{
  "generated_on": "2026-09-21",
  "fixture_sha256": "c31141afd73bdd40e09fceecd2053c069121c23c5c364d3105e73ff8e079f1a5",
  "commands": {
    "pdftotext": "pdftotext version 26.08.0",
    "tesseract": "tesseract 5.5.3"
  },
  "pages": [
    {
      "page": 1,
      "page_type": "born-digital",
      "visible_reference": "PAGE 1 — DIGITAL TEXT Batch AX-104 contains 27 folios. Keep section 4.2 with the appendix.",
      "pdftotext": "PAGE 1 — DIGITAL TEXT Batch AX-104 contains 27 folios. Keep section 4.2 with the appendix.",
      "tesseract_from_render": "PAGE 1 — DIGITAL TEXT Batch AX-104 contains 27 folios. Keep section 4.2 with the appendix.",
      "extraction_matches_visible_reference": true,
      "ocr_matches_visible_reference": true
    },
    {
      "page": 2,
      "page_type": "image-only",
      "visible_reference": "PAGE 2 — IMAGE ONLY Batch BX-205 contains 38 folios. Keep section 5.3 with the appendix.",
      "pdftotext": "",
      "tesseract_from_render": "PAGE 2 — IMAGE ONLY Batch BX-205 contains 38 folios. Keep section 5.3 with the appendix.",
      "extraction_matches_visible_reference": false,
      "ocr_matches_visible_reference": true
    },
    {
      "page": 3,
      "page_type": "existing-correct-ocr",
      "visible_reference": "PAGE 3 — EXISTING OCR Batch CX-306 contains 49 folios. Keep section 6.4 with the appendix.",
      "pdftotext": "PAGE 3 — EXISTING OCR Batch CX-306 contains 49 folios. Keep section 6.4 with the appendix.",
      "tesseract_from_render": "PAGE 3 — EXISTING OCR Batch CX-306 contains 49 folios. Keep section 6.4 with the appendix.",
      "extraction_matches_visible_reference": true,
      "ocr_matches_visible_reference": true
    },
    {
      "page": 4,
      "page_type": "existing-wrong-ocr",
      "visible_reference": "PAGE 4 — MISLEADING OCR Batch DX-407 contains 58 folios. Keep section 7.5 with the appendix.",
      "pdftotext": "PAGE 4 — MISLEADING OCR Batch DX-4O7 contains 53 folios. Keep section 7.8 with the appendix.",
      "tesseract_from_render": "PAGE 4 — MISLEADING OCR Batch DX-407 contains 58 folios. Keep section 7.5 with the appendix.",
      "extraction_matches_visible_reference": false,
      "ocr_matches_visible_reference": true
    }
  ],
  "scope": "Controlled English fixture; not an OCR accuracy benchmark."
}