Japan data.go.jp Open Data Catalog (CKAN API)
Pricing
from $2.00 / 1,000 record scrapeds
Unofficial client for Japan's government open-data catalog data.go.jp (CKAN API v3, no key). Search datasets across all ministries and municipalities, or download a dataset's CSV records as structured JSON (Shift_JIS aware). Per-dataset license shown via licenseId; default PDL1.0.
search_datasets (default) finds datasets on data.go.jp (CKAN package_search) and returns their metadata incl. resource URLs and per-resource license ids — set at least one of query / organization / tags / group / format. get_records downloads a dataset's CSV resource(s) and returns the rows as structured JSON — set datasetId or resourceUrl. Example: "search_datasets".
search_datasets mode: free-text query over dataset titles/descriptions (CKAN q), Japanese works best, e.g. "人口" (population), "医療機関", "気象". At least one of query / organization / tags / group / format is required in search_datasets mode.
search_datasets mode (optional): restrict to a publisher by its CKAN organization name, e.g. "org_1600" (厚生労働省 / MHLW) or "org_1100" (総務省 / MIC). Find names in a search_datasets run (the organizationId field).
search_datasets mode (optional): restrict to a category by its CKAN group name (find group names in a search_datasets run's groups field).
search_datasets mode (optional): restrict to datasets that have a resource in this format, e.g. "CSV", "XLSX", "PDF", "JSON" (case-insensitive). Useful to find datasets you can then pull with get_records.
get_records mode (REQUIRED there unless resourceUrl is given): the CKAN dataset name/slug to read, e.g. "mhlw_20260624_2024" — take it from a search_datasets run (the datasetId field). The Actor resolves the dataset's CSV/TSV resources and downloads their rows.
get_records mode (alternative to datasetId): fetch this specific CSV resource URL directly (an absolute http(s) URL from a dataset's resources[].url).
get_records mode: character encoding of the CSV body. auto (default) sniffs UTF-8 (incl. BOM) then falls back to Shift_JIS — many Japanese open-data CSVs are Shift_JIS even when the HTTP header claims UTF-8. Spellings such as "utf8", "Shift-JIS", "sjis", "cp932" are accepted.
Maximum number of records (datasets or CSV rows) to output, 1-100000. PPE charges per record. Example: 1000.
Hard safety cap on upstream requests per run, 1-25 (search pages 1000 datasets per request; get_records may fetch several resources). Politeness (1 connection, >= 1.2 s spacing, exponential backoff on 429/5xx) is enforced in code. Example: 10.
Apify proxy settings. Default is NO proxy (direct connection) — a public API rarely needs one. The Actor backs off exponentially on 429/5xx and fails visibly on a persistent block; it never attempts rate-limit evasion.
{ "useApifyProxy": false}