Closes REVIEW.md §2. 167 of 371 roadmap words had no lexicon entry, so the word rail silently showed nothing. Now: Roadmap words: 371/371 resolve Spiral targets: 35/35 resolve Deck words: 386/386 resolve and npm run dict:assert makes it a blocking build failure, not a silent empty rail. No runtime morphological analyser ships. lib/conjugation.js surfaceForms() runs at BUILD time over every verb and adjective, so looking up a conjugated form is an index hit on the surface table. The frequency join had to be inverted. A subtitle frequency list holds surface forms; a dictionary holds lemmas whose -다 citation form barely occurs in running text, so joining on headword gives verbs a frequency of roughly zero. Expanding each lemma into the forms it plausibly takes and summing recovers 하다 from 118 to 89,041. Forms claimed by more than one lemma are dropped rather than split, so homographs don't inherit each other's mass. Those expansions score frequency only — the surface table itself stays strictly surfaceForms() output plus the headword. Bands are one per curriculum phase. Phase 1 admits no frequency band at all: during the writing-system phase every word must be phonologically legal for the unit reached, and a rank ceiling would hand the learner a 겹받침 during unit 1.4. shared/phonology.mjs lifts validate.mjs's own feature ladder to enforce that; it agrees with the validator on all 371 words. Sources are chosen automatically — KRDICT when vendored, otherwise the kaikki.org extract. KRDICT's download is a JS form behind anti-bot protection, so it cannot be fetched by CI; the derived band files are committed instead, which the app needs offline regardless. Attribution and the share-alike terms are in NOTICE.md. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
130 lines
5.3 KiB
TypeScript
130 lines
5.3 KiB
TypeScript
/* The vocabulary bands and the phonological ladder — the two things that
|
|
decide what the tutor is allowed to reach for. */
|
|
|
|
import { describe, it, expect } from "vitest";
|
|
import { BANDS, REFERENCE_BAND, bandForPhase, bandForUnit, bandOf, ceilingForBand } from "@shared/bands.mjs";
|
|
import { featureLevel, isReadableAt, phonologyViolations, LADDER_COMPLETE } from "@shared/phonology.mjs";
|
|
import { flatten } from "@lib/gate.js";
|
|
import curriculum from "@data/curriculum.json";
|
|
import type { Curriculum } from "@lib/gate.js";
|
|
|
|
const units = flatten(curriculum as unknown as Curriculum);
|
|
const indexOf = (id: string) => units.findIndex((u) => u.id === id);
|
|
|
|
describe("bands", () => {
|
|
it("gives one band per curriculum phase, widening as it goes", () => {
|
|
expect(BANDS).toHaveLength(6);
|
|
const ceilings = BANDS.map((b) => b.maxFreq);
|
|
for (let i = 1; i < ceilings.length; i++) {
|
|
expect(ceilings[i]!, `band ${i} must not be narrower than band ${i - 1}`).toBeGreaterThan(
|
|
ceilings[i - 1]!,
|
|
);
|
|
}
|
|
});
|
|
|
|
it("maps phases and unit ids onto bands", () => {
|
|
expect(bandForPhase(1)).toBe(0);
|
|
expect(bandForPhase(6)).toBe(5);
|
|
expect(bandForUnit("1.1")).toBe(0);
|
|
expect(bandForUnit("3.4")).toBe(2);
|
|
expect(bandForUnit("6.8")).toBe(5);
|
|
});
|
|
|
|
it("clamps a phase outside the curriculum rather than throwing", () => {
|
|
expect(bandForPhase(0)).toBe(0);
|
|
expect(bandForPhase(99)).toBe(5);
|
|
});
|
|
|
|
it("admits phase 1 to curated words only — no frequency band at all", () => {
|
|
expect(ceilingForBand(0)).toBe(0);
|
|
// Even the single most frequent word in Korean is not admitted by rank.
|
|
expect(bandOf({ source: "kaikki", freqRank: 1, level: null })).toBeGreaterThan(0);
|
|
// But a curated word is always available.
