glossarist 0.4.49 → 0.4.51

package/src/diff/diff-patch.ts

@@ -1,260 +0,0 @@
-// @ts-nocheck — TEMPORARY during TS migration. TODO(Phase 2e): remove and type fully.
-import { Concept } from '../models/concept.js';
-import { ConceptSource } from '../models/concept-source.js';
-import { ConceptDate } from '../models/concept-date.js';
-import { RelatedConcept } from '../models/related-concept.js';
-import { Designation } from '../models/designation.js';
-import { DetailedDefinition } from '../models/detailed-definition.js';
-import { HyperedgeRegistry } from '../models/hyperedge-registry.js';
-import { Added, Removed, Changed } from './change.js';
-import { ListDiff } from './list-diff.js';
-import { TextDiff, TextHunk } from './text-diff.js';
-import { identityOf } from './identity.js';
-import {
-  ConceptDiff,
-  ConceptLevelDiff,
-  LocalizedConceptDiff,
-  MetadataDiff,
-} from './concept-diff.js';
-
-// Wire-name (camelCase JS field → snake_case JSON key) for metadata
-// fields patched onto Concept and LocalizedConcept.
-//
-// These mirror `Concept.wireNameFor` / `LocalizedConcept.wireNameFor`.
-// The duplication is intentional: diff-patch.js cannot import those
-// models lazily without making applyDiff async (a breaking API change),
-// and the field set is small and stable. A spec
-// (test/diff/field-sync.test.js) asserts the two stay in sync.
-
-const CONCEPT_METADATA_JSON_KEYS = Object.freeze({
-  status: 'status',
-  term: 'term',
-  uri: 'uri',
-  schemaVersion: 'schema_version',
-});
-
-const LOC_METADATA_JSON_KEYS = Object.freeze({
-  entryStatus: 'entry_status',
-  classification: 'classification',
-  reviewType: 'review_type',
-  domain: 'domain',
-  release: 'release',
-  lineageSourceSimilarity: 'lineage_source_similarity',
-  script: 'script',
-  system: 'system',
-  reviewDate: 'review_date',
-  reviewDecisionDate: 'review_decision_date',
-  reviewDecisionEvent: 'review_decision_event',
-  reviewStatus: 'review_status',
-  reviewDecision: 'review_decision',
-  reviewDecisionNotes: 'review_decision_notes',
-});
-
-export function applyDiff(oldConcept, diff) {
-  const json = oldConcept.toJSON();
-
-  applyConceptLevelPatch(json, diff.concept);
-  applyLanguagePatch(json, diff);
-
-  for (const [lang, lcDiff] of Object.entries(diff.localizations)) {
-    applyLocalizedPatch(json, lang, lcDiff);
-  }
-
-  return Concept.fromJSON(json);
-}
-
-export function reverseDiff(diff) {
-  const concept = new ConceptLevelDiff({
-    sources: reverseListDiff(diff.concept.sources),
-    dates: reverseListDiff(diff.concept.dates),
-    relatedConcepts: reverseListDiff(diff.concept.relatedConcepts),
-    relations: reverseListDiff(diff.concept.relations),
-    groups: reverseListDiff(diff.concept.groups),
-    sections: reverseListDiff(diff.concept.sections),
-    tags: reverseListDiff(diff.concept.tags),
-    metadata: reverseMetadataDiff(diff.concept.metadata),
-  });
-
-  const languages = reverseListDiff(diff.languages);
-
-  const localizations = {};
-  for (const [lang, lcDiff] of Object.entries(diff.localizations)) {
-    localizations[lang] = new LocalizedConceptDiff({
-      languageCode: lcDiff.languageCode,
-      designations: reverseListDiff(lcDiff.designations),
-      definitions: reverseListDiff(lcDiff.definitions),
-      notes: reverseListDiff(lcDiff.notes),
-      examples: reverseListDiff(lcDiff.examples),
-      sources: reverseListDiff(lcDiff.sources),
-      dates: reverseListDiff(lcDiff.dates),
-      related: reverseListDiff(lcDiff.related),
-      metadata: reverseMetadataDiff(lcDiff.metadata),
-      totalItems: lcDiff._totalItems,
-    });
-  }
-
-  return new ConceptDiff({
-    oldId: diff.newId,
-    newId: diff.oldId,
-    concept,
-    languages,
-    localizations,
-    totalItems: diff._totalItems,
-  });
-}
-
-function applyConceptLevelPatch(json, conceptDiff) {
-  json.sources = applyListPatch(json.sources ?? [], conceptDiff.sources, ConceptSource.identityOf);
-  json.dates = applyListPatch(json.dates ?? [], conceptDiff.dates, ConceptDate.identityOf);
-  json.related = applyListPatch(json.related ?? [], conceptDiff.relatedConcepts, RelatedConcept.identityOf);
-
-  // Unified hyperedge diff → typed wire keys. Iterate the registry;
-  // for each registered class, partition the unified diff by that
-  // class and apply to the class's wireKey. Adding a new hyperedge
-  // type means this loop picks it up automatically — no edits here.
