{
  "artifacts_index": [
    {
      "content_hash": null,
      "document_id": "doc01",
      "id": "doc01:image",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": null,
      "provenance": null,
      "type": "image",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc01",
      "id": "doc01:precomputed_corr:raw_text",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": "ocr",
      "provenance": null,
      "type": "raw_text",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc01",
      "id": "doc01:precomputed_tess:raw_text",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": "ocr",
      "provenance": null,
      "type": "raw_text",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc02",
      "id": "doc02:image",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": null,
      "provenance": null,
      "type": "image",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc02",
      "id": "doc02:precomputed_corr:raw_text",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": "ocr",
      "provenance": null,
      "type": "raw_text",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc02",
      "id": "doc02:precomputed_tess:raw_text",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": "ocr",
      "provenance": null,
      "type": "raw_text",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc03",
      "id": "doc03:image",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": null,
      "provenance": null,
      "type": "image",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc03",
      "id": "doc03:precomputed_corr:raw_text",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": "ocr",
      "provenance": null,
      "type": "raw_text",
      "uri": "<PATH>"
    },
    {
      "content_hash": null,
      "document_id": "doc03",
      "id": "doc03:precomputed_tess:raw_text",
      "pipeline_name": "ocr_then_correct",
      "produced_by_step": "ocr",
      "provenance": null,
      "type": "raw_text",
      "uri": "<PATH>"
    }
  ],
  "manifest": {
    "adapter_kwargs": {
      "corrector": {
        "source_label": "corr"
      },
      "ocr": {
        "source_label": "tess"
      }
    },
    "code_version": "charac-1.0",
    "corpus_name": "charac",
    "dependencies_lock": {
      "CacheControl": "0.14.4",
      "Jinja2": "3.1.6",
      "MarkupSafe": "3.0.3",
      "PyGObject": "3.48.2",
      "PyJWT": "2.7.0",
      "PyYAML": "6.0.1",
      "Pygments": "2.20.0",
      "RapidFuzz": "3.14.5",
      "annotated-doc": "0.0.4",
      "annotated-types": "0.7.0",
      "anyio": "4.13.0",
      "argcomplete": "3.1.4",
      "ast_serialize": "0.5.0",
      "bandit": "1.9.4",
      "blinker": "1.7.0",
      "boolean.py": "5.0",
      "certifi": "2026.2.25",
      "charset-normalizer": "3.4.6",
      "click": "8.4.0",
      "colorama": "0.4.6",
      "conan": "2.27.0",
      "coverage": "7.14.0",
      "cryptography": "41.0.7",
      "cyclonedx-python-lib": "11.7.0",
      "dbus-python": "1.3.2",
      "defusedxml": "0.7.1",
      "distro": "1.9.0",
      "fastapi": "0.136.1",
      "fasteners": "0.20",
      "filelock": "3.29.0",
      "h11": "0.16.0",
      "httpcore": "1.0.9",
      "httplib2": "0.20.4",
      "httptools": "0.7.1",
      "httpx": "0.28.1",
      "idna": "3.11",
      "iniconfig": "2.3.0",
      "jiwer": "4.0.0",
      "launchpadlib": "1.11.0",
      "lazr.restfulclient": "0.14.6",
      "lazr.uri": "1.0.6",
      "librt": "0.11.0",
      "license-expression": "30.4.4",
      "markdown-it-py": "4.2.0",
      "mdurl": "0.1.2",
      "msgpack": "1.1.2",
      "mypy": "2.1.0",
      "mypy_extensions": "1.1.0",
      "numpy": "2.4.6",
      "oauthlib": "3.2.2",
      "packageurl-python": "0.17.6",
      "packaging": "24.0",
      "patch-ng": "1.18.1",
      "pathspec": "1.1.1",
      "picarones": "1.1.0.dev311",
      "pillow": "12.2.0",
      "pip": "24.0",
      "pip-api": "0.0.34",
      "pip-requirements-parser": "32.0.1",
      "pip_audit": "2.10.0",
      "platformdirs": "4.9.6",
      "pluggy": "1.6.0",
      "py-serializable": "2.1.0",
      "pydantic": "2.13.4",
      "pydantic_core": "2.46.4",
      "pyparsing": "3.1.1",
      "pytesseract": "0.3.13",
      "pytest": "9.0.3",
      "pytest-cov": "7.1.0",
      "pytest-timeout": "2.4.0",
      "python-apt": "2.7.7+ubuntu5.2",
      "python-dateutil": "2.9.0.post0",
      "python-dotenv": "1.2.2",
      "python-multipart": "0.0.29",
      "requests": "2.33.1",
      "rich": "15.0.0",
      "setuptools": "68.1.2",
      "six": "1.16.0",
      "sortedcontainers": "2.4.0",
      "starlette": "1.0.0",
      "stevedore": "5.8.0",
      "toml": "0.10.2",
      "tomli": "2.4.1",
      "tomli_w": "1.2.0",
      "tqdm": "4.67.3",
      "typing-inspection": "0.4.2",
      "typing_extensions": "4.15.0",
      "urllib3": "2.6.3",
      "uvicorn": "0.47.0",
      "uvloop": "0.22.1",
      "wadllib": "1.3.6",
      "watchfiles": "1.2.0",
      "websockets": "16.0",
      "wheel": "0.42.0",
      "xmltodict": "0.13.0",
      "yq": "3.1.0"
    },
    "metadata": {
      "orchestrator": "picarones.app.services.run_orchestrator"
    },
    "n_documents": 3,
    "pipeline_names": [
