The final slice: land the format-neutral document model with REAL consumers
and unify the Markdown parser — no duplication, byte-identical output.
Neutral model (pkg/docforge/model.go): Document / Block / InlineSpan.
BlockKind values are the stylemap keys. A hyperlink is a span with Link set
+ Children (the label's spans), preserving link boundaries so adjacent
same-URL links stay distinct — byte-exact with the pre-model walker.
Markdown importer (pkg/docforge/markdown): Import(md) → Document. The SINGLE
Markdown parser for docforge — block split, marker detection, inline
bold/italic/link tokenisation, {{placeholder}} pass-through (the b78a984
fix). Relocated out of the docx walker.
docx renderer (pkg/docforge/docx/markdown.go): now RENDERS a Document →
OOXML (RenderDocumentToOOXML); RenderMarkdownToOOXML[WithStyles] = render(
markdown.Import(md)). The shipped submission walker routes through the model,
so there is one parser, not two. The comprehensive byte-exact render tests
(RenderMarkdownToOOXML_*) all PASS unchanged = output identical.
Exporter interface (pkg/docforge/exporter.go, PRD §4 B4): Exporter{Format,
MIMEType, RenderBody(Document)} with the .docx impl (pkg/docforge/docx/
exporter.go). The seam a future PDF/HTML exporter slots into.
Tests: parser tests relocated to the markdown pkg (parseSpans/detectBlockMarker)
+ new importer Document tests + exporter conformance test.
Verification: go build/vet clean; gofmt clean; full NO-DB test suite GREEN
(authoritative — proves no regression); docforge byte-exact render oracle
PASS; composer live test renders through the rewired walker (PASS); bun build
+ bun test 274/274. The shared-DB live run fails ~85 tests across unrelated
services from a harness pq-42P08 $1-type seeding quirk + a stale
deadline_rules test — systemic/environmental (the no-DB run is clean), not
this change.
docforge train complete: 8 slices, the engine extracted + cleaned + a working
author→generate→export loop on uploaded templates, plus the neutral model +
importer + exporter seam for future formats/consumers.
m/paliad#157
A submission draft can now render from an uploaded docforge template
instead of a legacy Gitea base. DB-VERIFIED against TEST_DATABASE_URL (the
head greenlit option C) before commit — not just compiled.
Schema: migration 159 adds submission_drafts.template_version_id (nullable,
FK template_versions ON DELETE SET NULL) — the snapshot pin (PRD A3). A
later template edit creates a new version; the pinned draft keeps rendering
its version.
Draft service: TemplateVersionID on the model + draftColumns + the JOIN
list + DraftPatch (two-level pointer like base_id) + Update SET. Column-sync
verified live (Create_seeds_section_rows + the new pin test both pass).
Export/preview (handlers): a template-version path checked FIRST — load the
carrier via TemplateStore.GetVersion, render via the existing Export/
RenderPreview (the carrier already carries {{slots}}; no Composer/sections
needed). Falls through to base_id / v1 if the pin is missing. Both preview
sites + the view assembly branch on it.
Store: TemplateMeta.VersionID exposes the current version's row id (slice-4
gap — a consumer needs it to pin); populated in List/Get/GetVersion + the
authoring JSON. New GET /api/templates (authenticated, firm-filtered) is the
picker list any lawyer reads; admin authoring endpoints stay gated.
Frontend: the submission editor's base picker now offers uploaded templates
as a 'tpl:<version_id>' optgroup; selecting one PATCHes template_version_id
(clearing base_id) and vice versa — mutually exclusive render paths.
Live test (submission_draft_template_live_test.go, gated): pin round-trips
Update→Get, the uploaded carrier renders ({{firm.name}}→HLC via Export), and
clearing nulls it — all PASS against real Postgres.
Verification: go build/vet/gofmt clean; bun build + bun test 274/274; slice-7
+ slice-4 store + draft/composer live tests PASS against TEST_DATABASE_URL.
Pre-existing env failures (approval/projection seed $1-type quirk,
migration136 stale deadline_rules table) are unrelated — confirmed my branch
touches none of that code.
m/paliad#157
The WYSIWYG authoring surface at /admin/templates (admin-gated page route):
- templates-authoring.tsx — page shell (upload form, template list,
workspace: palette / run-addressable preview / placed slots).
- client/templates-authoring.ts — hydrates it: lists templates, uploads a
.docx (multipart), renders the run-span preview, builds the variable
palette from the Go catalogue (GET /api/docforge/variables), and wires
the select-then-pick gesture: select text within one .docforge-run, click
a palette variable → POST the slot → re-render with the response. Reuses
the docforge-editor lib (escapeHtml, catalogue client). Cross-run
selections rejected with a hint (v1: single-run text slots).
- build.ts emits dist/templates-authoring.html + bundles the client.
- handleTemplatesAuthoringPage serves the shell; GET /admin/templates
registered under adminGate.
- 12 i18n keys (DE+EN) for the page; i18n-keys.ts regenerated (3079).
