timothy
fba5233caf
168 records -> docs/decisions/records/<area>/<topic>.md (163 active, 23 dirs) and docs/decisions/archive/<area>/<topic>.md (5 archived). The filename IS the key, so one-active-record-per-key becomes a filesystem property rather than a validator check, and supersession becomes a `git mv`. WHY: the monolith was a concurrency problem before an aesthetic one. A 3,900-line append target made parallel sessions collide -- PR #605 and PR #614 both hit append-vs-append conflicts during routine rebases, and hand-resolving those inside the corpus is exactly the operation the rationale-rewrite guard exists to police. HOW IT IS VERIFIED: a ~170-file diff cannot be meaningfully read, so correctness does not rest on reading it. The parser was taught BOTH formats first, so the body-diff guard parses the old form at the merge-base and the new form at head -- the migration validates itself, no bypass. The proof is a field-level equivalence harness: 168 records before and after, zero lost, zero gained, zero field mismatches, zero rationale bodies differing. Reviewers should scrutinise the harness; it is the actual evidence. What measuring caught that reading would not have: - ~500 lines sit OUTSIDE any record -- decisions.md's lifecycle schema and each topic file's preamble, mostly the only copy. Source files are kept and stripped, never deleted. They also cannot be filed per-area: topic files hold several areas and 4 of 23 areas span several files. - Archive discovery was a non-recursive glob; after the split it found ZERO archived records, surfacing as four bogus "supersedes points to unknown key" errors rather than an obvious failure. - ~32 live docs point into the corpus BY DATE, which the split dangles. Each stripped file now ends with a generated "Records formerly in this file" index, which also rescues the identical breadcrumbs in old issue comments. - decisions.md's "In this file:" list was 97 same-file anchor bullets that the split makes WRONG, not merely stale. Dropped; the generated index replaces them with links that resolve. The equivalence harness now runs against a checked-in FIXTURE, not the live corpus. The earlier version migrated the real tree, which made it a one-shot: the moment the migration landed there was nothing left to move and the tests failed for reasons unrelated to the code. A fixture keeps them testing the SCRIPT rather than the repo's current state. Keys preserved verbatim, warts included: `sched` (12) and `scheduling` (1) remain two directories for one concept. Renaming a key is not a move -- it changes identity, breaks the equivalence proof, and invalidates MemPalace's per-key drawers. Taxonomy normalisation is separate work. refs #610
3.5 KiB
key, title, status, since, supersedes, superseded-by, rule, signals, mechanics
| key | title | status | since | supersedes | superseded-by | rule | signals | mechanics |
|---|---|---|---|---|---|---|---|---|
| spa.smartcollection-rule-builder | 2026-07-18 — SmartCollection rule builder: compile-only closed subset, no stored AST, one-level nesting (#176) | superseded | 2026-07-18 | none | spa.rulebuilder-nesting@2026-07-25 | (superseded) The SmartCollection visual rule builder compiles to/from a closed subset of the Lucene grammar over the existing stored query string — no new AST, one level of group nesting. | SmartCollection, rule builder, Lucene compile/parse · paths: `web/src/builder/rules/`, `compile.ts`, `parse.ts`, `roundtrip.test.ts` · issues: #176, #69 | superseded by `spa.rulebuilder-nesting` (ersatztv#436) — the compile-only closed-subset and field-catalog stances carry forward there; only the one-level nesting cap was reversed. See `docs/decisions.md` → that record; `spa-conventions.md` §12 |
The SmartCollection create/edit dialog gained a visual rule builder (web/src/builder/rules/)
alongside the existing raw-Lucene textarea. The SmartCollection still stores a plain Lucene query
string — no new stored rule AST, no schema change. The builder compiles its in-memory rule tree
into a closed subset of the Lucene grammar (compile.ts) and parses exactly that subset back out
(parse.ts, the exact inverse — returns null, not a best-effort guess, for anything outside the
subset); escaping is total, so any builder-authored query round-trips losslessly, proven by a 500-tree
property test (roundtrip.test.ts, including Lucene special characters). Opening an existing
SmartCollection tries the parse first and falls back to raw-text mode on null (fuzzy queries,
boosts, mixed AND/OR at one nesting level, or nesting deeper than one level).
Why compile-only over persisting an authoritative rule AST: an AST would still need a Lucene→rules parser to open every pre-existing free-text query — including every query the Auto-Tune feature (#69) generates — so the AST would buy almost nothing (it still can't represent arbitrary hand-written Lucene) while costing a dual-provider EF migration and a second source of truth to keep in sync with the Lucene grammar. Compile-only keeps the query string as the single source of truth and treats the builder as a structured editor over it, not a new storage model.
One-level-nesting "Kodi" model. types.ts defines a top Group (match: all|any) over Rules
and/or one level of sub-Groups — enough to express type:movie AND (genre:Horror OR genre:Thriller), which covers the smart-playlist patterns Kodi-style rule builders are known for.
Arbitrary/recursive nesting was scoped out as YAGNI; revisit only if a real query needs it.
Field vocabulary comes from a new catalog endpoint, not a hardcoded list. GET /api/v1/search/fields (read-only, MCP-introspectable; see api-conventions.md) returns the curated,
typed, labeled field set derived from LuceneSearchIndex — name/label/type/group/values — and is the
single source of truth the builder's field pickers (fieldCatalog.ts's useSearchFields hook) and
operator/value-input choices are driven from, so the builder's vocabulary can't drift from what the
index actually supports.
Deferred as separate follow-up issues (explicitly out of scope for #176): facet-value typeahead
for value inputs, relative-date operators, nesting deeper than one level, and inline adoption of
RuleBuilder by ChannelBuilder / Auto-Tune (it was built reusable for exactly that reuse — see
spa-conventions.md §12).