-  for (const cls of HyperedgeRegistry.allClasses()) {
-    const partition = partitionListDiff(conceptDiff.relations, cls);
-    if (!partition.hasChanges) continue;
-    const wireKey = cls.wireKey;
-    // v1 wire keys (partitive_hyperedges) on the input are migrated
-    // to v2 wire key here: read from either, write to v2 only.
-    const v1ReadKeys = cls.v1WireKeys ?? [];
-    const existing = v1ReadKeys
-      .map(k => json[k])
-      .find(arr => Array.isArray(arr)) ?? json[wireKey] ?? [];
-    json[wireKey] = applyListPatch(existing, partition, cls.identityOf);
-    for (const k of v1ReadKeys) {
-      if (json[k] != null && json[wireKey] != null) delete json[k];
-    }
-  }
-
-  json.tags = applyListPatch(json.tags ?? [], conceptDiff.tags);
-
-  applyMetadataPatch(json, conceptDiff.metadata, CONCEPT_METADATA_JSON_KEYS);
-}
-
-function applyLanguagePatch(json, conceptDiff) {
-  if (!json.localizations) json.localizations = {};
-  for (const entry of conceptDiff.languages.added) {
-    if (!(entry.value in json.localizations)) {
-      json.localizations[entry.value] = {};
-    }
-  }
-  for (const entry of conceptDiff.languages.removed) {
-    delete json.localizations[entry.value];
-  }
-}
-
-function applyLocalizedPatch(json, lang, lcDiff) {
-  if (!json.localizations) json.localizations = {};
-  if (!(lang in json.localizations)) {
-    json.localizations[lang] = {};
-  }
-  const loc = json.localizations[lang];
-
-  loc.terms = applyListPatch(loc.terms ?? [], lcDiff.designations, Designation.identityOf);
-  loc.definition = applyListPatch(loc.definition ?? [], lcDiff.definitions, DetailedDefinition.identityOf);
-  loc.notes = applyListPatch(loc.notes ?? [], lcDiff.notes, DetailedDefinition.identityOf);
-  loc.examples = applyListPatch(loc.examples ?? [], lcDiff.examples, DetailedDefinition.identityOf);
-  loc.sources = applyListPatch(loc.sources ?? [], lcDiff.sources, ConceptSource.identityOf);
-  loc.dates = applyListPatch(loc.dates ?? [], lcDiff.dates, ConceptDate.identityOf);
-  loc.related = applyListPatch(loc.related ?? [], lcDiff.related, RelatedConcept.identityOf);
-
-  applyMetadataPatch(loc, lcDiff.metadata, LOC_METADATA_JSON_KEYS);
-}
-
-function applyListPatch(existingItems, listDiff, identityFn) {
-  const fn = identityFn ?? identityOf;
-  let result = [...existingItems];
-
-  for (const entry of listDiff.added) {
-    result.push(toJsonValue(entry.value));
-  }
-
-  for (const entry of listDiff.removed) {
-    const targetKey = fn(entry.value);
-    result = result.filter(item => fn(item) !== targetKey);
-  }
-
-  for (const entry of listDiff.changed) {
-    const oldKey = fn(entry.oldValue);
-    result = result.map(item =>
-      fn(item) === oldKey ? toJsonValue(entry.newValue) : item,
-    );
-  }
-