Verification: go build/vet/test green (13 pkgs); bun run build.ts clean
(i18n scan passes); bun test 274/274; gofmt-clean. The docx surgery + store
+ catalogue are unit/live-tested. VERIFICATION CEILING: the integrated live
flow (upload→render→select→inject→save in a browser) needs the app running
with DATABASE_URL + Supabase auth + Playwright — verified post-merge, not in
this env.
m/paliad#157
Admin-gated authoring API over docforge.TemplateStore + the docx authoring
engine (handlers/templates.go, routes under adminGate):
GET /api/admin/templates — catalog list
POST /api/admin/templates — multipart upload → ImportForAuthoring
(validate + detect slots) → Create v1
GET /api/admin/templates/{id} — authoring view (run-addressable
preview + slots)
POST /api/admin/templates/{id}/slots — InjectSlot at the selection →
AddVersion (re-detect slots from the
new carrier so template_slots mirror
the carrier's actual {{tokens}})
docforge.ErrTemplateNotFound → 404; injection failures (bad selection/key)
→ 400 with the engine's message for the UI to surface. Upload capped at
10 MB. Slot placement creates a version per placement (immutable snapshot);
batching a session into one version on explicit save is a documented
refinement.
Verification: go build/vet clean, handlers test green, gofmt-clean. The docx
surgery + store are unit/live-tested; the integrated HTTP flow is verified
post-merge (needs DATABASE_URL + Supabase auth).
m/paliad#157
The verifiable backend heart of the authoring surface, before the HTTP +
frontend layers.
pkg/docforge/docx/authoring.go:
- ImportForAuthoring(carrier) → AuthoringView{PreviewHTML, Slots}: parses
an uploaded .docx into a run-addressable preview (one
<span class="docforge-run" data-run="N"> per <w:t>, document order)
plus the {{placeholder}} slots already present.
- InjectSlot(carrier, runIndex, selectedText, slotKey) → new carrier:
replaces the selection inside run N with a {{slot_key}} token. Keys on
the selected TEXT (not a byte/UTF-16 offset) so umlauts can't desync the
client selection from the server slice; preview + injection walk runs in
the identical paragraph→<w:t> order so data-run indices line up.
- v1 scope: text slots in body paragraphs; out-of-run / cross-run / not-
found selections return an error the UI turns into a hint.
6 unit tests cover run-addressable preview, slot detection, injection +
round-trip re-import, umlaut/run-targeting, and the error paths (selection
absent, out-of-range run, invalid slot key) — all passing.
Wired PgTemplateStore through the stack (main.go → handlers.Services →
dbServices) so the upcoming authoring endpoints can reach it.
Verification: go build/vet clean, full module test green (13 pkgs), new
files gofmt-clean. The HTTP endpoints + frontend authoring page land next;
their live flow needs the post-merge e2e/manual loop (DB+Supabase).
m/paliad#157
Establish the shared frontend editor package and make the Go resolvers the
single source of truth for variable labels.
Go — catalogue SSOT:
- VariableResolver gains Keys() []VariableKey; ResolverSet gains
Catalogue(). The 7 submission resolvers implement Keys() with the
bilingual labels ported from the TS VARIABLE_LABELS table (incl. the
legacy rule.* aliases). Keys() is entity-independent, so
SubmissionVariableCatalogue() builds a metadata-only ResolverSet.
- GET /api/docforge/variables serves the catalogue (auth-gated, static).
- Tests: docforge ResolverSet (BuildBag merge + Catalogue order) and the
submission catalogue integrity (no dupes, labels present, spot-checks).
Frontend — frontend/src/lib/docforge-editor/ (new shared package):
- dom.ts: escapeHtml + cssEscape (pure), with bun tests. Dedupes the two
identical escapeHtml/escapeHTML copies + the cssEscape copy that lived
in the submission editor.
- catalogue.ts: fetchVariableCatalogue() + labelMap() — the client for
the Go catalogue.
- submission-draft.ts now imports escapeHtml/cssEscape from the lib and
fetches the catalogue on boot into state.varLabels (labelFor reads it,
falling back to the raw key if the fetch fails — graceful degrade). The
hardcoded VARIABLE_LABELS table is removed; VARIABLE_GROUPS stays
(presentation: which keys to show + how to section them, legitimately
frontend).
Scope note: the DOM-coupled editor plumbing (wireDraftVars/focus
preservation/autosave debounce) is extracted in slice 6 alongside its first
reuse — the authoring page — rather than speculatively now (extract with the
consumer; same principle as slices 2-3). Slice 5 lands the pure utilities +
the catalogue, which the slice-6 authoring palette consumes.
Verification: go build/vet/test green (Go files gofmt-clean; handlers.go
pre-existing drift, added region clean); bun run build.ts clean;
bun test 274/274 (incl. 5 new docforge-editor tests).
m/paliad#157
Persistence foundation for authoring (slice 6) + generation-on-templates
(slice 7). docforge owns no tables — it defines the contract; paliad
implements it (litigationplanner pattern).
