Add the Next.js web app and shared lesson content.

Move lessons out of the C sources into JSON so the web client can use the same material as the desktop app, and point release builds at desktop/.

Co-authored-by: Cursor <cursoragent@cursor.com>

tools/extract/deutsch.mjs

@@ -0,0 +1,249 @@
+import { readFileSync } from 'node:fs';
+import { parseArray, trimNulls } from '../lib/cparse.mjs';
+
+/**
+ * Deutsch Unit 1 ("Neue Freunde").
+ *
+ * The desktop app defines each exercise's data as a `const` array next to its
+ * `build_*` function; the task order and titles come from
+ * `build_exercise_page` in main.c and `u1_ex_names` in globals.c.
+ */
+
+const opt = (v) => (v === null || v === undefined || v === '' ? undefined : String(v));
+
+/** Split the pipe-separated accepted-answer lists used by TypedQ. */
+const alternatives = (s) => String(s).split('|').map((a) => a.trim()).filter(Boolean);
+
+/** `{before, after, answer}` becomes a single-gap fill sentence. */
+function verbFill(row, meaning, extra = {}) {
+  const [before, after, answer] = row;
+  return {
+    kind: 'fill',
+    segments: [before ?? '', after ?? ''],
+    answers: [String(answer)],
+    input: 'select',
+    ...(meaning ? { meaning: String(meaning) } : {}),
+    ...extra,
+  };
+}
+
+export function extractDeutschUnit1(desktopDir) {
+  const u1 = readFileSync(`${desktopDir}/exercise_u1.c`, 'utf8');
+  const common = readFileSync(`${desktopDir}/exercise_common.c`, 'utf8');
+
+  const tasks = [];
+
+  /* 1 — Dialog: three dialogues sharing one word pool, one gap per line. */
+  {
+    const meanings = parseArray(common, 'ex1_meaning');
+    const pool = parseArray(common, 'ex1_pool');
+    const dialogues = [
+      ['Sich vorstellen', parseArray(common, 'ex1_d1')],
+      ['Begrüßung', parseArray(common, 'ex1_d2')],
+      ['Verabschiedung', parseArray(common, 'ex1_d3')],
+    ];
+
+    let i = 0;
+    const items = dialogues.map(([title, rows]) => ({
+      kind: 'dialogue',
+      title,
+      rows: rows.map(([speaker, before, answer, after]) => ({
+        speaker: speaker ?? '',
+        before: before ?? '',
+        answer: String(answer),
+        after: after ?? '',
+        meaning: opt(meanings[i++]),
+      })),
+    }));
+
+    tasks.push({ number: 1, title: 'Dialog', subtitleKey: 'sub_dialog', pool, items });
+  }
+
+  /* 2 — Sätze bilden */
+  {
+    const meanings = parseArray(common, 'ex2_meaning');
+    tasks.push({
+      number: 2,
+      title: 'Sätze bilden',
+      subtitleKey: 'sub_assembly',
+      items: parseArray(common, 'ex2_items').map(([prompt, words], i) => ({
+        kind: 'assembly',
+        prompt: opt(prompt) ?? '',
+        words: trimNulls(words).map(String),
+        meaning: opt(meanings[i]),
+      })),
+    });
+  }
+
+  /* 3 + 9 — multiple choice */
+  const choiceTask = (number, title, subtitleKey, qsName, meanName) => {
+    const meanings = parseArray(u1, meanName);
+    return {
+      number,
+      title,
+      subtitleKey,
+      items: parseArray(u1, qsName).map(([prompt, options, n, correct], i) => ({
+        kind: 'choice',
+        prompt: String(prompt),
+        options: trimNulls(options).map(String).slice(0, n),
+        correct,
+        meaning: opt(meanings[i]),
+      })),
+    };
+  };
+  tasks.push(choiceTask(3, 'Was ist richtig?', 'sub_choice_num', 'ex3_questions', 'ex3_meaning'));
+
+  /* 4 — Freie Antwort */
+  tasks.push({
+    number: 4,
+    title: 'Freie Antwort',
+    subtitleKey: 'sub_free',
+    items: parseArray(u1, 'ex4_questions').map(([question, questionCs, sample, sampleCs]) => ({
+      kind: 'free',
+      question: String(question),
+      questionCs: opt(questionCs),
+      sample: String(sample),
+      sampleCs: opt(sampleCs),
+    })),
+  });
+
+  /* 5 — Zahlen: digits shown as the clue, pick the German word. */
+  {
+    const meanings = parseArray(u1, 'ex5_meaning');
+    tasks.push({
+      number: 5,
+      title: 'Zahlen',
+      subtitleKey: 'sub_zahlen',
+      pool: parseArray(u1, 'ex5_pool'),
