calliope-ts 0.0.1 → 0.0.2

package/tsconfig.json

@@ -0,0 +1,29 @@
+{
+  "compilerOptions": {
+    "target": "ES2022",
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "lib": ["ES2022", "DOM", "DOM.Iterable"],
+    "outDir": "./dist",
+    "rootDir": "./src",
+    "strict": true,
+    "esModuleInterop": true,
+    "skipLibCheck": true,
+    "forceConsistentCasingInFileNames": true,
+    "resolveJsonModule": true,
+    "declaration": true,
+    "declarationMap": true,
+    "allowJs": true,
+    "allowSyntheticDefaultImports": true,
+    "paths": {
+      "@/*": ["./*"]
+    }
+  },
+  "include": [
+    "src"
+  ],
+  "exclude": [
+    "node_modules",
+    "dist"
+  ]
+}

package/vitest.config.d.ts

@@ -0,0 +1,2 @@
+declare const _default: import("vite").UserConfig;
+export default _default;

package/vitest.config.js

@@ -0,0 +1,14 @@
+// vitest.config.ts
+import { defineConfig } from 'vitest/config';
+export default defineConfig({
+    test: {
+        globals: true,
+        environment: 'node',
+        include: ['tests/**/*.test.ts'],
+        testTimeout: 10000,
+        deps: {
+            // required for esm packages like finnlp
+            interopDefault: true,
+        },
+    },
+});

package/vitest.config.ts

@@ -0,0 +1,15 @@
+// vitest.config.ts
+import { defineConfig } from 'vitest/config';
+
+export default defineConfig({
+  test: {
+    globals: true,
+    environment: 'node',
+    include: ['tests/**/*.test.ts'],
+    testTimeout: 10000,
+    deps: {
+      // required for esm packages like finnlp
+      interopDefault: true,
+    },
+  },
+});

package/.claude/settings.local.json

@@ -1,8 +0,0 @@
-{
-  "permissions": {
-    "allow": [
-      "Bash(npx vitest *)",
-      "Bash(npx ts-node *)"
-    ]
-  }
-}

package/.opencode/skills/calliope-ts/SKILL.md

@@ -1,74 +0,0 @@
----
-name: calliope-ts
-description: Phonological-syntactic scansion of English poetry implementing McAleese (2008) methodology. Use when working on Calliope TS — stress assignment, meter detection, prosodic hierarchy, FinNLP parsing, CMU pronouncing lookup. Triggers: calliope, scansion, poetry meter, iambic, trochaic, anapestic, dactylic, nuclear stress, compound stress, prosodic hierarchy, FinNLP, en-pos, en-parse.
----
-
-# Calliope TS — Poetry Scansion Toolkit
-
-## Core Principle
-
-**Phonological scansion operates on PHONES (sound units), not graphemes (letter units).** Syntactic parsing (FinNLP) is correctly graphemic. Lexical stress lookup must use phonemic output from CMU dictionary via `Nounsing-Pro`.
-
-## Pipeline (7 steps per McAleese 2008)
-
-1. **Parse** (`parser.ts`) — FinNLP → ClsDocument with words, deps, POS
-2. **Lexical Stress** (`stress.ts`) — CMU dict lookup per word via `Nounsing-Pro` (phonemes → syllable stress)
-3. **Phrase-Level Stress** — Compound + Nuclear + Relative mapping (0→w, 1→n, 2+→s)
-4. **Prosodic Hierarchy** (`phonological.ts`) — CP → PP → IU (Selkirk)
-5. **Key-Stress Extraction** (`scansion.ts`) — Right-edge extraction per unit level
-6. **Metre Scoring** — 8 metres + tie-breaking, weighted by unit importance
-7. **Scansion Assembly** — Align strong stresses to metrical beats, place foot boundaries
-
-## Stress Symbols
-
-| Symbol | Level | Mapping |
-|--------|-------|---------|
-| w | Weakest | Numeric 0 |
-| n | Low | Numeric 1 |
-| m | Moderate/Secondary | Numeric 2 |
-| s | Strongest/Primary | Numeric 3+ |
-| x | Metrically insignificant | Key-stress masking |
-| W | Weak metrical position | |
-| S | Strong metrical position | |
-
-## Key Architectural Decisions
-
-### Nuclear Stress (2026-05-07 fix)
-- **Old**: 5-level right-to-left gradient (5→4→3→2→1) — killed triple meter detection
-- **New**: Only +1 boost for rightmost content word. `lexicalStress` field preserves CMU values for meter scoring
-- Meter detection uses `lexicalStress`, NOT nuclear-stress-modified values
-
-### Contraction Handling (2026-05-07 fix)
-- FinNLP `en-norm` deencliticizes: `we'll` → `we` + `will` (2 tokens)
-- Post-parse re-merging restores contracted form as 1 phonetic unit
-- 41 contraction patterns handled (pre-scan → parse → re-merge)
-
-### Punctuation Filtering
-- Punctuation tokens (`,` `.` `?` `!` `;` `:` `--` `...`) filtered from syllable arrays
-- Must NOT appear in scansion strings
-
-### Display (per-syllable)
-- All display functions operate per-SYLLABLE (not per-word)
-- 3-syllable word = 3 independently colored syllables
-
-## Known Gaps (Phase 5/6)
-
-- **Metrical promotion**: Function words in strong positions not promoted — nuclear stress used as proxy
-- **Contextual disambiguation**: Lines like "Burnt after them" need surrounding context for meter
-- **Anacrusis detection**: First weak syllable not properly excluded from foot count
-- **Elision**: Vowel merging across word boundaries not implemented
-- Full McAleese expert baseline matching (~65% target)
-
-## Test Commands
-
-```bash
-npm run build          # Compile TypeScript
-npx tsc --noEmit       # Typecheck
-npx vitest run tests/basic.test.ts   # Run all tests
-```
-
-## Dependencies
-
-- **finnlp** — Penn-treebank POS + Universal Dependency parsing
-- **nounsing-pro** — CMU dictionary lexical stress, and morphological/phonological nuance lookup
-- Never assume other libraries without checking package.json