Migration 158_docforge_templates (additive, generic — NOT submission_*-named
so a second docforge consumer reuses it):
- templates — catalog row; current_version_id pins the live
version (FK added post-create to break the
templates<->versions cycle; ON DELETE SET NULL).
- template_versions — immutable snapshots; carrier .docx in a bytea
column (the TemplateStore bytea backend) + stylemap
jsonb. Versioning = snapshot-at-create (PRD A3).
- template_slots — variable slots per version; anchor = sentinel token
locating the slot in the carrier OOXML (PRD §5
lean), slot_key = the bound variable.
RLS mirrors submission_bases: firm-shared SELECT for authenticated,
mutations admin-only + gated in Go (no mutation policy = denied).
docforge root: TemplateStore interface + neutral types (TemplateMeta,
Template, TemplateSlot, *Input, TemplateFilter) + ErrTemplateNotFound.
CarrierBytes is format-opaque []byte so the root never imports the docx
adapter; the exporter wraps (CarrierBytes, Stylemap) into a docx.Carrier.
paliad: PgTemplateStore (sqlx, follows the submission_base_service pattern):
List / Get (current version) / GetVersion (pinned snapshot) / Create
(version 1 + pin) / AddVersion (next version + re-pin), all transactional.
Gated live round-trip test (TEST_DATABASE_URL) covers carrier+stylemap+slot
round-trip and the version bump. No handler wires this yet (PRD: no UI in
slice 4).
Verification: go build ./... clean, go vet clean, gofmt clean, full module
test green, migration NoDuplicateSlot structural test green.
m/paliad#157
Move the variable-bag contract (PlaceholderMap, MissingPlaceholderFn,
DefaultMissingMarker) up to the pkg/docforge root (placeholder.go) — it is
format-neutral, consumed by the resolver layer and any future exporter.
The {{key}} substitution grammar (placeholderRegex, PUA preview sentinels,
replacePlaceholders) stays in pkg/docforge/docx: it is the .docx renderer's
own machinery, not a root concern.
New at the root (vars.go):
- VariableResolver{Namespace() string; Populate(bag PlaceholderMap)} —
a PUSH interface, deliberately not pull Resolve(key): some namespaces
emit a data-dependent key set (parties.claimant.0.name, .1.name, … one
per party) that a fixed key-by-key pull can't enumerate.
- ResolverSet + BuildBag() — composes resolvers into one bag, replacing
the hard-coded addFooVars-then-addBarVars sequencing in Build.
paliad side (submission_vars_resolvers.go): seven resolver types wrap the
UNCHANGED addXxxVars push-builders (firm/today/user/procedural_event/
project/parties/deadline), each capturing the entity it needs. The builder
bodies are byte-for-byte untouched, so the bag is identical by
construction; SubmissionVarsService.Build now wires the applicable
resolvers and calls ResolverSet.BuildBag(). Resolvers stay in paliad
because they read paliad's domain model; a second docforge consumer plugs
its own resolvers into a ResolverSet the same way.
Keys()/Catalogue() (the static key list that will data-drive the authoring
palette + kill the hardcoded VARIABLE_GROUPS in submission-draft.ts) is
deferred to the UI slice that consumes it, sourced from the frontend's
existing labels — building it now, ahead of its consumer, would be
speculative (PRD §4 B3 principle).
Verification: go build ./... clean, go vet clean, full module test green.
Alias-parity (procedural_event ≡ rule) and party-form tests pass unchanged
= bag byte-identical.
m/paliad#157
Move the full compose pipeline (anchor-pair splicing, append-before-sectPr,
hyperlink-rels patching, zip split/repack, final placeholder pass) into
pkg/docforge/docx/compose.go, decoupled from paliad's DB row types. The
engine now owns the entire .docx assembly.
New neutral types in docx:
- Carrier{Bytes, Stylemap} — the opaque base .docx, preserved
byte-for-byte outside the spliced regions (the lossless docforge
carrier for .docx).
- Section{Key, OrderIndex, Included, ContentMDDE, ContentMDEN} — the
format-neutral content input.
- Composer / NewComposer / ComposeOptions on those neutral types.
internal/services keeps SubmissionComposer + ComposeOptions as a thin
mapping wrapper (SubmissionSection -> docx.Section, Base.SectionSpec.Stylemap
+ BaseBytes -> docx.Carrier). handlers + the comprehensive compose_test are
unchanged; the test drives the wrapper end-to-end and its byte-exact OOXML
assertions pass = behaviour preserved.
Retired the slice-1 docx.XMLAttrEscape wrapper + its services forwarder:
compose now calls the local xmlAttrEscape inside the docx package.
Sequencing note: the paragraph-level neutral model (Document/Block/Slot the
PRD §3.2 sketches) is deferred to slice 6, where the authoring importer +
format exporters consume it. Building it now, ahead of any consumer, would
be speculative and risk the byte-identical guarantee for no gain (PRD §4 B3
principle). Carrier is the part of the model that earns its keep this cycle.
