Getting started
How collections are updated
Imports, changes and on-demand retrieval: how official documents reach LawDB.
From official document to search result
LawDB first imports a source’s holdings. Recurring update runs then collect new or changed documents. Official identifiers, metadata, language versions and texts are processed together; indexing and text processing may finish at different times.
Scheduled update intervals
These intervals describe configured import schedules when the relevant source is enabled in operation. They do not guarantee that every newly published document is fully available within that interval.
| Source | Schedule |
|---|---|
| NeuRIS legislation and case law | Every six hours |
| EU consolidations: CELLAR | Hourly |
| EU case law: InfoCuria | Hourly |
| International Court of Justice and Permanent Court of Arbitration | Daily |
| HUDOC and Council of Europe treaties | Daily when enabled; HUDOC also has a weekly full reconciliation |
| Gesetze im Internet and other holdings imports | Supplementary import runs; no fixed recurring run in this schedule |
What update runs cover
NeuRIS imports evaluate change information; CELLAR uses notifications for consolidated versions. Other sources are reconciled through date windows or their catalogues. Overlapping windows help capture late changes.
Historical versions and repealed acts may remain in the holdings. A historical result therefore does not necessarily represent current law. Check the version and status information returned.
On-demand retrieval
Search uses the local index by default. With include_external_sources: true, search_legal_documents can additionally query supported, enabled official sources. This is slower and is not available for every source.
Given a known identifier, missing documents or language texts may be retrieved from the source. status: fetching and retry_after_seconds mean retrieval is still running: repeat the same call after the stated delay. Retrieving an older stored document may also trigger a background refresh.
Availability and AI analyses
Source publication times, outages, queues, PDF processing and indexing can delay availability. unavailable_sources in a search response identifies sources that could not be reached.
AI summaries are generated separately and checked against source quotations. They are not available for every document and may appear later than the original text.