Backend: two new endpoints /reports/dashboard/yearly-comparison and
/reports/dashboard/top-icd for multi-year KW case counts and ICD
frequency aggregation.
Frontend: grouped bar chart comparing KW across years (2022+) and
a ranked top 10 ICD list with progress bars on the Dashboard.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds report_type support across the full stack:
- Backend: REPORT_TYPES mapping, fallgruppen filter in all 5 sheet
calculations, dynamic Excel columns, report_type DB column with
Alembic migration 007
- Frontend: report type dropdown in generation form, type column in
reports table, dynamic fallgruppen in ReportViewer
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Previously, generate_full_report() ignored the kw parameter for data
filtering — it was only stored as metadata. This caused all reports to
contain data up to the latest available KW, making historical reports
(e.g., for KW 8) identical to the current one.
Now all 5 sheet calculation functions accept an optional max_kw parameter.
When generating a report for a specific KW, only cases with kw <= max_kw
are included. Dashboard and vorjahr callers are unaffected (max_kw=None).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add VERSICHERUNG_FILTER="DAK" to config and apply it to all case
queries: list, detail, pending-icd, pending-coding, coding queue,
dashboard KPIs, all 5 report sheets, and excel sync export.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
.cast(int) uses Python's builtin int, which lacks SQLAlchemy's
_isnull attribute. Replace all occurrences with .cast(Integer).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>