Verification: go build ./... clean, go vet clean, full module test green.
m/paliad#157
Relocate the in-house OOXML machinery out of internal/services into the
first docforge adapter, with zero behaviour change:
submission_merge.go -> pkg/docforge/docx/merge.go (placeholder
substitution renderer + preview-HTML emitter)
submission_md.go -> pkg/docforge/docx/markdown.go (Markdown->OOXML
walker incl. the b78a984 underscore-fix)
submission_render.go -> pkg/docforge/docx/dotm.go (.dotm->.docx)
+ their _test.go files (git-tracked renames, 84-99% identical)
internal/services keeps thin type-alias + forwarder shims
(docforge_shims.go) so every caller in services/handlers/main compiles
and behaves identically: PlaceholderMap, MissingPlaceholderFn,
SubmissionRenderer, HyperlinkAllocator (aliases); NewSubmissionRenderer,
DefaultMissingMarker, RenderMarkdownToOOXML[WithStyles], ConvertDotmToDocx,
SanitiseSubmissionFileName (forwarders). docx.XMLAttrEscape is exported so
submission_compose.go's hyperlink-rels inserts reuse the walker's escaping.
Three mis-filed pretty-printer tests (legalSourcePretty, ourSideDE/EN,
patentNumberUPC) that exercise the vars layer move back to
internal/services/submission_vars_pretty_test.go.
Placeholder grammar + PlaceholderMap stay co-located with the renderer in
docx for now; slice 3 hoists the format-neutral grammar to the docforge
root with the VariableResolver interface.
Verification: go build ./... clean, go vet clean, full module test green
(the byte-exact OOXML golden tests in merge/compose/render pass unchanged
= behaviour preserved). gofmt drift on the moved files is pre-existing
(72/169 services files already drift; no gofmt gate).
m/paliad#157
Extract the submission generator into pkg/docforge: neutral document model
+ opaque carrier (lossless .docx), VariableResolver interface per namespace,
pluggable importer/exporter (.docx first), WYSIWYG authoring page, generic
editor UI package. 8-slice train, extract-in-place migration that protects
the b78a984 underscore fix, the placeholderRegex + data-var contracts, and
the building-block/section model.
Includes all 13 of m's decisions (5 prose-grill metaphor + 8 structured).
upc-kommentar deferred as a live consumer (it is Bun/SvelteKit/TS, zero Go);
abstractions sized for a later HTTP veneer.
m/paliad#157
m had a one-off /tmp/paliad-deadline-export.py (work/head delegation
#2572) that dumped every published sequencing_rules row. Output
showed 37 entries on upc.inf.cfi including optional rules
(Lodging of translations, Review of CMO, ...) which fights the
engine's IncludeOptional=false default and m's "naked proceeding
with options but not always displayed" mental model.
Move to exports/gen-deadline-list.py as the canonical re-runnable
script and add a SQL-level priority filter that matches the
engine. Default suppresses priority='optional'; --include-optional
opts back in for an exhaustive catalog dump.
- DSN overridable via PALIAD_DEADLINE_EXPORT_DSN env var.
- argparse-driven: --include-optional / -o OUT / --generated-for LABEL.
- Header explains the mode so the PA reader knows what's suppressed.
- Regenerated exports/upc-deadlines-2026-05-28.md: now 178 rules across
25 proceedings (vs the unfiltered run). upc.inf.cfi section drops
from ~37 to 28 mandatory + conditional rules - the optional ones
are gone; trigger_event_id mandatory rules stay in the catalog
(they're a real PA-knowable surface; runtime anchor state is what
decides whether they project into a timeline, separate concern).
Run:
uv run exports/gen-deadline-list.py [--include-optional]
(m/paliad#153)
The Litigation Builder triplet renders /api/tools/fristenrechner output
verbatim and never applied the pre-existing filterByDetailMode pass that
the legacy /tools/verfahrensablauf page uses. With the engine fix
(3c840c0 — pkg/litigationplanner default IncludeOptional=false + trigger
event semantic anchoring) already in main, optional rules are dropped
server-side but rules with an unsatisfied trigger_event_id surface as
IsConditional. Without filterByDetailMode those still rendered as
"abhängig von ..." cards on the triplet, polluting m's "naked
proceeding with options but not always displayed" mental model.
upc.inf.cfi went from 7 mandatory backbone events to 29 visible cards
(22 conditional noise — Lodging of translations, Mängelbeseitigung,
Antrag auf Verweisung, Wiedereinsetzung, ...). Live BEFORE/AFTER
captured in exports/screenshots/.
Fix layers:
- Go handler (internal/handlers/fristenrechner.go): accept
includeOptional + triggerEventAnchors from request body and
forward to services.CalcOptions. Default zero values match the
engine defaults (suppress optionals + no fabricated dates for
trigger_event_id rules), so the wire is unchanged when callers
don't set them.