      "ocr_then_correct"
    ],
    "pipeline_specs": [
      {
        "description": "",
        "initial_inputs": [
          "image"
        ],
        "name": "ocr_then_correct",
        "steps": [
          {
            "adapter_name": "ocr",
            "id": "ocr",
            "input_types": [
              "image"
            ],
            "inputs_from": {},
            "kind": "step",
            "output_types": [
              "raw_text"
            ],
            "params": {}
          },
          {
            "adapter_name": "corrector",
            "id": "corrector",
            "input_types": [
              "image",
              "raw_text"
            ],
            "inputs_from": {
              "raw_text": "ocr"
            },
            "kind": "step",
            "output_types": [
              "corrected_text"
            ],
            "params": {}
          }
        ]
      }
    ],
    "system_binaries_lock": {},
    "view_specs": [
      {
        "candidate_types": [
          "alto_xml",
          "canonical_document",
          "corrected_text",
          "page_xml",
          "raw_text"
        ],
        "char_exclude": null,
        "description": "Compare les sorties textuelles finales après projection éventuelle (ALTO/PAGE/markdown → texte plat).",
        "ignored_dimensions": [
          "block_structure",
          "confidence",
          "formatting",
          "geometry",
          "ids",
          "reading_order"
        ],
        "metric_names": [
          "cer",
          "mer",
          "wer",
          "wil"
        ],
        "name": "text_final",
        "normalization_profile": null,
        "projection": null,
        "projections_by_source_type": {
          "alto_xml": {
            "params": {},
            "projector_name": "alto_to_text",
            "source_type": "alto_xml",
            "target_type": "raw_text"
          },
          "canonical_document": {
            "params": {},
            "projector_name": "canonical_to_text",
            "source_type": "canonical_document",
            "target_type": "raw_text"
          },
          "page_xml": {
            "params": {},
            "projector_name": "page_to_text",
            "source_type": "page_xml",
            "target_type": "raw_text"
          }
        },
        "warnings": [
          "Cette vue compare les sorties textuelles finales après projection éventuelle.  Les pipelines qui produisent ALTO/PAGE/markdown sont projetés vers du texte plat — leurs structures spatiale et documentaire ne sont PAS évaluées ici.  Pour évaluer la qualité ALTO, voir AltoView (S15)."
        ]
      }
    ]
  },
  "pipeline_results": [
    {
      "document_id": "doc01",
      "pipeline_name": "ocr_then_correct",
      "step_results": [
        {
          "error": null,
          "produced_artifacts": {
            "raw_text": "doc01:precomputed_tess:raw_text"
          },
          "step_id": "ocr",
          "succeeded": true
        },
        {
          "error": "missing_output: ['corrected_text']",
          "produced_artifacts": {},
          "step_id": "corrector",
          "succeeded": false
        }
      ],
      "succeeded": false
    },
    {
      "document_id": "doc02",
      "pipeline_name": "ocr_then_correct",
      "step_results": [
        {
          "error": null,
          "produced_artifacts": {
            "raw_text": "doc02:precomputed_tess:raw_text"
          },
          "step_id": "ocr",
          "succeeded": true
        },
        {
          "error": "missing_output: ['corrected_text']",
          "produced_artifacts": {},
          "step_id": "corrector",
          "succeeded": false
        }
      ],
      "succeeded": false
    },
    {
      "document_id": "doc03",
      "pipeline_name": "ocr_then_correct",
      "step_results": [
        {
          "error": null,
          "produced_artifacts": {
            "raw_text": "doc03:precomputed_tess:raw_text"
          },
          "step_id": "ocr",
          "succeeded": true
        },
        {
          "error": "missing_output: ['corrected_text']",
          "produced_artifacts": {},
          "step_id": "corrector",
          "succeeded": false
        }
      ],
      "succeeded": false
    }
  ],
  "view_results": [
    {
      "candidate_artifact_id": "doc01:precomputed_tess:raw_text",
      "document_id": "doc01",
      "failed_metrics": {},
      "ground_truth_artifact_id": "doc01:gt:raw_text",
      "ignored_dimensions": [
        "block_structure",
        "confidence",
        "formatting",
        "geometry",
        "ids",
        "reading_order"
      ],
      "metric_values": {
        "cer": 0.05,
        "mer": 0.25,
        "wer": 0.25,
        "wil": 0.4375
      },
      "pipeline_name": "ocr_then_correct",
      "projection_report": null,
      "view_name": "text_final",
      "warnings": [
        "Cette vue compare les sorties textuelles finales après projection éventuelle.  Les pipelines qui produisent ALTO/PAGE/markdown sont projetés vers du texte plat — leurs structures spatiale et documentaire ne sont PAS évaluées ici.  Pour évaluer la qualité ALTO, voir AltoView (S15)."