-  return result;
-}
-
-function applyMetadataPatch(target, metadataDiff, keyMap) {
-  for (const [field, change] of Object.entries(metadataDiff.changes)) {
-    const jsonKey = keyMap[field];
-    if (!jsonKey) continue;
-    if (change.newValue == null) {
-      delete target[jsonKey];
-    } else {
-      target[jsonKey] = change.newValue;
-    }
-  }
-}
-
-function reverseListDiff(listDiff) {
-  const added = listDiff.removed.map(r => new Added({ value: r.value, path: r.path }));
-  const removed = listDiff.added.map(a => new Removed({ value: a.value, path: a.path }));
-  const changed = listDiff.changed.map(c => new Changed({
-    oldValue: c.newValue,
-    newValue: c.oldValue,
-    textDiff: c.textDiff ? reverseTextDiff(c.textDiff) : null,
-    path: c.path,
-  }));
-  return new ListDiff({ added, removed, changed });
-}
-
-function reverseMetadataDiff(metadataDiff) {
-  const changes = {};
-  for (const [field, change] of Object.entries(metadataDiff.changes)) {
-    changes[field] = new Changed({
-      oldValue: change.newValue,
-      newValue: change.oldValue,
-    });
-  }
-  return new MetadataDiff({ changes });
-}
-
-function reverseTextDiff(textDiff) {
-  const hunks = textDiff.hunks.map(h => {
-    if (h.type === 'added') return new TextHunk({ type: 'removed', text: h.text });
-    if (h.type === 'removed') return new TextHunk({ type: 'added', text: h.text });
-    return new TextHunk({ type: 'equal', text: h.text });
-  });
-  return new TextDiff({
-    oldText: textDiff.newText,
-    newText: textDiff.oldText,
-    hunks,
-  });
-}
-
-function toJsonValue(value) {
-  if (value == null) return null;
-  if (typeof value === 'string' || typeof value === 'number' || typeof value === 'boolean') {
-    return value;
-  }
-  if (typeof value.toJSON === 'function') return value.toJSON();
-  return value;
-}
-
-// Project a unified relations ListDiff onto a single concrete class.
-// Entries whose value (old or new) is an instance of `Cls` survive;
-// everything else is filtered out. Cross-type changed entries (e.g.
-// old was PartitiveHyperedge, new is GenericHyperedge) are not yet
-// supported — those would need to surface as a remove + add pair.
-function partitionListDiff(listDiff, Cls) {
-  if (!listDiff) return new ListDiff({ added: [], removed: [], changed: [] });
-  return new ListDiff({
-    added: listDiff.added.filter(e => e.value instanceof Cls),
-    removed: listDiff.removed.filter(e => e.value instanceof Cls),
-    changed: listDiff.changed.filter(e =>
-      (e.oldValue == null || e.oldValue instanceof Cls) &&
-      (e.newValue == null || e.newValue instanceof Cls)),
-  });
-}

package/src/diff/diff-renderer.ts

@@ -1,293 +0,0 @@
-// @ts-nocheck — TEMPORARY during TS migration. TODO(Phase 2e): remove and type fully.