- TS calc surface (frontend/src/client/views/verfahrensablauf-core.ts):
add the same two fields to CalcParams + forward in the fetch body;
surface rulesAwaitingAnchor on DeadlineResponse mirroring
Timeline.RulesAwaitingAnchor.
- Builder triplet (frontend/src/client/builder.ts hydrateTriplet):
apply filterByDetailMode(detailgrad) before renderColumnsBody, with
detailgrad sourced from the proceeding row. "selected" (default)
drops conditional + optional rules; "all_options" passes
includeOptional=true so the engine returns the optional rules the
user can opt into.
- Legacy /tools/verfahrensablauf (frontend/src/client/verfahrensablauf.ts):
pass includeOptional based on detailMode + a small hasOptionalOptIn
helper so per-rule rule:<uuid>=true deviations still surface their
optional rule even in "selected" mode (the engine has no rule:<uuid>
awareness; without the opt-in the user's pick would silently no-op).
Tests:
- frontend/src/client/views/verfahrensablauf-core.test.ts: pin the
fetch body shape - includeOptional=true and triggerEventAnchors={...}
round-trip through the request; empty/default values are omitted so
the wire stays minimal.
bun build + bun test (269 pass) + go vet + go test
./internal/handlers/... ./pkg/litigationplanner/... all clean.
(m/paliad#153)
PRD §2.3 + §10. Implements the dual-write rule (load-bearing
complexity per PRD §10): project-backed scenarios mirror flag
toggles to paliad.projects.scenario_flags and filed event states
to paliad.deadlines, while kontextfrei scenarios continue writing
only to paliad.scenario_events. Visible affordances: page-header
Akte picker, enabled "Aus Akte" mode tab, Akte banner on the
project-backed canvas, cross-surface scenario-flag-changed
dispatch + listener for live peer-surface coherence.
Backend
- ScenarioBuilderService takes ProjectService + ScenarioFlagsService
deps so dual-write hits live tables.
- CreateScenarioFromProject seeds a scenario from a project: copies
proceeding_type_id + scenario_flags, normalises our_side to the
builder's binary claimant|defendant axis, surfaces existing
rule-bound deadlines as scenario_events (filed when completed,
planned otherwise).
- PatchProceeding on a project-backed top-level triplet dual-writes
scenario_flags to projects.scenario_flags via flagDeltaFromBuilder.
- PatchEvent transitioning to state='filed' on a project-backed
scenario upserts paliad.deadlines (status='completed', completed_
at, source='rule') inside the same tx as the scenario_events
UPDATE — canvas and project surfaces never diverge mid-flight.
- POST /api/builder/scenarios/from-project handler wires the entry
point.
Frontend
- builder-akte.ts: project list fetch + dropdown render, Akte
banner, createScenarioFromProject POST helper.
- builder.ts: mode branching — picking an Akte (search hit or
page-header pick) creates a project-backed scenario and loads it;
loaded scenarios reflect their origin_project_id on the picker +
banner; flag toggles on Akte-backed top-level triplets dispatch
scenario-flag-changed so the Verfahrensablauf strip / project
surfaces refresh; the builder listens to inbound scenario-flag-
changed and refetches its scenario when the changed project
matches origin_project_id.
- procedures.tsx: enable the previously-disabled Aus Akte tab.
- i18n + CSS: builder.akte.banner.prefix key (DE+EN); lime-tinted
banner styling.
Tests
- TestScenarioBuilderAkteDualWrite (live DB) pins the dual-write
contract: Akte flag toggle → projects.scenario_flags updated,
Akte filed event → deadlines row inserted; kontextfrei flag
toggle leaves projects.scenario_flags untouched, kontextfrei
filed event leaves deadlines untouched.
- Existing TestScenarioBuilderService passes against the new
signature (nil deps short-circuit dual-write paths).
Verification: go test ./... + go vet ./... + bun run build all
clean. Playwright smoke against the static dist build confirms
the Akte tab + picker render correctly, fetchAkteProjects fires
on mount, and the scenario-flag-changed CustomEvent dispatches +
receives without runtime errors.
t-paliad-347
PRD §2.2 + §3.1: the page-header search box drives a typed dropdown
returning grouped event / scenario / project hits, and the "Ereignis"
entry mode is enabled. Picking an event creates a scratch scenario
with one triplet anchored on that event's proceeding type, with the
event card auto-anchored (lime band + "━━━━ DU BIST HIER ━━━━" divider
above the next-coming events).
Backend: new GET /api/builder/search reuses
DeadlineSearchService.SearchEvents for the events corpus (UPC v1),
filters owned scenarios by ILIKE on name, and reuses ProjectService.List
for the Akten group (team-RLS via visibilityPredicate). Each group is
capped independently (default 8 events / 5 scenarios / 5 projects, max
30). Missing services degrade gracefully — empty group, not 503.
