{
  "repository": "https://github.com/Murrough-Foley/web-content-extraction-benchmark",
  "revision": "c039d5ee9f5a3a984a0e167e63aacd04e76e78a9",
  "source_test_count": 511,
  "source_test_sha256": "080a35a399880bb74ce23f40d9a1008f974500716af183a990f8701ab3e5e7bb",
  "qualified_test_count": 504,
  "qualified_source_sha256": "a76148767d25beb1adb49bc4aa3c2afb97c69f8464e8383f55a8a8c153b42097",
  "original_selection_manifest_sha256": "2f42daa5a793cdf64654bcc1e58e332315c0d972bbc9e2e06510cfa59ea94754",
  "original_distribution": {
    "parent": {
      "article": 257,
      "collection": 34,
      "product": 28,
      "listing": 40,
      "forum": 51,
      "documentation": 42,
      "service": 59
    },
    "sample": {
      "article": 256,
      "collection": 33,
      "product": 27,
      "listing": 39,
      "documentation": 42,
      "service": 57,
      "forum": 50
    }
  },
  "changes": "Remove seven source-invalid tasks; retain every selected HTML and ground-truth file byte-for-byte. No reference edits.",
  "license": "CC-BY-4.0",
  "copyright": "Copyright (c) 2026 Murrough Foley",
  "evaluator_sha256": "ae4ad6299e190177fbb04a7c1190077fc6086b3fa0f31fb37a6538a5e979c559",
  "citation": {
    "title": "WCXB: A Multi-Type Web Content Extraction Benchmark",
    "author": "Murrough Foley",
    "year": 2026,
    "url": "https://github.com/Murrough-Foley/web-content-extraction-benchmark"
  }
}