test(pdf): diagnose S01 metric mismatch without changing fidelity gates

2026-10-29 20:45:33 +01:00
parent 26d531cc90
commit 4024a89a95
8 changed files with 408 additions and 0 deletions
@@ -0,0 +1,75 @@
+"""Explain each unchanged S01 failure using actual DOM/PDF baseline and font-cell math.
+
+This report is evidence only. It never replaces the historical oracle or grants approval.
+"""
+import argparse
+import hashlib
+import importlib.util
+import json
+from pathlib import Path
+
+import pdfplumber
+from pypdf import PdfReader
+
+spec = importlib.util.spec_from_file_location('baseline_math', Path(__file__).with_name('baseline-math.py'))
+math = importlib.util.module_from_spec(spec); spec.loader.exec_module(math)
+
+
+def sha(file): return hashlib.sha256(file.read_bytes()).hexdigest()
+
+
+def main():
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument('corpus', type=Path)
+    parser.add_argument('dom', type=Path)
+    parser.add_argument('output', type=Path)
+    args = parser.parse_args()
+    if args.output.exists(): raise ValueError('Keep earlier reports immutable')
+    measured = json.loads(args.dom.read_text(encoding='utf-8'))
+    oracle = json.loads((args.corpus/'fidelity-results.json').read_text(encoding='utf-8'))
+    ink = json.loads((args.corpus/'name-box-metrics.json').read_text(encoding='utf-8'))
+    rows = []
+    for dom in measured['rows']:
+        file = args.corpus/dom['fixture']/'export.pdf'
+        assert sha(file) == dom['pdfSha256'], 'Native PDF changed since DOM measurement binding'
+        original = next(row for row in oracle['fixtures'] if row['fixture'] == dom['fixture'])
+        comparison = next(row for row in original['comparisons'] if row['token'] == 'S01')
+        assert all(failure.startswith('S01: position delta ') for failure in original['failures'])
+        pixels = next(row for row in ink['fixtures'] if row['fixture'] == dom['fixture'])
+        with pdfplumber.open(file) as document:
+            page = document.pages[0]
+            markers = [word for word in page.extract_words() if word['text'] == 'S01']
+            assert len(markers) == 1, 'Missing or duplicate source marker'
+            marker = markers[0]
+            char = next(char for char in page.chars if char['text'] == 'S' and abs(char['x0']-marker['x0'])<0.01 and abs(char['top']-marker['top'])<0.01)
+            assert abs(char['matrix'][1]) < 1e-6 and abs(char['matrix'][2]) < 1e-6, 'Unexpected rotated source'
+            font = next(font.get_object() for font in PdfReader(file).pages[0]['/Resources']['/Font'].values() if str(font.get_object()['/BaseFont']).lstrip('/') == char['fontname'])
+            if '/DescendantFonts' in font: font = font['/DescendantFonts'][0].get_object()
+            descriptor = font['/FontDescriptor'].get_object()
+            descent, ascent = float(descriptor['/Descent']), float(descriptor['/Ascent'])
+            baseline = (page.height-char['matrix'][5])*96/72
+            values = math.compare(dom['rangeTop'],dom['sentinelBaseline'],baseline,char['top']*96/72,char['size']*96/72,descent)
+            assert values['cellFormulaResidualPx'] < 0.01, 'Font-cell formula does not explain native coordinates'
+            assert abs(values['oldCellDeltaMm']-comparison['deltaYmm']) < 0.001, 'Original native oracle mismatch'
+            assert values['baselineDeltaMm'] <= 2 and pixels['visibleInkTopDeltaPx'] <= 3, 'Actual baseline/visible-ink difference needs separate investigation'
+            values.update(fixture=dom['fixture'], originalPass=original['pass'], originalFailures=original['failures'],
+                          classification='INCORRECT_TEST_ASSUMPTION' if not original['pass'] else 'PASS_CONTROL_WITH_SAME_METRIC_MISMATCH',
+                          secondaryCause='PLATFORM_FONT_DESCRIPTOR_METRICS', pdfSha256=sha(file), font=char['fontname'],
+                          descriptorAscent=ascent,descriptorDescent=descent,domRangeTopPx=dom['rangeTop'],
+                          domBaselinePx=dom['sentinelBaseline'],pdfBaselinePx=baseline,pdfCellTopPx=char['top']*96/72,
+                          fontSizePt=char['size'], visibleInkTopDeltaPx=pixels['visibleInkTopDeltaPx'])
+            rows.append(values)
+    assert len(rows)==36
+    report=dict(status='DIAGNOSTIC_COMPLETE_OWNER_DISPOSITION_REQUIRED', policy='NO_ORIGINAL_GATE_OR_GOLDEN_CHANGE',
+                originalOracleSha256=sha(args.corpus/'fidelity-results.json'),domReportSha256=sha(args.dom),
+                inkReportSha256=sha(args.corpus/'name-box-metrics.json'),
+                total=36,originalPass=sum(row['originalPass'] for row in rows),
+                classifiedFailures=sum(not row['originalPass'] for row in rows),
+                maxBaselineDeltaMm=max(row['baselineDeltaMm'] for row in rows),
+                maxCellFormulaResidualPx=max(row['cellFormulaResidualPx'] for row in rows),
+                maxVisibleInkTopDeltaPx=max(row['visibleInkTopDeltaPx'] for row in rows),rows=rows)
+    args.output.write_bytes((json.dumps(report,ensure_ascii=False,indent=2)+'\n').encode())
+    print(f"Classified {report['classifiedFailures']}/36 original failures; max actual baseline delta {report['maxBaselineDeltaMm']:.4f}mm; original gate unchanged")
+
+
+if __name__ == '__main__': main()