Frontend: builder-search.ts owns the dropdown (debounced 180ms,
arrow-key navigation, Enter to pick, abort on next query). builder.ts
gains mode state ("cold" | "event" | "akte"), wires the mode bar +
search input, and runs applyAnchorHighlight after triplet hydration —
the helper finds the .fr-col-item with the picked rule_id, adds the
.builder-anchor-card lime band, and inserts a full-width
.builder-anchor-divider after the anchor's row in the columns grid
via JS row-index math (the grid is row-major with 3 header cells
+ 3-cells-per-row body).
Filter pill reset: setMode() clears the search input and closes the
dropdown when switching entry modes. Forum/proc/party/kind chips are
not yet rendered separately (they live in the search dropdown today);
the reset hook attaches there too when those land in a follow-up.
Verification:
- bun build (frontend bundles + i18n scan clean)
- go vet ./... + go test ./... (all packages pass)
- Playwright: mode switch focuses search, debounced fetch fires,
typed result groups render with N · M · K pluralization, event
pick creates scratch scenario + adds proceeding, anchor card
+ DU BIST HIER divider render in the columns grid (screenshots
confirmed visually)
The Litigation Builder's "+ Verfahren hinzufügen" silently failed in
prod after t-paliad-343 B2 shipped — clicking a Verfahren chip in the
picker did nothing, no visible error.
Root cause: the wire shape FristenrechnerType (the response of
/api/tools/proceeding-types) carried code+name+nameEN+group but not
id. Builder.ts mountAddProceedingPicker's callback POSTed
`{proceeding_type_id: meta.id}` to
/api/builder/scenarios/{id}/proceedings — meta.id was undefined,
JSON.stringify dropped the key, the server returned 400 ("invalid
input: proceeding_type_id is required"), and fetchJSON swallowed the
error to console. The user saw "nothing happens".
Fix:
- Add `ID int json:"id"` to lp.FristenrechnerType.
- SELECT id in FristenrechnerService.ListProceedings + Scan into the
new field.
- Defensive guard in builder.ts openAddProceedingPicker — refuse to
POST without a positive integer id and log a clear error, so a
future wire-shape regression cannot recreate the silent-fail.
- Regression test in pkg/litigationplanner/types_wire_test.go pins the
contract (id present in JSON, round-trips as integer).
Side-benefit: fristenrechner-wizard.ts:599-628 documented this exact
gap as a known limitation ("S5/follow-up can extend the wire shape to
include id"). That workaround can now be retired in a follow-up.
Refs m/paliad#153 (Litigation Builder)
GetScenarioDeep returned nil slices for proceedings/events/shares when
a scenario had zero rows, which Go's encoding/json serialises as `null`
rather than `[]`. The builder's renderCanvas then unconditionally calls
`state.active.proceedings.filter(...)` on a null and dies with
`procedures.js:101 TypeError: Cannot read properties of null
(reading 'filter')` — every cold-open scenario crashed the page before
the empty canvas could render.
Backend (root cause): initialise Proceedings / Events / Shares to empty
slices in BuilderScenarioDeep before SelectContext, so the wire shape
is always arrays. Existing rows still load via SelectContext, which
truncates the placeholder and refills from the DB.
Frontend (defence in depth): on loadScenario(), normalise each of the
three arrays to `[]` if the server response is not an array. Catches a
future regression (or an older deployed build) without re-introducing
the same crash class.
bun build clean, go vet + go test ./... green.
Two SELECTs still referenced paliad.deadlines.rule_id after mig 140
(Slice B.4) dropped that column in favour of sequencing_rule_id:
- internal/services/deadline_service.go:268 — DeadlineService.
ListVisibleForUser. Powers /api/events?type=deadline (dashboard
deadline rail, /deadlines page, every status bucket). Threw
`pq: column f.rule_id does not exist` on every request → 500
for any authenticated user hitting the dashboard.
- internal/services/projection_service.go:1250 — collectActualsForOverrides.
Same column on `paliad.deadlines d`. Logged once per projection
pass (`ERROR service: projection: deadlines: ...`); aliased the
rename to `rule_id` so the receiving struct tag still scans.
Live container logs confirmed the failure mode — a 60-row burst of
`pq: column f.rule_id does not exist at position 3:36 (42703)` starting
the minute the post-B0 container came up (mig 140 had applied to the
DB but the SELECT still used the dropped name). EXPLAIN against the
live schema after the edit plans cleanly; the LEFT JOIN to
paliad.deadline_rules_unified on sequencing_rule_id was already correct
(only the SELECT projection was stale).
Root cause: mig 140 commit (1129bab) renamed the JOIN to
`f.sequencing_rule_id` but left the SELECT clause on the older name.
The model tag is already `db:"sequencing_rule_id" json:"rule_id"`, so
the wire shape is unchanged — only the column reference flips.
bun build clean, go vet ./... clean, go test ./... green.
Builds on B1 (commit 6c1d8cc). After this slice a user can compose a
multi-proceeding scenario kontextfrei: stack proceedings, flip
perspective per-triplet, toggle scenario flags, auto-spawn child
proceedings on flag transitions, and mark individual event cards as
planned / filed / skipped — all auto-saved to paliad.scenario_*.
