CoolFace
Datasetpublic

justicedao/ipfs_netherlands_laws

IPFS Netherlands Laws Hugging Face target: justicedao/ipfs_netherlands_laws. This dataset packages Netherlands law records with deterministic IPFS Content IDs. Each row includes a cid and content_address; article rows also include the parent law_cid. This is a quality-audited catalog-backed Netherlands snapshot from official Dutch government sources. It is not the full Dutch legal corpus: the persistent catalog contains 42,956 discovered BWBR identifiers, of which 5,000 are… See the full description on the dataset page: https://huggingface.co/datasets/justicedao/ipfs_netherlands_laws.

sourceHugging Faceotherupdated 3mo agoView on Hugging Face
0likes38downloads
Dataset Card

IPFS Netherlands Laws

Hugging Face target: justicedao/ipfs_netherlands_laws.

This dataset packages Netherlands law records with deterministic IPFS Content IDs. Each row includes a cid and content_address; article rows also include the parent law_cid.

This is a quality-audited catalog-backed Netherlands snapshot from official Dutch government sources. It is not the full Dutch legal corpus: the persistent catalog contains 42,956 discovered BWBR identifiers, of which 5,000 are complete in this snapshot: 4,999 parsed law documents plus one permanent official HTTP 400 failure logged outside the packaged rows.

Scrape date: 2026-06-27T14:04:23.964996

Exact corpus size in this package: 4999 law records, 89737 article records, and 94736 CID index rows.

Full BWB discovery inventory: unknown unique BWBR identifiers from unknown official SRU record(s), with unknown failed discovery page(s).

This refresh includes parser coverage improvements for older/French heading styles such as Article I.er, plus run metadata diagnostics that distinguish article-producing laws, parser-missing article cases, and genuinely unnumbered/non-article documents.

Quality audit note: this snapshot was rebuilt after parser-noise cleanup for official website UI chrome such as relation, permanent-link, print, save, and wetstechnical-information controls. Duplicate article identifiers caused by repeated article numbers in different hierarchy branches are disambiguated with deterministic hierarchy/text-derived suffixes. The quality audit reports duplicate IDs/CIDs as zero, clean status inheritance, clean hierarchy/citation reconstruction, and passing sampled CID/vector/BM25/KG retrieval validation.

Historical/former laws are preserved. Consumers should use law_status, is_current, valid_from, valid_to, effective_date, retrieved_at, status_source, status_confidence, and status_note to distinguish current law from historical, repealed, superseded, or unknown-status records. These fields are part of the content-addressed payload, so CIDs change when official status/version metadata changes. This dataset is not legal advice and does not validate legal force beyond the official metadata parsed.

Scrape command:

bash
python -m ipfs_datasets_py.processors.legal_scrapers.netherlands_laws scrape --output-dir /home/sofiya/ipfs_datasets_py/ipfs_datasets_py/processors/legal_scrapers/netherlands_laws/datasets/raw/nl_catalog_batch_250_20260626 --max_seed_pages 0 --crawl_depth 0 --rate_limit_delay 1.1 --skip_existing true --resume true

Current package counts:

  • —Laws: 4999
  • —Articles: 89737
  • —CID index rows: 94736
  • —Unique laws discovered before any document cap: 500
  • —Documents failed: 0
  • —Article-producing laws: 4847
  • —Non-article-producing laws: 152
  • —Current laws: 1241
  • —Historical/repealed/superseded laws: 3736
  • —Unknown-status laws: 22
  • —Ambiguous-status laws: 22

Remaining limitations: this package should only be described as the full Dutch corpus when the run metadata shows uncapped official discovery and every discovered official BWBR document was parsed or explicitly logged as failed/skipped. Otherwise it is a verified shard or discovered-corpus subset.