debug_exclusion_overlap.py

@@ -0,0 +1,43 @@
+"""Debug how exclusions work with overlapping regions"""
+from natural_pdf import PDF
+from natural_pdf.analyzers.guides import Guides
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Add exclusion
+pdf.add_exclusion(lambda page: page.find(text="PREMISE").above(), label="header")
+
+# Get the exclusion region
+exclusions = page._get_exclusion_regions()
+excl_region = exclusions[0]
+print(f"Exclusion region: {excl_region.bbox}")
+print(f"Exclusion bottom: {excl_region.bbox[3]}")
+
+# Create a test cell that overlaps the exclusion
+# Cell 1 from before: (32.06, 0.5, 73.18288, 79.53999999999996)
+test_cell = page.region(32.06, 0.5, 73.18288, 79.53999999999996)
+
+print(f"\nTest cell: {test_cell.bbox}")
+print(f"Cell overlaps exclusion: top={test_cell.bbox[1]} < excl_bottom={excl_region.bbox[3]}")
+
+# Extract text from different y-ranges
+print("\nText in different parts of the cell:")
+
+# Part above exclusion line (should be empty)
+upper_part = page.region(32.06, 0.5, 73.18288, 59.12)
+print(f"Upper part (0.5 to 59.12): '{upper_part.extract_text(apply_exclusions=True)}'")
+
+# Part below exclusion line (should have text)  
+lower_part = page.region(32.06, 59.12, 73.18288, 79.54)
+print(f"Lower part (59.12 to 79.54): '{lower_part.extract_text()}'")
+
+# The whole cell
+print(f"Whole cell with exclusions: '{test_cell.extract_text(apply_exclusions=True)}'")
+print(f"Whole cell without exclusions: '{test_cell.extract_text(apply_exclusions=False)}'")
+
+# Check what text elements are in this region
+print("\nText elements in cell:")
+cell_texts = test_cell.find_all('text')
+for t in cell_texts[:5]:
+    print(f"  '{t.text}' at y={t.top:.2f}-{t.bottom:.2f}")

temp/debug_exclusions_guides.py

@@ -0,0 +1,67 @@
+"""Debug why exclusions aren't working with guides.extract_table()"""
+from natural_pdf import PDF
+from natural_pdf.analyzers.guides import Guides
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Check initial text
+print("Initial text:")
+print(page.extract_text()[:200])
+print()
+
+# Add exclusions
+pdf.add_exclusion(lambda page: page.find(text="PREMISE").above())
+pdf.add_exclusion(lambda page: page.find("text:regex(Page \d+ of)").expand())
+
+# Check text after exclusions
+print("Text after exclusions:")
+print(page.extract_text()[:100])
+print()
+
+# Debug exclusion regions
+print("Checking exclusion regions:")
+exclusions = page._get_exclusion_regions(debug=True)
+print(f"\nTotal exclusions: {len(exclusions)}")
+for i, exc in enumerate(exclusions):
+    print(f"  {i}: {exc.bbox}")
+print()
+
+# Create guides
+headers = (
+    page
+    .find(text="NUMBER")
+    .right(include_source=True)
+    .expand(top=3, bottom=3)
+    .find_all('text')
+)
+
+guides = Guides(page)
+guides.vertical.from_content(headers, align='left')
+guides.horizontal.from_stripes()
+
+# Build grid to see what regions are created
+print("\nBuilding grid...")
+grid_result = guides.build_grid(include_outer_boundaries=True)
+table_region = grid_result["regions"]["table"]
+print(f"Table region: {table_region}")
+print(f"Table bbox: {table_region.bbox if table_region else 'None'}")
+
+# Check if table region respects exclusions
+if table_region:
+    print("\nExtracting text from table region directly:")
+    table_text = table_region.extract_text()[:200]
+    print(f"Table text: {table_text}")
+    
+    # Now extract table
+    print("\nExtracting table with apply_exclusions=True:")
+    result = guides.extract_table(include_outer_boundaries=True, apply_exclusions=True, header=False)
+    df = result.to_df()
+    print(df.head())
+    
+    # Check if excluded content is in the table
+    table_str = df.to_string()
+    has_feb = "FEBRUARY 2014" in table_str or "FEBR" in table_str
+    has_alphabetic = "ALPHABETIC LISTING" in table_str
+    print(f"\nContains 'FEBRUARY': {has_feb}")
+    print(f"Contains 'ALPHABETIC LISTING': {has_alphabetic}")