      ]
    },
    {
      "candidate_artifact_id": "doc01:precomputed_corr:raw_text",
      "document_id": "doc01",
      "failed_metrics": {},
      "ground_truth_artifact_id": "doc01:gt:raw_text",
      "ignored_dimensions": [
        "block_structure",
        "confidence",
        "formatting",
        "geometry",
        "ids",
        "reading_order"
      ],
      "metric_values": {
        "cer": 0.05,
        "mer": 0.25,
        "wer": 0.25,
        "wil": 0.4375
      },
      "pipeline_name": "ocr_then_correct",
      "projection_report": null,
      "view_name": "text_final",
      "warnings": [
        "Cette vue compare les sorties textuelles finales après projection éventuelle.  Les pipelines qui produisent ALTO/PAGE/markdown sont projetés vers du texte plat — leurs structures spatiale et documentaire ne sont PAS évaluées ici.  Pour évaluer la qualité ALTO, voir AltoView (S15)."
      ]
    },
    {
      "candidate_artifact_id": "doc02:precomputed_tess:raw_text",
      "document_id": "doc02",
      "failed_metrics": {},
      "ground_truth_artifact_id": "doc02:gt:raw_text",
      "ignored_dimensions": [
        "block_structure",
        "confidence",
        "formatting",
        "geometry",
        "ids",
        "reading_order"
      ],
      "metric_values": {
        "cer": 0.05,
        "mer": 0.25,
        "wer": 0.25,
        "wil": 0.4375
      },
      "pipeline_name": "ocr_then_correct",
      "projection_report": null,
      "view_name": "text_final",
      "warnings": [
        "Cette vue compare les sorties textuelles finales après projection éventuelle.  Les pipelines qui produisent ALTO/PAGE/markdown sont projetés vers du texte plat — leurs structures spatiale et documentaire ne sont PAS évaluées ici.  Pour évaluer la qualité ALTO, voir AltoView (S15)."
      ]
    },
    {
      "candidate_artifact_id": "doc02:precomputed_corr:raw_text",
      "document_id": "doc02",
      "failed_metrics": {},
      "ground_truth_artifact_id": "doc02:gt:raw_text",
      "ignored_dimensions": [
        "block_structure",
        "confidence",
        "formatting",
        "geometry",
        "ids",
        "reading_order"
      ],
      "metric_values": {
        "cer": 0.05,
        "mer": 0.25,
        "wer": 0.25,
        "wil": 0.4375
      },
      "pipeline_name": "ocr_then_correct",
      "projection_report": null,
      "view_name": "text_final",
      "warnings": [
        "Cette vue compare les sorties textuelles finales après projection éventuelle.  Les pipelines qui produisent ALTO/PAGE/markdown sont projetés vers du texte plat — leurs structures spatiale et documentaire ne sont PAS évaluées ici.  Pour évaluer la qualité ALTO, voir AltoView (S15)."
      ]
    },
    {
      "candidate_artifact_id": "doc03:precomputed_tess:raw_text",
      "document_id": "doc03",
      "failed_metrics": {},
      "ground_truth_artifact_id": "doc03:gt:raw_text",
      "ignored_dimensions": [
        "block_structure",
        "confidence",
        "formatting",
        "geometry",
        "ids",
        "reading_order"
      ],
      "metric_values": {
        "cer": 0.05,
        "mer": 0.25,
        "wer": 0.25,
        "wil": 0.4375
      },
      "pipeline_name": "ocr_then_correct",
      "projection_report": null,
      "view_name": "text_final",
      "warnings": [
        "Cette vue compare les sorties textuelles finales après projection éventuelle.  Les pipelines qui produisent ALTO/PAGE/markdown sont projetés vers du texte plat — leurs structures spatiale et documentaire ne sont PAS évaluées ici.  Pour évaluer la qualité ALTO, voir AltoView (S15)."
      ]
    },
    {
      "candidate_artifact_id": "doc03:precomputed_corr:raw_text",
      "document_id": "doc03",
      "failed_metrics": {},
      "ground_truth_artifact_id": "doc03:gt:raw_text",
      "ignored_dimensions": [
        "block_structure",
        "confidence",
        "formatting",
        "geometry",
        "ids",
        "reading_order"
      ],
      "metric_values": {
        "cer": 0.05,
        "mer": 0.25,
        "wer": 0.25,
        "wil": 0.4375
      },
      "pipeline_name": "ocr_then_correct",
      "projection_report": null,
      "view_name": "text_final",
      "warnings": [
        "Cette vue compare les sorties textuelles finales après projection éventuelle.  Les pipelines qui produisent ALTO/PAGE/markdown sont projetés vers du texte plat — leurs structures spatiale et documentaire ne sont PAS évaluées ici.  Pour évaluer la qualité ALTO, voir AltoView (S15)."
      ]
    }
  ]
}