+      items: parseArray(u1, 'ex5_data').map(([digits, answer], i) => ({
+        kind: 'fill',
+        segments: ['', ''],
+        answers: [String(answer)],
+        input: 'select',
+        clue: String(digits),
+        meaning: opt(meanings[i]),
+      })),
+    });
+  }
+
+  /* 6 — Wie viel?: the emoji is repeated `count` times as the clue. */
+  {
+    const meanings = parseArray(u1, 'ex6_meaning');
+    tasks.push({
+      number: 6,
+      title: 'Wie viel?',
+      subtitleKey: 'sub_wieviel',
+      pool: parseArray(u1, 'ex6_pool'),
+      items: parseArray(u1, 'ex6_data').map(([emoji, count, noun, answer], i) => ({
+        kind: 'fill',
+        segments: ['', ` ${noun}`],
+        answers: [String(answer)],
+        input: 'select',
+        emoji: Array.from({ length: count }, () => emoji).join(' '),
+        meaning: opt(meanings[i]),
+      })),
+    });
+  }
+
+  /* 7 — Zahlenreihe */
+  {
+    const meanings = parseArray(u1, 'ex7_meaning');
+    tasks.push({
+      number: 7,
+      title: 'Zahlenreihe',
+      subtitleKey: 'sub_reihe',
+      pool: parseArray(u1, 'ex7_pool'),
+      items: parseArray(u1, 'ex7_data').map((row, i) => verbFill(row, meanings[i])),
+    });
+  }
+
+  /* 8 — Verb einsetzen */
+  {
+    const meanings = parseArray(u1, 'ex8_meaning');
+    tasks.push({
+      number: 8,
+      title: 'Verb einsetzen',
+      subtitleKey: 'sub_verb',
+      pool: parseArray(u1, 'ex8_pool'),
+      items: parseArray(u1, 'ex8_data').map((row, i) => verbFill(row, meanings[i])),
+    });
+  }
+
+  tasks.push(choiceTask(9, 'Wer? Wie? Wo?', 'sub_wer', 'ex9_questions', 'ex9_meaning'));
+
+  /* 10 — Wörter trennen: split a run-together string into words. */
+  {
+    const meanings = parseArray(u1, 'ex10_meaning');
+    tasks.push({
+      number: 10,
+      title: 'Wörter trennen',
+      subtitleKey: 'sub_assembly',
+      items: parseArray(u1, 'ex10_items').map(([prompt, words], i) => ({
+        kind: 'assembly',
+        prompt: opt(prompt) ?? '',
+        words: trimNulls(words).map(String),
+        meaning: opt(meanings[i]),
+      })),
+    });
+  }
+
+  /* 11 + 13 — sort items into groups */
+  const assignTask = (number, title, subtitleKey, itemsName, groupsName, meanName) => {
+    const meanings = parseArray(u1, meanName);
+    return {
+      number,
+      title,
+      subtitleKey,
+      items: [{
+        kind: 'assign',
+        groups: parseArray(u1, groupsName).map(String),
+        items: parseArray(u1, itemsName).map(([emoji, label, group], i) => ({
+          emoji: emoji ?? null,
+          label: String(label),
+          group,
+          meaning: opt(meanings[i]),
+        })),
+      }],
+    };
+  };
+  tasks.push(assignTask(11, 'Grußformen', 'sub_gruss', 'ex11_items', 'ex11_groups', 'ex11_meaning'));
+
+  /* 12 — Was macht er/sie? */
+  {
+    const meanings = parseArray(u1, 'ex12_meaning');
+    tasks.push({
+      number: 12,
+      title: 'Was macht er/sie?',
+      subtitleKey: 'sub_bild',
+      pool: parseArray(u1, 'ex12_pool'),
+      items: parseArray(u1, 'ex12_data').map(([emoji, before, after, answer], i) =>
+        verbFill([before, after, answer], meanings[i], { emoji: String(emoji) })),
+    });
+  }
+
+  tasks.push(assignTask(13, 'Länder', 'sub_land', 'ex13_items', 'ex13_groups', 'ex13_meaning'));
+
+  /* Vocabulary trainer branch: one section per textbook page. */
+  const vocab = parseArray(u1, 'u1_trans_sections').map(([header, arrayRef]) => {
+    const name = arrayRef?.__c;
+    if (!name) throw new Error(`vocab section '${header}' has no data array`);
+    return {
+      header: String(header),
+      items: parseArray(u1, name).map(([prompt, answers, meaning]) => ({
+        prompt: String(prompt),
+        answers: alternatives(answers),
+        meaning: opt(meaning),
+      })),
+    };
+  });
+
+  return {
+    id: 'unit-01',
+    subject: 'deutsch',
+    number: 1,
+    title: { cs: 'Neue Freunde', en: 'Neue Freunde' },
+    slides: [],
+    tasks,
+    vocab,
+  };
+}