-import { TextDiff } from './text-diff.js';
-import { resolveMultiplicity } from '../models/multiplicity.js';
-
-const COLORS = {
-  red: s => `\x1b[31m${s}\x1b[0m`,
-  green: s => `\x1b[32m${s}\x1b[0m`,
-  yellow: s => `\x1b[33m${s}\x1b[0m`,
-  cyan: s => `\x1b[36m${s}\x1b[0m`,
-  dim: s => `\x1b[2m${s}\x1b[0m`,
-  bold: s => `\x1b[1m${s}\x1b[0m`,
-};
-
-export function renderConceptDiff(diff, options = {}) {
-  const colors = options.colors ?? false;
-  const showUnchanged = options.showUnchanged ?? false;
-  const lines = [];
-
-  const pct = Math.round(diff.similarity * 100);
-  const idPart = diff.oldId && diff.newId && diff.oldId !== diff.newId
-    ? `${diff.oldId} → ${diff.newId}`
-    : (diff.newId ?? diff.oldId ?? '?');
-
-  const similarityStr = colors ? colorizeSimilarity(pct) : `${pct}%`;
-  lines.push(`Concept "${idPart}" — ${similarityStr} similar`);
-
-  if (!diff.hasChanges && !showUnchanged) {
-    lines.push('  (no changes)');
-    return lines.join('\n');
-  }
-
-  if (diff.concept.hasChanges || showUnchanged) {
-    lines.push('');
-    lines.push('Concept-level:');
-    lines.push(renderConceptLevel(diff.concept, colors));
-  }
-
-  if (diff.languages.hasChanges || showUnchanged) {
-    lines.push('');
-    lines.push('Languages:');
-    lines.push(renderLanguageSet(diff.languages, colors));
-  }
-
-  for (const lang of diff.localizationLanguages) {
-    const lcDiff = diff.localization(lang);
-    if (!lcDiff) continue;
-    if (lcDiff.hasChanges || showUnchanged) {
-      lines.push('');
-      lines.push(`Localization (${lang}):`);
-      lines.push(renderLocalized(lcDiff, colors));
-    }
-  }
-
-  return lines.join('\n');
-}
-
-export function renderCollectionDiff(diff, options = {}) {
-  const colors = options.colors ?? false;
-  const lines = [];
-
-  const pct = Math.round(diff.similarity * 100);
-  const similarityStr = colors ? colorizeSimilarity(pct) : `${pct}%`;
-
-  lines.push(`Collection comparison — ${similarityStr} similar overall`);
-  lines.push(`  Old: ${diff.oldCount} concepts`);
-  lines.push(`  New: ${diff.newCount} concepts`);
-  lines.push(`  Matched: ${diff.matched.length}`);
-  lines.push(`  Added: ${diff.added.length}`);
-  lines.push(`  Removed: ${diff.removed.length}`);
-
-  if (diff.added.length > 0) {
-    lines.push('');
-    lines.push(colors ? COLORS.green('Added concepts:') : 'Added concepts:');
-    for (const entry of diff.added) {
-      lines.push(`  + ${entry.value}`);
-    }
-  }
-
-  if (diff.removed.length > 0) {
-    lines.push('');
-    lines.push(colors ? COLORS.red('Removed concepts:') : 'Removed concepts:');
-    for (const entry of diff.removed) {
-      lines.push(`  - ${entry.value}`);
-    }
-  }
-
-  const changedDiffs = Object.entries(diff.conceptDiffs)
-    .filter(([, d]) => d.hasChanges)
-    .sort(([, a], [, b]) => a.similarity - b.similarity);
-
-  if (changedDiffs.length > 0) {
-    lines.push('');
-    lines.push('Changed concepts:');
-    for (const [id, conceptDiff] of changedDiffs) {
-      const conceptPct = Math.round(conceptDiff.similarity * 100);
-      const pctStr = colors ? colorizeSimilarity(conceptPct) : `${conceptPct}%`;
-      lines.push(`  ${id}: ${pctStr}`);
-    }
-  }
-
-  return lines.join('\n');
-}
-
-export function renderTextDiff(textDiff, options = {}) {
-  if (!(textDiff instanceof TextDiff)) {
-    textDiff = TextDiff.fromJSON(textDiff);
-  }
-  const colors = options.colors ?? false;
-  const lines = [];
-
-  for (const hunk of textDiff.hunks) {
-    if (hunk.type === 'equal') {
-      lines.push(`  ${hunk.text}`);
-    } else if (hunk.type === 'added') {
-      lines.push(colors ? `+ ${COLORS.green(hunk.text)}` : `+ ${hunk.text}`);
-    } else if (hunk.type === 'removed') {
-      lines.push(colors ? `- ${COLORS.red(hunk.text)}` : `- ${hunk.text}`);
-    }
-  }
-
-  return lines.join('\n');
-}
-
-function renderConceptLevel(diff, colors) {
-  const lines = [];
-  lines.push(...renderListDiff('Sources', diff.sources, colors, itemLabel));
-  lines.push(...renderListDiff('Dates', diff.dates, colors, itemLabel));
-  lines.push(...renderListDiff('Related', diff.relatedConcepts, colors, itemLabel));
-  // Unified hyperedge section. Replaces the per-type sections
-  // (Partitive relations, Generic relations) — the underlying diff
-  // is a single ListDiff over the unified .relations array.
