Ametlikud Teadaanded

Official Notices

Access

• URI contract and current type slugs: https://www.ametlikudteadaanded.ee/avalik/uriotsing
• Interactive search: POST https://www.ametlikudteadaanded.ee/avalik/otsing
• Public notice URIs: https://www.ametlikudteadaanded.ee/ee/...

Retrieve

Build a URI in this order:

/ee/{publisher}/{main-type}/{subtype}/{year}/{month}/{day}/{notice-number}/{format}

Trailing components may be omitted. Use - for an unused publisher, main type, or subtype in the middle. The optional format is xml, rdf, or txt; without it the response is HTML. For example, /ee/-/advokatuur/xml returns a public XML collection for the advokatuur main type.

Read publisher and type slugs from the URI contract page. XML collections use root element at:teadaanded; each at:teadaanne includes the notice number, canonical URL, type, legal basis, purpose, publisher, publication data, and notice-specific fields.

For ad hoc name/keyword searches, POST do_search=1, o__search_term, and optional o__teate_liigid, o__teate_alaliik_list, o__avaldamise_kuupaev_alates, and o__avaldamise_kuupaev_kuni to /avalik/otsing. Structured person search is o__isiku_tyyp=F with o__eesnimi / o__perekonnanimi / o__fkood.

Three things about that POST, all verified 2026-08-01:

• It needs a session cookie or it silently does not search. The register is behind Cloudflare bot management; without a __cf_bm cookie the POST answers HTTP 200 with the empty search form back — no error, no status code to check. GET any page on the host first and carry its cookies into the POST.
• GET /avalik/otsing is HTTP 405 ("kasutatav ainult järgmiste päringumeetoditega: POST"). Fetch the URI-contract page instead when you just want the form vocabulary.
• Dates are dd.mm.yyyy only. 20.07.2026 filters; 2026-07-20 is accepted and matches nothing.

claim `GET /avalik/otsing` is HTTP 405
GET https://www.ametlikudteadaanded.ee/avalik/otsing
expect-status 405
expect /kasutatav ainult järgmiste päringumeetoditega: POST/
# The control is the URI-contract page, on the same host, which must still
# answer 200. Without it, a day when the whole register sits behind a bot check
# would "confirm" this limitation while the source was unusable.
control GET https://www.ametlikudteadaanded.ee/avalik/uriotsing

o__teate_liigid takes a numeric main-type id (1 = Pärimisteated; 51 types in the form's own select). These ids are a different vocabulary from the URI slugs above — parimisteated is the slug, 1 is the id.

Inheritance proceedings (pärimisteated)

This is the public route to Estonian inheritance proceedings. The pärimisregister itself is ID-authenticated (see ## Limits), but every proceeding is announced here under pärimisseaduse (PärS) § 168, so "has a pärimismenetlus been started for X" is answerable from these notices alone.

Main type parimisteated, one day per request:

/ee/-/parimisteated/-/2026/07/20/xml → 55 notices (verified 2026-07-27)

Subtypes, as the {subtype} component (all five verified live; an empty day returns a valid 146-byte collection, not an error):

• parimismenetluse-algatamine — proceeding opened (49 of that day's 55)
• parijate-valjaselgitamise-yleskutsemen — call to identify heirs
• parimismenetluse-algat-ja-parijate-valjaselgitamise-yleskutsemen — both at once
• yleskutsemen-teade-parandaja-kohust-kindlakstegemiseks — call to establish the estate's debts
• parimistunnistuse-keht-tunnist-ja-taiend-parimismen-labiviimine — succession certificate invalidated / supplementary proceeding

Each at:teadaanne carries teate_number, url, liik, mall, oiguslik_alus_omastav with an oigusliku_aluse_viide deep link into Riigi Teataja, the notary as andmeandja/avaldaja with contact details, avaldamise_kpv, and puudutatud_isik naming the deceased. The renunciation deadline itself is PärS, not this feed — read the act through legal-acts-data and cite the § rather than restating a remembered period.

Return

• Preserve notice number, canonical URL, publication date, publisher, main/subtype, legal basis, status, matched party fields, query URI, and retrieval time.
• Keep notice publication distinct from the underlying court, insolvency, procurement, or administrative action.

Limits

• URI searches return at most 1,000 results. Narrow by type and date when completeness matters.
• A collection URI without a date does not answer. /ee/-/parimisteated/-/xml hangs and returns nothing (25 s, 0 bytes, verified 2026-07-27). Always pin {year}/{month}/{day}; to cover a range, request each day.
• The open URI interface excludes archived notices and notices addressed to natural persons for service.
• Notice XML can contain personal data lawfully published in the source — pärimisteated name the deceased and carry their isikukood. Return only fields needed for the user's request.
• These notices say a proceeding was announced, not how it ended. The pärimisregister, which holds the outcome, is free but ID-authenticated (Notarite Koja iseteenindus / eesti.ee), so there is no public route to "who inherited".

Verify

• Require the contract page to contain the documented URI pattern and current type slugs.
• Require a collection URI ending in /xml to return parseable XML with at:teadaanded and at least one at:teadaanne containing teate_number and url.
• The search ran only if the response contains Otsingutulemused (hits) or Ei leitud ühtegi teadaannet! (a real zero). Neither present = the register handed back its form without searching (the cookie case above), and that page is 85,314 bytes where a genuine zero is 85,493 — nothing a size or status check separates. Reporting "no proceeding has been started" off that page is the worst answer this source can produce.
• Same URI without /xml is the register's human-readable day view (HTTP 200) — cite that, not the XML.

Module API

findNotices({term, firstName, lastName, type, from, to}) -> {source, page, ran, note, matched, truncated, rows} noticesOn(date, {type, subtype, publisher}) -> {source, page, date, note, totals, rows}