CURIA EU Court of Justice Judgment Scraper

The Court of Justice of the EU publishes thousands of judgments, AG opinions, and orders on curia.europa.eu — but there's no bulk download, no public API, and the search UI returns one page at a time. This actor scrapes the official CURIA juris search interface and returns structured records with ECLI identifiers, CELEX numbers, parties, dates, document types, and links to full text and PDFs on EUR-Lex.

Features

• Case-number mode — look up specific cases (e.g. C-311/18) and get every judgment, opinion, and order in the dossier
• Broad-search mode — filter by court, year, and document type to collect large batches across the dossier
• Returns ECLI and CELEX identifiers for unambiguous cross-referencing with EUR-Lex, EUR-Lex SPARQL, and other legal databases
• Direct links to full-text HTML and PDF versions of each document on EUR-Lex
• Handles the Angular SPA with fingerprint-matched Chrome requests — no CAPTCHAs, no manual intervention

Use Cases

• Legal research & LegalTech — build annotated case law corpora, citation graphs, or case chronologies for briefs and analysis
• Academic research — systematic study of EU case law trends, subject-matter frequency, or AG opinion uptake rates
• Compliance & regulatory monitoring — track decisions in specific areas (data protection, competition, state aid) by year and court
• Data journalism — quantitative analysis of CJEU output, parties, and subject matter over time
• RAG / AI legal assistants — index CJEU judgments as structured chunks with reliable identifiers for retrieval

How It Works

CURIA's search frontend is an Angular SPA hosted on infocuria.curia.europa.eu. When you enter a case number, the site redirects to a per-case affair page that expands the full document list inline. The actor uses a Playwright-based renderer with real Chrome fingerprinting (so the backend's headless-browser guard doesn't fire) and extracts case metadata and every document entry directly from the rendered DOM.

In broad-search mode, the actor loads the case listing filtered by court, year, and document type, returning one compact record per case row.

No proxy is required. The site does not use Cloudflare or any third-party bot-management service.

Input

Case-number mode (recommended for targeted lookups)

{
  "caseNumbers": ["C-311/18", "C-362/14"],
  "maxItems": 50
}

Broad-search mode (batch collection by filter)

{
  "court": "C",
  "year": "2023",
  "documentType": "Judgment",
  "maxItems": 200
}

Input Parameters

Parameter	Type	Required	Description
caseNumbers	array of strings	No	CJEU case numbers to look up (e.g. C-311/18, T-125/22). Leave empty to use the filter parameters below.
court	string	No	Filter by court. C = Court of Justice, T = General Court, F = Civil Service Tribunal (archived). Blank = all courts.
year	string	No	Filter to a specific year, e.g. 2023. Used only when caseNumbers is empty.
documentType	string	No	Filter by document type: Judgment, Opinion (AG opinions), Order. Blank = all types.
maxItems	integer	Yes	Maximum number of records to return. Default: 100.

Tip: caseNumbers and filter parameters can be combined. When caseNumbers is provided, the actor uses the direct per-case URL for each number and the court filter is applied as a URL hint; year and documentType are used for broad-search requests only.

Output

Each item in the dataset represents one document (judgment, opinion, or order). In broad-search mode, each item represents one case row with only the case number and parties populated.

Output Fields

Field	Description
caseNumber	Case number, e.g. C-311/18
ecli	European Case Law Identifier, e.g. ECLI:EU:C:2020:559
parties	Applicant and defendant parties, e.g. Facebook Ireland and Schrems
court	Court name: Court of Justice or General Court
documentType	Document type: Judgment, Opinion of AG, Order, or Abstract
dateOfDocument	Date the document was issued (DD/MM/YYYY format as shown on CURIA)
subjectMatter	Subject matter / keywords from the case title (when available)
judgmentTextUrl	URL to the full HTML text on EUR-Lex
judgmentPdfUrl	URL to the PDF version on EUR-Lex
celexNumber	CELEX identifier, e.g. 62018CJ0311 — use this to look up the document in EUR-Lex APIs
caseUrl	URL to the case listing page on CURIA
scrapedAt	ISO-8601 timestamp of when the record was scraped

Example Record

{
  "caseNumber": "C-311/18",
  "ecli": "ECLI:EU:C:2020:559",
  "parties": "Facebook Ireland and Schrems",
  "court": "Court of Justice",
  "documentType": "Judgment",
  "dateOfDocument": "16/07/2020",
  "subjectMatter": "",
  "judgmentTextUrl": "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:62018CJ0311",
  "judgmentPdfUrl": "https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:62018CJ0311",
  "celexNumber": "62018CJ0311",
  "caseUrl": "https://infocuria.curia.europa.eu/tabs/affair?publishedId=C-311%2F18",
  "scrapedAt": "2026-06-12T16:14:00.000Z"
}

FAQ

Does this cover all three CJEU courts? Yes — Court of Justice (C-series), General Court (T-series), and the archived Civil Service Tribunal (F-series). Use the court filter to narrow to one.

What's the difference between the ECLI and CELEX number? ECLI (ECLI:EU:C:2020:559) is the pan-European case identifier used across legal databases. CELEX (62018CJ0311) is the EUR-Lex internal identifier used in that database's API and URLs. Both are returned when available.

How far back does the data go? CURIA's juris search covers the full case law archive. The scraper does not impose a year floor. Use the year filter to target specific years.

Are AG opinions included? Yes. Advocate-General opinions appear as separate document entries within a case dossier and are returned with documentType: "Opinion of AG".

What if a case has multiple documents (e.g. judgment + AG opinion + order)? Each document becomes a separate output record. A case like C-311/18 (Data Protection Commissioner v Facebook Ireland and Schrems II) has three document entries.

Is there a pagination limit? Broad-search mode retrieves one page of the case list (up to ~20 rows). For systematic collection of large result sets, provide explicit caseNumbers lists or run multiple filtered batches by year.

Can I get the full text of the judgment? The actor returns the EUR-Lex URL (judgmentTextUrl) and PDF URL (judgmentPdfUrl). It does not fetch or return the full document body — use the links to retrieve the text via EUR-Lex directly or with a separate HTTP request.

Need More Features?

Open a GitHub issue or contact us at orbtop.com. Common extension requests: full-text PDF extraction, pagination across multiple result pages, bulk case-number input from a spreadsheet URL, and filtering by subject-matter keyword.

Why Use This Actor?

CURIA's search is built for human browsers — there is no API, no sitemap, no bulk export. The Angular SPA requires JavaScript execution and browser fingerprint matching. This actor handles all of that and delivers clean structured output in one run. The ECLI and CELEX identifiers make it straightforward to join with EUR-Lex datasets, CrossRef citations, or your own legal research database.