PRD §7.1 B2 acceptance shipped:
- Multi-triplet stack: top-level proceedings sorted by ordinal,
child proceedings nested inline with a left lime border.
- Per-triplet controls bar: perspective radio (none / claimant /
defendant), Detailgrad pill (selected / all options), Entfernen
action. Each control PATCHes the proceeding row and re-renders the
affected triplet.
- Per-triplet flag strip: every paliad.scenario_flag_catalog row
rendered as a checkbox, bound to scenario_proceedings.scenario_flags.
Active flags also surface as chips in the triplet header for quick
legibility.
- Spawn nesting: when `with_ccr` flips ON on upc.inf.cfi the builder
auto-POSTs an upc.ccr.cfi child proceeding linked via
parent_scenario_proceeding_id; flip OFF deletes the child (events
cascade via the schema). The SPAWN_MAP table is data-driven so
future spawn flags slot in.
- 3-state event cards (planned / filed / skipped):
overlayEventStates walks the rendered .fr-col-item nodes (the
data-rule-id hook added to verfahrensablauf-core in this slice)
and stamps each card with data-builder-state + per-state action
buttons (File / Skip / Reset to planned). Filed cards prompt for
a date; skipped cards prompt for an optional reason. POSTs or
PATCHes paliad.scenario_events keyed by sequencing_rule_id.
- Per-card optional horizon chip: stores horizon_optional on the
scenario_event row, increment / decrement chip on every card.
The full surface awaits a calc-engine "optionals available"
counter (PRD §3.4 follow-up); the persistence layer + UX hook are
in place so the wiring lands without another schema touch.
- Page-header Stichtag drives default dates for every triplet (the
triplet's per-stichtag override path is wired but the per-triplet
Stichtag input is a B3+ affordance).
verfahrensablauf-core.renderColumnsBody now stamps data-rule-id (and
data-submission-code as a future hook) on every .fr-col-item root —
non-breaking enhancement; the legacy /tools/* pages don't read either
attribute. Verified by re-running the existing 57-test suite.
Backend: one new read-only endpoint
GET /api/builder/scenario-flag-catalog passes through
ScenarioFlagsService.ListCatalog so the builder doesn't need a
per-project round-trip to render flag toggles.
bun run build clean (3050 i18n keys), go vet ./... clean, go test ./...
clean, frontend bun test (verfahrensablauf-core suite) 57 / 57 pass.
Replaces cronus's U0-U4 catalog at /tools/procedures with a
persistence-backed builder shell on top of B0's API surface
(/api/builder/scenarios/*, t-paliad-340).
PRD §7.1 B1 acceptance shipped:
- Page header: scenario picker, name action, Akte picker stub,
Stichtag input, search input, save status indicator.
- Entry-mode radio (cold-open active; event-triggered + akte
rendered disabled for B3/B4 layout stability).
- Empty canvas with "Neues Szenario starten" CTA and a 5-most-recent
list rendered when the user has saved scenarios.
- Side panel "Meine Szenarien" with the Aktiv bucket; clicking an
item loads the scenario into the canvas.
- Add-proceeding inline picker (Forum chip row → Verfahren chip row
→ Hinzufügen). UPC v1; other forums chipped but disabled.
- First proceeding triplet renders end-to-end via
verfahrensablauf-core.calculateDeadlines + renderColumnsBody (the
existing 3-column proaktiv|court|reaktiv body, read-only in B1).
- Auto-save with 500ms debounce on name + stichtag patches; save
status flips idle → saving → saved/error in the page header.
New client modules under frontend/src/client/:
- builder.ts — orchestrator (URL state, fetch, auto-save loop,
canvas render, scenario-list re-paint).
- builder-picker.ts — inline Forum/Verfahren popover for the
add-proceeding affordance.
- builder-triplet.ts — single-triplet header + body wrapper.
procedures.tsx rewritten as the shell scaffolding (sidebar, page
header, mode radio, two-column body); procedures.ts now boots the
builder instead of toggling the 4-tab catalog.
Legacy U0-U4 modules (verfahrensablauf.ts, verfahrensablauf-state.ts,
VerfahrensablaufBody.tsx, procedures' tab toggle in client/procedures.ts,
fristenrechner-* mounts) are no longer reachable from /tools/procedures
but kept in the tree for the B6 cleanup sweep per PRD §7.4.
i18n.ts grew 60 keys × 2 langs under builder.*. global.css grew a
self-contained .builder-* block at the file tail.
bun run build, go vet ./..., and go test ./... all green.
Two paired engine semantics fixes:
1. trigger_event_id is now the authoritative semantic anchor. When a
rule carries trigger_event_id, the engine no longer falls back to
the proceeding's trigger date — it resolves the anchor via
CalcOptions.TriggerEventAnchors keyed by paliad.trigger_events.code.
