4d6fd8a4ab
Add 7 adversarial PDF fixtures exercising Phase 1 error-recovery paths: - xref_30pct_bad_offsets.pdf: 100 objects, 30 bad xref offsets - missing_mediabox_all_pages.pdf: 10 pages, no /MediaBox at any level - missing_endobj.pdf: object 5 missing endobj marker - truncated_mid_stream.pdf: FlateDecode stream truncated mid-decompression - int_overflow_bbox.pdf: /BBox value 99999999999999999 (i32 overflow) - nested_failure.pdf: every page has at least one diagnostic - combined_failures.pdf: combines multiple failure modes (keystone INV-8 test) Each fixture has a sibling .expected_diagnostics.json file with threshold counts (>= not == per EC-07/EC-09 to tolerate drift). Integration test harness (error_recovery_integration.rs): - assert_diagnostic_count_at_least() helper for threshold checking - assert_no_panic() helper using std::panic::catch_unwind for INV-8 - Individual test functions for each fixture - Cumulative test_inv_8_no_panics_across_all_fixtures() All 8 tests pass. INV-8 verified: zero panics across all fixtures. Closes: pdftract-4w0v4
71 lines
2.5 KiB
Python
71 lines
2.5 KiB
Python
#!/usr/bin/env python3
|
|
"""Generate xref_30pct_bad_offsets.pdf - 100-object PDF where 30 xref entries point to wrong offsets."""
|
|
|
|
# Generate a PDF with 100 objects where 30% have bad xref offsets
|
|
objects = []
|
|
xref_entries = []
|
|
|
|
# Object 0 is always free
|
|
xref_entries.append("0000000000 65535 f")
|
|
|
|
# Generate 100 objects (1-100)
|
|
# First 70 are valid, last 30 have bad offsets
|
|
for i in range(1, 101):
|
|
if i <= 70:
|
|
# Valid objects
|
|
if i == 1:
|
|
objects.append(f"{i} 0 obj\n<< /Type /Catalog /Pages 2 0 R >>\nendobj\n")
|
|
elif i == 2:
|
|
objects.append(f"{i} 0 obj\n<< /Type /Pages /Kids [3 0 R] /Count 1 >>\nendobj\n")
|
|
elif i == 3:
|
|
objects.append(f"{i} 0 obj\n<< /Type /Page /Parent 2 0 R /MediaBox [0 0 612 792] /Contents 100 0 R /Resources << /Font << /F1 99 0 R >> >> >>\nendobj\n")
|
|
elif i == 99:
|
|
objects.append(f"{i} 0 obj\n<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica >>\nendobj\n")
|
|
elif i == 100:
|
|
objects.append(f"{i} 0 obj\n<< /Length 44 >>\nstream\nBT\n/F1 12 Tf\n100 700 Td\n(Test) Tj\nET\nendstream\nendobj\n")
|
|
else:
|
|
# Dummy objects
|
|
objects.append(f"{i} 0 obj\n<< /Type /Dummy /Data {i} >>\nendobj\n")
|
|
else:
|
|
# Objects with bad xref offsets - these will exist in the PDF but xref will point to wrong places
|
|
objects.append(f"{i} 0 obj\n<< /Type /Dummy /Data {i} >>\nendobj\n")
|
|
|
|
# Calculate the actual offsets for the valid xref entries
|
|
pdf_body = "%PDF-1.4\n"
|
|
offset = len(pdf_body)
|
|
|
|
obj_offsets = []
|
|
for obj in objects:
|
|
obj_offsets.append(offset)
|
|
pdf_body += obj
|
|
offset += len(obj)
|
|
|
|
# Build the xref table with 30% bad offsets
|
|
xref_table = "xref\n0 101\n"
|
|
for i in range(101):
|
|
if i == 0:
|
|
xref_table += "0000000000 65535 f\n"
|
|
elif i <= 70:
|
|
# Valid offset
|
|
xref_table += f"{obj_offsets[i-1]:010d} 00000 n\n"
|
|
else:
|
|
# Bad offset - point to somewhere in the middle of the PDF
|
|
bad_offset = 99999
|
|
xref_table += f"{bad_offset:010d} 00000 n\n"
|
|
|
|
# Add trailer and EOF
|
|
trailer = f"""trailer
|
|
<< /Size 101 /Root 1 0 R >>
|
|
startxref
|
|
{offset}
|
|
%%EOF
|
|
"""
|
|
|
|
PDF_CONTENT = pdf_body + xref_table + trailer
|
|
|
|
with open('xref_30pct_bad_offsets.pdf', 'wb') as f:
|
|
f.write(PDF_CONTENT.encode('latin-1'))
|
|
|
|
print("Generated xref_30pct_bad_offsets.pdf")
|
|
print("100 objects, 30 with bad xref offsets (objects 71-100)")
|
|
print("Expected: 70 objects extracted, 30+ STRUCT_INVALID_XREF_ENTRY diagnostics")
|