The per-Property fact layer deferred by ADR-0004: one row per
(property, building_part, override_component) holding the resolved
landlord-override enum as a denormalised text snapshot, plus the raw
spreadsheet description it resolved from.
Schema only — no writer yet. The bulk_upload_finaliser application will
populate it (recalculate-on-rerun via upsert on the unique key). Design
and rationale (snapshot-not-FK, drop source, recalculate semantics) in
docs/design/bulk-upload-finaliser.md.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
For uploads with classifier columns, surface the classifier's
description→enum assumptions on the awaiting_review screen so the user can
correct any (written source='user') and acknowledge before Finalise.
Previously only the multi-entry order step existed, so non-multi-entry
uploads got no classification review at all and the assumptions were
applied silently.
- detectMultiEntry: capture a sample whenever classifier columns are
mapped (largest-count row if multi-entry, else first classified row);
the sample now carries all classifier columns. "sample != null" means
"there is something to verify"; largestCount >= 2 stays the multi-entry
signal.
- setVerifyAck + verify-classification PATCH route + useConfirmVerification.
- VerifyClassificationPanel (Step 1); MultiEntryOrderingPanel slimmed to
order-only with read-only classification annotation; canFinalize gated
on both steps where each applies.
- Unit tests for detectMultiEntry + ordering helpers.
The verify_ack column + 0219 migration landed separately via #303.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The "Verify classification" acknowledgement flag for ADR-0004 Step 1.
Gates Finalise whenever an upload has classifier columns, independent of
multi-entry, so it lives in its own column rather than on
multiEntryOrdering.
Plain additive column (boolean NOT NULL DEFAULT false), no data backfill,
so it applies cleanly with either `drizzle-kit migrate` or `push`. The
feature that reads/writes it lands separately on
feature/frontend_landlord_overrides.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
#298: read the classifier's resolved enums for the multi-entry sample's
entries (joined from landlord_*_overrides by normalized description) and show
each beside the raw text, or "not classified" when absent.
#299: make each classification editable — a dropdown of the category's valid
enum values whose selection upserts the override by (portfolio, description)
with source='user', so the classifier never re-clobbers it. UI notes the
portfolio-wide blast radius. Verification ack is folded into the existing
order-confirm (no separate flag/migration); editing is optional review.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Additive nullable jsonb column for the user-confirmed building-part ordering
(ADR-0004, issue #297), generated off main. No data migration. The jsonb shape
type is co-located with the column so the schema is self-contained.
Split out as its own migration PR so the DB change can be approved and deployed
independently of the feature's app code.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds the multiEntryOrdering jsonb column + interactive order picker: the
largest-count multi-entry sample is shown with a building-part dropdown per
file position (one Main building + Extensions), validated as a permutation.
A PATCH route persists { count: permutation } + confirmed, and Finalise is
disabled until the ordering is confirmed when the upload is multi-entry.
Migration for the new column is intentionally not included here — generated
after origin/main is merged (its multi_entry_summary migration lands first).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Additive nullable jsonb column for multi-entry building-part detection
(ADR-0004), generated off main. No data migration. The jsonb shape type is
co-located with the column so the schema is self-contained.
This is the only migration the multi-entry feature needs that isn't already
on main (the column_mapping inversion, sub_task.service, and the
landlord_overrides tables are all already merged). Split out so the DB change
can be approved and deployed independently of the feature's app code.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Removed properties have no coordinator or confirmed survey date by definition,
so those columns were always blank. Outcome notes are the field that actually
explains why the property dropped out of bookings.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Previously, any HubSpot outcome would override bookingStatus="Do Not Book"
and keep the property in the normal pipeline. That was too permissive —
outcomes like "Tenant Refusal" or "Not Viable" combined with Do Not Book
should classify the property as Removed from Bookings, not lurk in Queries
or the survey-issues bucket.
Now only completed survey outcomes (Surveyed, Surveyed - Pending Upload,
EPC Completed) override Do Not Book. Any other outcome + Do Not Book
falls through to Removed from Bookings, surfaces in the Halted or Removed
panel, and gets the matching stage badge in the Properties tab. The
redundant "Removed from Bookings" chip in the drill-down table is gone
since the stage classification now carries that signal cleanly.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Splits the DB migration artifacts off the frontend branch so they can
land independently:
- 0215_invert_column_mapping: one-shot data migration inverting
bulk_address_uploads.column_mapping from header->field to field->header
(drops 'skip' entries). One-shot — see file header and ADR-0003.
- 0216_add_subtask_service: adds sub_task.service (nullable text) to tag
which pipeline a subtask belongs to (address2uprn vs
landlord_description_overrides).
Includes the subtask.ts schema source for 0216 so drizzle's snapshot
stays consistent. No app code depends on these yet.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Both surface in the column-toggle dropdown and the CSV export, hidden by
default like the other optional columns. surveyedDate sits next to
designDate (chronological survey -> design order); epcPrn sits after the
lodgement date since the PRN is the output of lodgement.
epcPrn was already in the DB schema but absent from the HubspotDeal type
and mapper, so the plumbing is added alongside.
Extracts the CSV-formatting logic out of PropertyTable into a pure
propertyCsv module with tests, locking in the export contract (header
order, en-GB date formatting, null cells empty) so future column drift
is caught.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Softer, less technical language for landlord-facing copy. The subtitle
now spells out the two states (halted before a survey, or removed from
the project entirely) so the relationship between the section header
and the two cards is explicit.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Properties that are intentionally not progressing (bookingStatus = "Do Not
Book" with no outcome, or batch = "Removed from Program") were landing in
the "Queries" bucket, inflating it with non-actionable rows. Two new
terminal DisplayStage values now classify these explicitly, with
precedence Removed from Program > Removed from Bookings > Queries. Both
are excluded from pipeline funnel and stage-progress denominators
(sibling to Queries) and surface as their own cards under "Excluded from
Pipeline" on the analytics tab. Drill-down rows in Survey Issues get
slate chips when a deal carries either flag, preserving outcome history
for properties surveyed before being de-scoped.
Also removes the unused SurveyedResultsPieChart chain (component,
computeOutcomeSlices, OutcomeSlice, outcomePieSlices field).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Coordinators record non-damp/mould observations (e.g. wasp nests) in the
same comments field, but the section was framed entirely around damp and
mould. Reframe the panel copy and table titles around "condition issues",
keep "Damp, Mould" up front so the Awaab's Law urgency still leads, and
mark the damp/mould rows specifically with a red badge column so they
don't blend into the broader list.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The risk drill-down's coordinator-stage table showed "Yes" for every row,
which carried no useful signal. It also missed properties where the
coordinator wrote a comment without setting the growth flag.
Include rows where dampmould_growth is "yes" (case-insensitive) OR the
comment is populated, and render the comment in the cell — truncated
with a popover for the full text, or a "no note from coordinator"
placeholder when the row is here only because the flag was ticked.
Also drop the typo in the schema property name
(damnp -> damp); SQL column unchanged.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