-  lines.push(...renderListDiff('Hyperedges', diff.relations, colors, relationLabel));
-  lines.push(...renderListDiff('Groups', diff.groups, colors, itemLabel));
-  lines.push(...renderListDiff('Sections', diff.sections, colors, itemLabel));
-  lines.push(...renderListDiff('Tags', diff.tags, colors, itemLabel));
-  lines.push(...renderMetadataDiff(diff.metadata));
-  return lines.filter(Boolean).join('\n');
-}
-
-function renderLanguageSet(diff, colors) {
-  const lines = [];
-  for (const entry of diff.added) {
-    lines.push(colors ? `  + ${COLORS.green(entry.value)}` : `  + ${entry.value}`);
-  }
-  for (const entry of diff.removed) {
-    lines.push(colors ? `  - ${COLORS.red(entry.value)}` : `  - ${entry.value}`);
-  }
-  return lines.join('\n');
-}
-
-function renderLocalized(diff, colors) {
-  const lines = [];
-  lines.push(...renderListDiff('Designations', diff.designations, colors, designationLabel));
-  lines.push(...renderListDiff('Definitions', diff.definitions, colors, definitionLabel));
-  lines.push(...renderListDiff('Notes', diff.notes, colors, definitionLabel));
-  lines.push(...renderListDiff('Examples', diff.examples, colors, definitionLabel));
-  lines.push(...renderListDiff('Sources', diff.sources, colors, itemLabel));
-  lines.push(...renderListDiff('Dates', diff.dates, colors, itemLabel));
-  lines.push(...renderListDiff('Related', diff.related, colors, itemLabel));
-  lines.push(...renderMetadataDiff(diff.metadata));
-  return lines.filter(Boolean).join('\n');
-}
-
-function renderListDiff(label, listDiff, colors, labelFn) {
-  if (!listDiff.hasChanges) return [];
-  const lines = [`  ${label}:`];
-  for (const entry of listDiff.added) {
-    const text = `    + ${labelFn(entry.value)}`;
-    lines.push(colors ? COLORS.green(text) : text);
-  }
-  for (const entry of listDiff.removed) {
-    const text = `    - ${labelFn(entry.value)}`;
-    lines.push(colors ? COLORS.red(text) : text);
-  }
-  for (const entry of listDiff.changed) {
-    const oldLabel = labelFn(entry.oldValue);
-    const newLabel = labelFn(entry.newValue);
-    lines.push(`    ~ ${oldLabel} → ${newLabel}`);
-  }
-  return lines;
-}
-
-function renderMetadataDiff(metadataDiff) {
-  if (!metadataDiff.hasChanges) return [];
-  const lines = [`  Metadata:`];
-  for (const [field, change] of Object.entries(metadataDiff.changes)) {
-    const oldVal = formatScalar(change.oldValue);
-    const newVal = formatScalar(change.newValue);
-    lines.push(`    ~ ${field}: ${oldVal} → ${newVal}`);
-  }
-  return lines;
-}
-
-function designationLabel(d) {
-  if (!d) return '?';
-  const text = d.designation ?? '?';
-  const status = d.normativeStatus ? ` (${d.normativeStatus})` : '';
-  return `${text}${status}`;
-}
-
-function definitionLabel(d) {
-  if (!d) return '?';
-  return d.content ?? '?';
-}
-
-function relationLabel(r) {
-  if (!r) return '?';
-  const c = r.comprehensive;
-  const head = c?.id ?? c?.source ?? c?.text ?? '?';
-  // PartitiveHyperedge exposes .partitives as the wire alias for
-  // .members; GenericHyperedge uses .members directly. Read both.
-  const members = Array.isArray(r.partitives) ? r.partitives : r.members;
-  const memberText = Array.isArray(members)
-    ? members.map(m => {
-        const ref = m?.ref ?? m ?? {};
-        let tail = '';
-        const mult = resolveMultiplicity(m);
-        if (mult && mult !== 'compulsory') tail += ` (${mult})`;
-        // PartitiveMember uses is_delimiting (boolean flag).
-        if (m?.is_delimiting === true) tail += ' ⊘';
-        // GenericMember uses delimitingCharacteristic (string value per
-        // ISO 704:2022 §5.5.4.2.1) — show the delimiting characteristic
-        // text so reviewers can see the intension difference.