temp/debug_extra_guide.py

@@ -0,0 +1,41 @@
+"""Debug the extra guide issue."""
+from natural_pdf import PDF
+from natural_pdf.analyzers.guides import Guides
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Get headers
+headers = (
+    page
+    .find("text:contains(NUMBER)")
+    .right(include_source=True)
+    .expand(top=3, bottom=3)
+    .find_all('text')
+)
+
+print("Headers 3-5:")
+for i, h in enumerate(headers[3:5]):
+    print(f"  {i}: '{h.text}' bbox={h.bbox}")
+
+# Create guides with just these two headers
+guides = Guides(page)
+guides.vertical.from_content(headers[3:5], align='left', outer=False)
+
+print(f"\nResulting guides: {guides.vertical}")
+print(f"Expected: [328.32012, 539.63316]")
+
+# Let's also check what happens with each header individually
+print("\nTesting each header individually:")
+for i, h in enumerate(headers[3:5]):
+    g = Guides(page)
+    g.vertical.from_content([h], align='left', outer=False)
+    print(f"  Header {i} guides: {g.vertical}")
+
+# Check if it's related to the ElementCollection
+print("\nTesting with manual list of text:")
+text_list = [h.text for h in headers[3:5]]
+print(f"Text list: {text_list}")
+guides2 = Guides(page)
+guides2.vertical.from_content(text_list, align='left', outer=False)
+print(f"Guides from text list: {guides2.vertical}")

temp/debug_outer_boundaries.py

@@ -0,0 +1,46 @@
+"""Debug outer boundaries issue with exclusions"""
+from natural_pdf import PDF
+from natural_pdf.analyzers.guides import Guides
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Add exclusions
+pdf.add_exclusion(lambda page: page.find(text="PREMISE").above())
+
+# Create guides
+headers = (
+    page
+    .find(text="NUMBER")
+    .right(include_source=True)
+    .expand(top=3, bottom=3)
+    .find_all('text')
+)
+
+guides = Guides(page)
+guides.vertical.from_content(headers, align='left')
+guides.horizontal.from_stripes()
+
+print("Horizontal guides (sorted):")
+for i, h in enumerate(sorted(guides.horizontal)):
+    print(f"  {i}: {h:.2f}")
+
+print(f"\nFirst content guide: {sorted(guides.horizontal)[0]:.2f}")
+print(f"Page height: {page.height}")
+
+# Test without outer boundaries
+print("\n\nWithout outer boundaries:")
+result1 = guides.extract_table(include_outer_boundaries=False, apply_exclusions=True, header=False)
+df1 = result1.to_df()
+print(f"Shape: {df1.shape}")
+print("First row, first column:", df1.iloc[0, 0] if not df1.empty else "Empty")
+
+# Test with outer boundaries
+print("\n\nWith outer boundaries:")
+result2 = guides.extract_table(include_outer_boundaries=True, apply_exclusions=True, header=False)
+df2 = result2.to_df()
+print(f"Shape: {df2.shape}")
+print("First row, first column:", df2.iloc[0, 0] if not df2.empty else "Empty")
+
+# The issue: include_outer_boundaries adds guides at 0 and 612,
+# which creates cells that span into the exclusion zone

