test: update structured render contracts

Author: audreyt

data/README.md

@@ -20,10 +20,10 @@ bun run extract:readings -- --full
 ```
 
 Files:
-- `koktai-sinogram-readings.jsonl` — per-sinogram blocks.
-- `koktai-word-readings.jsonl` — word entries with reduced 台語 tokens.
-- `extract-stats.json` — corpus counters and coverage.
-- `extract-anomalies.jsonl` — preserved parse/round-trip anomaly records.
-- `koktai-pron.tei.xml` — TEI pronunciation/usg layer for sinogram blocks.
+- koktai-sinogram-readings.jsonl — per-sinogram blocks.
+- koktai-word-readings.jsonl — word entries with reduced 台語 tokens.
+- extract-stats.json — corpus counters and coverage.
+- extract-anomalies.jsonl — preserved parse/round-trip anomaly records.
+- koktai-pron.tei.xml — TEI pronunciation/usg layer for sinogram blocks.
 
-Current committed extraction stats: 26 volumes, 31,143 sinogram blocks, 34,382 word records, 45,067 senses, 31,061 parsed reading lines, 480,942 reading values, 0 round-trip failures, 0 anomalies.
+Current extraction stats: 26 volumes, 11,810 sinogram blocks, 34,382 word records, 45,067 senses, 31,143 reading lines, 31,061 parsed reading lines, 480,942 reading values, 0 round-trip failures, 0 anomalies.

scripts/extract-readings.ts

@@ -142,6 +142,7 @@ Regenerate:
 
 \`\`\`bash
 bun run extract:readings
+bun run export:tei
 \`\`\`
 
 Optional full token streams (gitignored):
@@ -151,10 +152,13 @@ bun run extract:readings -- --full
 \`\`\`
 
 Files:
-- \`koktai-sinogram-readings.jsonl\` — per-sinogram blocks
-- \`koktai-word-readings.jsonl\` — word entries (reduced taigi)
-- \`extract-stats.json\` — corpus counters
-- \`extract-anomalies.jsonl\` — parse/roundtrip issues
+- koktai-sinogram-readings.jsonl — per-sinogram blocks.
+- koktai-word-readings.jsonl — word entries with reduced 台語 tokens.
+- extract-stats.json — corpus counters and coverage.
+- extract-anomalies.jsonl — preserved parse/round-trip anomaly records.
+- koktai-pron.tei.xml — TEI pronunciation/usg layer for sinogram blocks.
+
+Current extraction stats: ${result.stats.volumes} volumes, ${result.stats.sinograms.toLocaleString("en-US")} sinogram blocks, ${result.stats.words.toLocaleString("en-US")} word records, ${result.stats.senses.toLocaleString("en-US")} senses, ${result.stats.readingLines.toLocaleString("en-US")} reading lines, ${result.stats.readingLinesParsed.toLocaleString("en-US")} parsed reading lines, ${result.stats.readings.toLocaleString("en-US")} reading values, ${result.stats.roundtripFailures} round-trip failures, ${result.stats.anomalies} anomalies.
 `;
 
 writeFileSync(join(outDir, "README.md"), readme, "utf8");

tests/render-contract.test.ts

@@ -8,8 +8,8 @@ const indexPageSource = readFileSync("src/pages/index.astro", "utf8");
 
 describe("dictionary render contracts", () => {
   test("volume ranges use wave dash between bopomofo syllables", () => {
-    expect(volumePageSource).toContain("${sections[0].syllable}～${sections[sections.length - 1].syllable}");
-    expect(volumePageSource).not.toContain("${sections[0].syllable}—${sections[sections.length - 1].syllable}");
+    expect(volumePageSource).toContain("${railSections[0]!.syllable}～${railSections[railSections.length - 1]!.syllable}");
+    expect(volumePageSource).not.toContain("${railSections[0]!.syllable}—${railSections[railSections.length - 1]!.syllable}");
     expect(indexPageSource).toContain("{v.first}～{v.last}");
     expect(indexPageSource).not.toContain("{v.first}—{v.last}");
   });