Missing anchor renders the rule as IsConditional (empty date) and
propagates via courtSet so descendants also surface as conditional.
Fixes the RoP.109.5 bug where the engine fabricated a date 2 weeks
before the user's SoC instead of waiting for the oral_hearing date.
2. priority='optional' rules are suppressed from the default
Calculate output. Callers (paliad /tools/procedures,
youpc.org/deadlines) opt in via CalcOptions.IncludeOptional=true to
restore the legacy "show optional applications" behaviour. The
suppression cascades through skippedIDs so child rules drop too.
Wire shape additions:
- CalcOptions.IncludeOptional bool
- CalcOptions.TriggerEventAnchors map[string]string
- Timeline.RulesAwaitingAnchor int (count of suppressed-by-missing-
anchor rules, for caller telemetry / "N rules need an anchor" UX)
Existing before-court-set-anchor tests opt in to IncludeOptional=true
to preserve their non-optional-related test intent.
Refs: youpcorg/head delegations #2568 + #2570, m/paliad#153 (Litigation
Builder PRD path).
The Markdown inline scanner (parseInlineSpans) treats _ and * as
italic delimiters. A placeholder like {{project.case_number}} fed
through the scanner had its underscores consumed as italic markers,
leaving {{project.casenumber}} in the composed OOXML. The v1
placeholder pass then looked up the wrong key, surfacing
[KEIN WERT: project.casenumber] in the preview. The form ↔ preview
highlighting also stopped working because data-var attributes
mismatched between the input (snake_case) and the rendered span
(stripped).
parseInlineSpans now detects {{ at the cursor and skips ahead to
the matching }}, copying the entire placeholder verbatim into the
current text run. Unmatched {{ falls through to the existing
character handling so legal prose with stray braces still renders.
Tests: regression test for underscored keys (single + multiple +
mixed-with-italics), direct guard on parseInlineSpans, and an
italic-around-placeholder structural test.
t-paliad-340 — B0 of edison's 7-slice train (PRD §7.1). DB-only: schema
+ RLS land, dev-only test route exercises the surface, no user-facing
change. B1 wires the actual builder UI on top.
Migration 157 (additive on the legacy mig-145 scenarios table — 0 rows
in prod, safe to relax):
- paliad.scenarios gets owner_id / status / origin_project_id /
promoted_project_id / stichtag / notes. spec drops NOT NULL and the
scenarios_unique_per_scope constraint drops (the builder allows
multiple scratch + Unbenanntes Szenario rows per user).
- New tables: scenario_proceedings, scenario_events, scenario_shares.
- paliad.projects.origin_scenario_id for the promote-to-project audit
trail (the FK lands now; the wizard ships in B5).
- paliad.can_see_scenario(uuid) STABLE SECURITY DEFINER helper covering
owner / share / global_admin / two legacy paths.
- Replacement RLS on scenarios + RLS on the three new tables; legacy
service + handlers stay live and unchanged.
PRD §5.1 deviations called out in the migration header:
- proceeding_type_id is integer (live schema), not uuid (PRD draft).
- FK target is paliad.users, matching the rest of paliad's schema.
Go surface:
- ScenarioBuilderService — list/create/get-deep/patch scenarios,
add/patch/delete proceedings, add/patch/delete events,
add/delete shares. Writes wrap in transactions with set_config(
paliad.audit_reason, ..., true) per event_choice_service.go pattern.
- /api/builder/scenarios/* — handlers register under a builder/
prefix so the legacy /api/scenarios surface still works.
- /dev/scenario-builder — single-page HTML form gated to
PaliadinOwnerEmail, exercises the B0 surface without Postman.
- Live-DB integration test (TEST_DATABASE_URL gated) covers
create + list + deep-get + share + visibility negatives + patch.
Audit-first: every DDL block ran clean via BEGIN/ROLLBACK against
the live DB before commit; end-to-end sanity (insert chain + CHECK
constraints + CASCADE-on-delete) verified via the Supabase MCP.
bun build clean. go vet + go test -short ./... green.
PRD for the columnar litigation planner replacing today's 4-tab catalog
at /tools/procedures with a Litigation Builder backed by a new Scenario
DB. Captures 20 chip-picker decisions (5 batches via AskUserQuestion)
covering: unified-builder shape with 3 entry modes (cold-open /
event-triggered / Akte), separate paliad.scenarios table with
multi-proceeding constellations, auto-save + named-list, per-proceeding
flags + perspective + Detailgrad, 3-state event cards
(planned/filed/skipped), per-event-card optional horizon, vertical
stacked column-triplets with inline spawn children, universal search
(events + scenarios + Akten), 3-step promote-to-project wizard,
read-only team sharing, desktop v1 + mobile basic-read.
Includes data model deltas (4 new tables + 1 column on
paliad.projects), 6-slice migration plan from the current live U0-U4
catalog, and coder hand-off notes. Cross-proceeding peer triggers and
DE/EPA/DPMA full expansion deferred to v1.1.