temp/debug_st_search.py

@@ -0,0 +1,33 @@
+"""Debug searching for 'ST' text."""
+from natural_pdf import PDF
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Get the original ST element
+headers = (
+    page
+    .find("text:contains(NUMBER)")
+    .right(include_source=True)
+    .expand(top=3, bottom=3)
+    .find_all('text')
+)
+original_st = headers[4]
+print(f"Original 'ST' element: '{original_st.text}' at {original_st.bbox}")
+
+# Search for 'ST' using find
+found_st = page.find('text:contains("ST")')
+print(f"\nFound 'ST' using find: '{found_st.text}' at {found_st.bbox}")
+
+# Find all elements containing 'ST'
+all_st = page.find_all('text:contains("ST")')
+print(f"\nAll elements containing 'ST':")
+for i, elem in enumerate(all_st[:10]):  # First 10
+    print(f"  {i}: '{elem.text}' at x={elem.x0:.2f}, bbox={elem.bbox}")
+
+# Check what's at position 332.88
+print(f"\nLooking for element at x≈332.88:")
+all_text = page.find_all('text')
+for elem in all_text:
+    if 332 < elem.x0 < 334:
+        print(f"  Found: '{elem.text}' at x={elem.x0:.5f}, bbox={elem.bbox}")

temp/fix_page_exclusions.py

@@ -0,0 +1,42 @@
+"""Script to fix the remaining exclusion bugs in page.py"""
+
+import re
+
+# Read the file
+with open('/Users/soma/Development/natural-pdf/natural_pdf/core/page.py', 'r') as f:
+    content = f.read()
+
+# Fix 1: Line 1132 in find() method
+# Change: if apply_exclusions and self._exclusions and results_collection:
+# To: if apply_exclusions and results_collection:
+content = re.sub(
+    r'(\s+)if apply_exclusions and self\._exclusions and results_collection:',
+    r'\1if apply_exclusions and results_collection:',
+    content
+)
+
+# Fix 2: Line 1227 in find_all() method
+# Same change pattern
+content = re.sub(
+    r'(\s+)if apply_exclusions and self\._exclusions and results_collection:',
+    r'\1if apply_exclusions and results_collection:',
+    content
+)
+
+# Fix 3: Line 1599 in get_elements() method
+# Change: if apply_exclusions and self._exclusions:
+# To: if apply_exclusions:
+content = re.sub(
+    r'(\s+)if apply_exclusions and self\._exclusions:',
+    r'\1if apply_exclusions:',
+    content
+)
+
+# Write the fixed content back
+with open('/Users/soma/Development/natural-pdf/natural_pdf/core/page.py', 'w') as f:
+    f.write(content)
+
+print("Fixed exclusion checks in page.py")
+print("- find() method: removed self._exclusions check")
+print("- find_all() method: removed self._exclusions check")
+print("- get_elements() method: removed self._exclusions check")

temp/test_exclusion_with_debug.py

@@ -0,0 +1,30 @@
+"""Test exclusion with detailed debugging"""
+from natural_pdf import PDF
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Add exclusion
+pdf.add_exclusion(lambda page: page.find(text="PREMISE").above(), label="header")
+
+# First, verify the exclusion works on the page itself
+print("Page-level text extraction:")
+print("Without exclusions:", page.extract_text(apply_exclusions=False)[:100])
+print("With exclusions:", page.extract_text(apply_exclusions=True)[:100])
+
+# Now test on a specific region that should be excluded
+print("\n\nRegion in excluded area (0, 0, 200, 50):")
+excluded_region = page.region(0, 0, 200, 50)
+print("Without exclusions:", repr(excluded_region.extract_text(apply_exclusions=False)))
+print("With exclusions:", repr(excluded_region.extract_text(apply_exclusions=True)))
+
+# Test the actual problematic cell region
+print("\n\nProblematic cell region (32.06, 0.5, 73.18, 79.54):")
+cell_region = page.region(32.06, 0.5, 73.18288, 79.54)
+print("Without exclusions:", repr(cell_region.extract_text(apply_exclusions=False)))
+print("With exclusions:", repr(cell_region.extract_text(apply_exclusions=True)))
+
+# Check if the region inherits the page
+print(f"\nCell region's page: {cell_region.page}")
+print(f"Cell region's _page: {getattr(cell_region, '_page', 'Not found')}")
+print(f"Same as original page: {cell_region.page is page if hasattr(cell_region, 'page') else 'N/A'}")