-        if (m?.delimitingCharacteristic) {
-          const c = Object.values(m.delimitingCharacteristic)[0];
-          if (typeof c === 'string' && c.length > 0) tail += ` ⟨${c}⟩`;
-        }
-        return `${ref.id ?? ref.source ?? ref.text ?? '?'}${tail}`;
-      }).join(', ')
-    : '';
-  const completeness = r.completeness ? ` (${r.completeness})` : '';
-  const criterion = r.criterion
-    ? ` / ${Object.values(r.criterion)[0] ?? ''}`
-    : '';
-  // Each hyperedge leaf declares a short static kindLabel (PART, GEN, …).
-  // Read it via the constructor so adding a new hyperedge type needs no
-  // renderer edit. The blank-string default keeps backward-compat with
-  // any AbstractHyperedge subclass that doesn't declare kindLabel.
-  const kindLabel = r?.constructor?.kindLabel ?? '';
-  const typeTag = kindLabel ? `${kindLabel} ` : '';
-  return `${typeTag}${head} → {${memberText}}${completeness}${criterion}`;
-}
-
-// Aggregate multiplicity + delimiting stats across a concept's
-// PartitiveRelations. Used by dataset-level summaries, not by the
-// per-relation diff label.
-//
-// v4: presence and count are independent dimensions. Stats include
-// byPresence and byCount breakdowns so tools can query each axis
-// independently ("show me all optional parts" or "show me all
-// multiple-count parts").
-export function multiplicityStats(relations) {
-  const byMultiplicity = {};
-  const byPresence = {};
-  const byCount = {};
-  let delimitingCount = 0;
-  let total = 0;
-  for (const rel of relations ?? []) {
-    for (const m of rel?.partitives ?? []) {
-      total += 1;
-      const mult = resolveMultiplicity(m);
-      byMultiplicity[mult] = (byMultiplicity[mult] ?? 0) + 1;
-      const p = m?.presence ?? 'required';
-      const c = m?.count ?? 'exactly_one';
-      byPresence[p] = (byPresence[p] ?? 0) + 1;
-      byCount[c] = (byCount[c] ?? 0) + 1;
-      if (m?.is_delimiting === true) delimitingCount += 1;
-    }
-  }
-  return { total, byMultiplicity, byPresence, byCount, delimiting: delimitingCount };
-}
-
-function itemLabel(item) {
-  if (item == null) return '?';
-  if (typeof item === 'string') return item;
-  if (typeof item === 'number') return String(item);
-  if (typeof item.toJSON === 'function') {
-    return JSON.stringify(item.toJSON());
-  }
-  return String(item);
-}
-
-function formatScalar(val) {
-  if (val == null) return '∅';
-  if (typeof val === 'string') return val;
-  return String(val);
-}
-
-function colorizeSimilarity(pct) {
-  if (pct >= 95) return COLORS.green(`${pct}%`);
-  if (pct >= 80) return COLORS.yellow(`${pct}%`);
-  return COLORS.red(`${pct}%`);
-}

package/src/diff/identity.ts

@@ -1,33 +0,0 @@
-// Shared identity helper for the diff layer.
-//
-// Per invariant N1, identity is a model concern. Models expose
-// `identity()` returning a stable string. This helper dispatches: if
-// the value is a model with `identity()`, call it; otherwise fall
-// back to canonical JSON of `toJSON()`.
-
-import { canonicalJson } from './canonical-json.js';
-
-interface Identifiable {
-  identity(): string;
-}
-interface HasToJSON {
-  toJSON(): unknown;
-}
-
-function hasIdentity(v: unknown): v is Identifiable {
-  return v != null && typeof v === 'object' && typeof (v as { identity?: unknown }).identity === 'function';
-}
-
-function hasToJSON(v: unknown): v is HasToJSON {
-  return v != null && typeof v === 'object' && typeof (v as { toJSON?: unknown }).toJSON === 'function';
-}
-
-export function identityOf(item: unknown): string {
-  if (item == null) return '';
-  if (hasIdentity(item)) return item.identity();
-  if (typeof item === 'string') return item;
-  if (typeof item === 'number') return String(item);
-  if (typeof item === 'boolean') return String(item);
-  if (hasToJSON(item)) return canonicalJson(item.toJSON());
-  return canonicalJson(item);
-}