Python-KontAKTJournalize

Keeps a GetOrganized (GO) journaliseringssag (an AKT case) in sync with a KontAKT aktindsigt case, and journalises the correspondence + the final delivery onto it.

KontAKT triggers this whenever a case is created, when Emne / sagsbehandler / modtaget changes, when a reply is sent, and when the case is finished.

What it does

Queue-driven; the mode is set by mode in the input:

• create_case — creates the AKT case in GO with the MTM aktindsigt case profile (Sagsprofil = MTM Aktindsigt, Facet = A53 Aktindsigtsanmodning mv., Modtaget, Title = Emne) and sets the KontAKT caseworker as CaseOwner (the PeoplePicker "hack" — best-effort). Calls back with the new GO case number.
• update_metadata — updates Title / Modtaget (via /_goapi/Cases/Metadata) and CaseOwner when they change in KontAKT.
• journalize_email — fetches a sent e-mail from KontAKT, renders it to PDF (LibreOffice), then adds + journalises it on the GO case.
• journalize_ref — fired when ONE GO/Nova case is shared: fetches its delivered files from KontAKT's file store (by doc-id, via …/content), adds + journalises them on the GO case, and reports doc_id → go_doc_id back.
• journalize_folder — fired when the WHOLE case is shared: journalises every delivered file for the case, mapping each to its doc_id.
• delete_doc — deletes a document from GO (by go_doc_id) after it was deleted/removed in KontAKT.
• generate_aktliste — (re)generates ONE GO/Nova case's aktliste (PDF + Excel via oomtm.reports) and journalises both onto the GO case. Fired on every doc-list change (coalesced by reference) and at share time. Idempotent, no callback. The Filnavn column uses the editable display_name (rename-to-redact).

Documents are journalised into a sub-folder per GO/Nova case (named by {source_case_id}) under the GO case's Dokumenter library; files with no source case go to the root. AddToCase creates the folder itself; the large-file chunked path makes a placeholder folder first (oomtm.go).

Input

Field	Modes	Meaning
mode	all	create_case / update_metadata / journalize_email / journalize_ref / journalize_folder / delete_doc / generate_aktliste
kontakt_case_id	all	KontAKT case id (callback target)
go_case_no	most	the GO AKT-… case number
title	create/update	Emne → ows_Title
modtaget	create/update	YYYY-MM-DD HH:MM:SS → ows_Modtaget
caseworker_email	create/update	the caseworker to set as CaseOwner
email_id	journalize_email	KontAKT case_emails row to journalise
source_case_id	journalize_ref / generate_aktliste	the GO/Nova case to act on
go_doc_id	delete_doc	the GO DocId to delete

Large data (e-mail bodies, delivered-files list, aktliste rows) is fetched from KontAKT (…/emails/{email_id}, …/delivery-files[?source_case_id=], …/aktliste?source_case_id=); file bytes come from …/documents/{doc_id}/content.

Output / callbacks

• create → POST /api/v1/cases/{id}/go-journal/created {ok, go_case_no, note}
• update → POST /api/v1/cases/{id}/go-journal/updated {ok, note}
• email → POST /api/v1/cases/{id}/emails/{email_id}/journalized {ok, doc_id}
• journalize_ref / journalize_folder → POST /api/v1/cases/{id}/go-journal/documents-journalized {ok, mappings:[{doc_id, go_doc_id}], doc_count}
• delete_doc → no callback (best-effort; the KontAKT row is already gone)

(or {ok: false, note} on failure)

Required configuration

• Constant GOApiURL — GO base URL (e.g. https://ad.go.aarhuskommune.dk)
• Credential GOAktApiUser — GO NTLM username + password
• Credential KontAKTAPI — username = base URL, password = API key

Dependencies

The shared oomtm library (go, pdf). E-mail/HTML → PDF uses LibreOffice headless, auto-installed on first use by oomtm.pdf.ensure_libreoffice.

Caveats

• CaseOwner is set via a PeoplePicker + ValidateUpdateListItem hack (oomtm.go.set_case_owner). It resolves the case's ModernConfiguration path from the create response's CaseRelativeUrl (which equals ows_CaseUrl), so there's no tenant-specific guess. It's best-effort and never fails the job — if it can't be set, the case still exists and the owner can be set in GO.
• Large files (>10 MB) upload to GO via a chunked upload (startUpload/continueUpload/finishUpload to the case's Dokumenter library, then locate the DocId + set metadata) instead of the AddToCase byte-array.