temp/test_find_exclusions_fix.py

@@ -0,0 +1,53 @@
+"""Test that the find methods now work with PDF-level exclusions."""
+
+from natural_pdf import PDF
+
+# Load a test PDF
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Add PDF-level exclusion for the header
+pdf.add_exclusion(lambda page: page.find(text="PREMISE").above())
+
+# Test 1: find() should now exclude the header text
+print("Test 1: find() with PDF-level exclusions")
+result = page.find("text:contains(FEBRUARY)", apply_exclusions=True)
+if result is None:
+    print("✅ SUCCESS: find() correctly excluded header text")
+else:
+    print(f"❌ FAILED: find() returned {result.text}")
+
+# Test 2: find_all() should exclude header elements
+print("\nTest 2: find_all() with PDF-level exclusions")
+all_text = page.find_all("text", apply_exclusions=False)
+filtered_text = page.find_all("text", apply_exclusions=True)
+print(f"Without exclusions: {len(all_text)} elements")
+print(f"With exclusions: {len(filtered_text)} elements")
+if len(filtered_text) < len(all_text):
+    print("✅ SUCCESS: find_all() excluded some elements")
+else:
+    print("❌ FAILED: find_all() didn't exclude any elements")
+
+# Test 3: get_elements() should exclude header elements
+print("\nTest 3: get_elements() with PDF-level exclusions")
+all_elements = page.get_elements(apply_exclusions=False)
+filtered_elements = page.get_elements(apply_exclusions=True)
+print(f"Without exclusions: {len(all_elements)} elements")
+print(f"With exclusions: {len(filtered_elements)} elements")
+if len(filtered_elements) < len(all_elements):
+    print("✅ SUCCESS: get_elements() excluded some elements")
+else:
+    print("❌ FAILED: get_elements() didn't exclude any elements")
+
+# Test that excluded text is not in the filtered results
+print("\nChecking excluded text...")
+excluded_texts = ["FEBRUARY", "ALPHABETIC LISTING", "M27"]
+for text in excluded_texts:
+    found_in_filtered = any(
+        text in str(el.text) if hasattr(el, 'text') else False 
+        for el in filtered_text
+    )
+    if not found_in_filtered:
+        print(f"✅ '{text}' correctly excluded")
+    else:
+        print(f"❌ '{text}' still present in filtered results")