|
|
expect(bandOf({ source: "curated", freqRank: null, level: null })).toBe(0);
|
|
expect(bandOf({ source: "grammar", freqRank: null, level: null })).toBe(0);
|
|
});
|
|
|
|
it("puts a common word in an early band and a rare one late", () => {
|
|
const common = bandOf({ source: "kaikki", freqRank: 200, level: null });
|
|
const mid = bandOf({ source: "kaikki", freqRank: 4000, level: null });
|
|
const rare = bandOf({ source: "kaikki", freqRank: 12000, level: null });
|
|
expect(common).toBeLessThan(mid);
|
|
expect(mid).toBeLessThan(rare);
|
|
});
|
|
|
|
it("lets the curated level override raw frequency", () => {
|
|
// Same rank, different graded level — the graded one lands later.
|
|
const beginner = bandOf({ source: "krdict", freqRank: 900, level: "초급" });
|
|
const advanced = bandOf({ source: "krdict", freqRank: 900, level: "고급" });
|
|
expect(beginner).toBeLessThan(advanced);
|
|
});
|
|
|
|
it("sends unranked and very rare words to the reference band", () => {
|
|
expect(bandOf({ source: "kaikki", freqRank: null, level: null })).toBe(REFERENCE_BAND);
|
|
expect(bandOf({ source: "kaikki", freqRank: 999_999, level: null })).toBe(REFERENCE_BAND);
|
|
});
|
|
|
|
it("keeps the reference band above every gated band, so it can never leak", () => {
|
|
for (const b of BANDS) expect(REFERENCE_BAND).toBeGreaterThan(b.band);
|
|
});
|
|
});
|
|
|
|
describe("the phonological ladder", () => {
|
|
it("agrees with validate.mjs — no curriculum word breaks its own unit", () => {
|
|
const violations: string[] = [];
|
|
units.forEach((u, i) => {
|
|
const level = featureLevel(i, indexOf);
|
|
for (const w of u.words ?? []) {
|
|
const v = phonologyViolations(w, level);
|
|
if (v.length) violations.push(`${u.id} ${w}: ${v.join(", ")}`);
|
|
}
|
|
});
|
|
expect(violations).toEqual([]);
|
|
});
|
|
|
|
it("climbs monotonically through phase 1", () => {
|
|
const levels = ["1.1", "1.2", "1.3", "1.4", "1.5", "1.6", "1.7", "1.8"].map((id) =>
|
|
featureLevel(indexOf(id), indexOf),
|
|
);
|
|
for (let i = 1; i < levels.length; i++) {
|
|
expect(levels[i]!, `unit 1.${i + 1}`).toBeGreaterThan(levels[i - 1]!);
|
|
}
|
|
});
|
|
|
|
it("blocks a sound the learner has not reached", () => {
|
|
const at11 = featureLevel(indexOf("1.1"), indexOf);
|
|
// Unit 1.1 is batchim-free and uses only the ten basic vowels.
|
|
expect(isReadableAt("가", at11)).toBe(true);
|
|
expect(isReadableAt("밥", at11)).toBe(false); // final consonant
|
|
expect(isReadableAt("개", at11)).toBe(false); // compound vowel
|
|
expect(isReadableAt("까", at11)).toBe(false); // tense consonant
|
|
});
|
|
|
|
it("admits a double batchim only once 1.7 has been reached", () => {
|
|
const before = featureLevel(indexOf("1.6"), indexOf);
|
|
const after = featureLevel(indexOf("1.7"), indexOf);
|
|
expect(isReadableAt("값", before)).toBe(false);
|
|
expect(isReadableAt("값", after)).toBe(true);
|
|
});
|
|
|
|
it("admits a liaison context only once 1.5 has been reached", () => {
|
|
const before = featureLevel(indexOf("1.4"), indexOf);
|
|
const after = featureLevel(indexOf("1.5"), indexOf);
|
|
expect(phonologyViolations("음악", before)).toContain("liaison context");
|
|
expect(isReadableAt("음악", after)).toBe(true);
|
|
});
|
|
|
|
it("is a no-op once the sound phase is over", () => {
|
|
const at61 = featureLevel(indexOf("6.1"), indexOf);
|
|
expect(at61).toBe(LADDER_COMPLETE);
|
|
for (const w of ["괜찮다", "값", "읽었어", "많이"]) {
|
|
expect(isReadableAt(w, at61), w).toBe(true);
|
|
}
|
|
});
|
|
|
|
it("ignores non-Hangul characters rather than choking on them", () => {
|
|
expect(phonologyViolations("ABC 123", 0)).toEqual([]);
|
|
expect(phonologyViolations("", 0)).toEqual([]);
|
|
});
|
|
});
|