temp/test_find_exclusions_fix_no_recursion.py

@@ -0,0 +1,97 @@
+"""Test that the find methods now work with PDF-level exclusions (without recursion)."""
+
+from natural_pdf import PDF
+
+# Load a test PDF
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Add PDF-level exclusion using a Region directly to avoid recursion
+# First, get the header region without exclusions
+header_element = page.find(text="PREMISE", apply_exclusions=False)
+if header_element:
+    header_region = header_element.above()
+    pdf.add_exclusion(header_region)
+else:
+    print("WARNING: Could not find PREMISE text for exclusion")
+
+# Test 1: find() should now exclude the header text
+print("Test 1: find() with PDF-level exclusions")
+result = page.find("text:contains(FEBRUARY)", apply_exclusions=True)
+if result is None:
+    print("✅ SUCCESS: find() correctly excluded header text")
+else:
+    print(f"❌ FAILED: find() returned {result.text}")
+
+# Test 2: find_all() should exclude header elements
+print("\nTest 2: find_all() with PDF-level exclusions")
+all_text = page.find_all("text", apply_exclusions=False)
+filtered_text = page.find_all("text", apply_exclusions=True)
+print(f"Without exclusions: {len(all_text)} elements")
+print(f"With exclusions: {len(filtered_text)} elements")
+if len(filtered_text) < len(all_text):
+    print("✅ SUCCESS: find_all() excluded some elements")
+else:
+    print("❌ FAILED: find_all() didn't exclude any elements")
+
+# Test 3: get_elements() should exclude header elements
+print("\nTest 3: get_elements() with PDF-level exclusions")
+all_elements = page.get_elements(apply_exclusions=False)
+filtered_elements = page.get_elements(apply_exclusions=True)
+print(f"Without exclusions: {len(all_elements)} elements")
+print(f"With exclusions: {len(filtered_elements)} elements")
+if len(filtered_elements) < len(all_elements):
+    print("✅ SUCCESS: get_elements() excluded some elements")
+else:
+    print("❌ FAILED: get_elements() didn't exclude any elements")
+
+# Test that excluded text is not in the filtered results
+print("\nChecking excluded text...")
+excluded_texts = ["FEBRUARY", "ALPHABETIC LISTING", "M27"]
+for text in excluded_texts:
+    found_in_filtered = any(
+        text in str(el.text) if hasattr(el, 'text') else False 
+        for el in filtered_text
+    )
+    if not found_in_filtered:
+        print(f"✅ '{text}' correctly excluded")
+    else:
+        print(f"❌ '{text}' still present in filtered results")
+
+# Also test that the original table extraction issue is fixed
+print("\n\nTesting original table extraction issue...")
+from natural_pdf.analyzers.guides import Guides
+
+# Add exclusion for footer too
+footer_element = page.find("text:regex(Page \\d+ of)", apply_exclusions=False)
+if footer_element:
+    pdf.add_exclusion(footer_element.expand())
+
+headers = (
+    page
+    .find(text="NUMBER", apply_exclusions=False)
+    .right(include_source=True)
+    .expand(top=3, bottom=3)
+    .find_all('text')
+)
+
+guides = Guides(page)
+guides.vertical.from_content(headers, align='left')
+guides.horizontal.from_stripes()
+
+result = guides.extract_table(include_outer_boundaries=True, apply_exclusions=True, header=False)
+df = result.to_df()
+
+# Check if excluded content is in the table
+table_str = df.to_string()
+has_feb = "FEBRUARY" in table_str
+has_alphabetic = "ALPHABETIC LISTING" in table_str
+
+print(f"\nTable extraction with exclusions:")
+print(f"Contains 'FEBRUARY': {has_feb}")
+print(f"Contains 'ALPHABETIC LISTING': {has_alphabetic}")
+
+if not has_feb and not has_alphabetic:
+    print("✅ SUCCESS: Table extraction correctly excludes header/footer!")
+else:
+    print("❌ FAILED: Exclusions not working in table extraction")

temp/test_fix_real_pdf.py

@@ -0,0 +1,48 @@
+"""Test the fix with the actual PDF from the user's example."""
+from natural_pdf import PDF
+from natural_pdf.analyzers.guides import Guides
+
+pdf = PDF("pdfs/m27.pdf")
+page = pdf.pages[0]
+
+# Get headers using the user's exact code
+headers = (
+    page
+    .find("text:contains(NUMBER)")
+    .right(include_source=True)
+    .expand(top=3, bottom=3)
+    .find_all('text')
+)
+
+print("Headers found:")
+for i, h in enumerate(headers):
+    print(f"  {i}: '{h.text}' at x={h.x0:.2f}")
+
+# Create guides using ElementCollection
+guides = Guides(page)
+guides.vertical.from_content(headers, align='left', outer=False)
+
+print(f"\nResulting vertical guides: {sorted(guides.vertical)}")
+
+# Check specific headers that were problematic
+print("\nChecking headers 3-5:")
+for i, h in enumerate(headers[3:5]):
+    print(f"  Header {i+3}: '{h.text}' at x={h.x0:.5f}")
+
+# Test with just headers 3-5
+guides2 = Guides(page)
+guides2.vertical.from_content(headers[3:5], align='left', outer=False) 
+
+print(f"\nGuides from headers[3:5]: {guides2.vertical}")
+print(f"Expected: [328.32012, 539.63316]")
+
+# Verify the fix
+if 332.88095999999996 in guides2.vertical:
+    print("\n❌ FAILED: Extra guide at 332.88 is still present")
+else:
+    print("\n✅ SUCCESS: Extra guide at 332.88 is not present")
+
+# Test that outer guides work correctly too
+guides3 = Guides(page)
+guides3.vertical.from_content(headers[3:5], align='left', outer=True)
+print(f"\nWith outer=True: {guides3.vertical}")