Initial upload — Legesher native code submission portal

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+wheels/legesher_i18n-0.7.3-py3-none-any.whl filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,12 +1,60 @@
 ---
-title: Legesher Native Code
+title: Legesher Native Code Submission
 emoji: 🌍
-colorFrom: pink
-colorTo: blue
+colorFrom: blue
+colorTo: green
 sdk: gradio
 sdk_version: 6.11.0
 app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Native Code Submission Portal
+
+A Gradio web app for collecting natively-written Legesher Python code from native speakers of Chinese, Spanish, and Urdu.
+
+Part of **Language, Decoded — Expedition Tiny Aya** (Condition 4: Strictly Native Code).
+
+---
+
+## What this does
+
+Contributors open the app, pick an exercise, write their solution in their native language using Legesher, run it to confirm it works, then submit. Submissions are saved to a HuggingFace dataset.
+
+The app uses the Legesher Python API (`TokenTranslator`) to transpile native-language code to standard Python for execution — so contributors can see their code actually run in the browser.
+
+---
+
+## Run locally
+
+**Prerequisites:** Python 3.10+
+
+```bash
+cd native-code-submission
+pip install -r requirements.txt
+python app.py
+```
+
+---
+
+## Folder structure
+
+```
+native-code-submission/
+├── app.py              # Main Gradio app
+├── exercises.json      # Exercise prompts (zh/es/ur, tiers 1–3)
+├── requirements.txt    # Python dependencies
+├── references/         # Legesher keyword reference per language
+│   ├── zh.md
+│   ├── es.md
+│   └── ur.md
+└── wheels/             # Bundled Legesher wheels (not on PyPI)
+```
+
+---
+
+## Related issues
+
+- AYA-128 — Exercise design (Done)
+- AYA-129 — Native code criteria (Done) — [criteria doc](https://linear.app/legesher-research/document/native-code-criteria-qualification-disqualification-and-quality-rubric-5705fc83d6e8)
+- AYA-127 — Submission flow design (this)

app.py

@@ -0,0 +1,470 @@
+"""
+Native Code Submission Portal — Language, Decoded / Expedition Tiny Aya
+=======================================================================
+A Gradio app for collecting natively-written Legesher code from
+native speakers of Chinese, Spanish, and Urdu.
+
+Contributors write Python in their native language using Legesher,
+run it to verify it works, then submit it for research use.
+
+HF Space secrets required:
+  HF_TOKEN       — write token for pushing submissions to the dataset repo
+  HF_DATASET_ID  — target dataset repo (e.g. "legesher-research/native-code-submissions")
+"""
+
+import json
+import logging
+import os
+import subprocess
+import sys
+import tempfile
+import uuid
+from datetime import datetime, timezone
+from pathlib import Path
+
+import gradio as gr
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+# ---------------------------------------------------------------------------
+# Legesher setup
+# ---------------------------------------------------------------------------
+
+try:
+    from legesher_core import TokenTranslator
+    LEGESHER_OK = True
+    logger.info("Legesher loaded successfully.")
+except ImportError as e:
+    LEGESHER_OK = False
+    logger.warning(f"Legesher not available: {e}. Run/translate disabled.")
+
+# Cache translators so we don't reload on every button click
+_translators: dict[str, TokenTranslator] = {}
+
+def _get_native_to_en(lang: str) -> TokenTranslator:
+    if lang not in _translators:
+        en_to_native = TokenTranslator.from_language_pack(lang)
+        _translators[lang] = en_to_native.reverse()
+    return _translators[lang]
+
+# ---------------------------------------------------------------------------
+# Exercises
+# ---------------------------------------------------------------------------
+
+EXERCISES_PATH = Path(__file__).parent / "exercises.json"
+with open(EXERCISES_PATH, encoding="utf-8") as _f:
+    _ALL_EXERCISES: dict = json.load(_f)
+
+TIER_KEYS = ["tier1", "tier2", "tier3"]
+TIER_LABELS = ["Tier 1 — Basic (10–20 lines)", "Tier 2 — Applied (20–50 lines)", "Tier 3 — Domain (50–100 lines)"]
+
+LANGUAGES = {
+    "zh": "中文 (Chinese)",
+    "es": "Español (Spanish)",
+    "ur": "اردو (Urdu)",
+}
+
+# Reverse map: display label -> code (handles Gradio 6 returning label instead of value)
+_LABEL_TO_CODE = {v: k for k, v in LANGUAGES.items()}
+
+# ---------------------------------------------------------------------------
+# Keyword reference — loaded from REFERENCE.md files per language
+# ---------------------------------------------------------------------------
+
+_REFERENCES_DIR = Path(__file__).parent / "references"
+
+def load_reference(lang: str) -> str:
+    """Load the Legesher REFERENCE.md for the given language."""
+    path = _REFERENCES_DIR / f"{lang}.md"
+    if path.exists():
+        return path.read_text(encoding="utf-8")
+    return f"_Reference file not found for `{lang}`._"
+
+def normalize_lang(lang_input: str) -> str:
+    """Accept either a language code ('ur') or display label ('اردو (Urdu)') and return the code."""
+    if lang_input in LANGUAGES:
+        return lang_input
+    if lang_input in _LABEL_TO_CODE:
+        return _LABEL_TO_CODE[lang_input]
+    logger.warning(f"Unknown lang input: {lang_input!r}, falling back to zh")
+    return "zh"
+
+def get_exercise_choices(lang: str, tier_idx: int) -> list[str]:
+    tier_key = TIER_KEYS[tier_idx]
+    exercises = _ALL_EXERCISES.get(lang, {}).get(tier_key, [])
+    return [ex["title"] for ex in exercises]
+
+def get_exercise(lang: str, tier_idx: int, title: str) -> dict | None:
+    tier_key = TIER_KEYS[tier_idx]
+    for ex in _ALL_EXERCISES.get(lang, {}).get(tier_key, []):
+        if ex["title"] == title:
+            return ex
+    return None
+
+# ---------------------------------------------------------------------------
+# HuggingFace dataset submission
+# ---------------------------------------------------------------------------
+
+HF_TOKEN = os.environ.get("HF_TOKEN", "")
+HF_DATASET_ID = os.environ.get("HF_DATASET_ID", "legesher-research/native-code-submissions")
+
+def push_submission(row: dict) -> bool:
+    """Push a single submission row to the HF dataset as a JSONL entry."""
+    if not HF_TOKEN:
+        logger.warning("HF_TOKEN not set — submission not saved to HF.")
+        return False
+    try:
+        from huggingface_hub import HfApi
+        api = HfApi(token=HF_TOKEN)
+        filename = f"submissions/{row['language']}/{row['id']}.json"
+        content = json.dumps(row, ensure_ascii=False, indent=2).encode("utf-8")
+        api.upload_file(
+            path_or_fileobj=content,
+            path_in_repo=filename,
+            repo_id=HF_DATASET_ID,
+            repo_type="dataset",
+            commit_message=f"[submission] {row['language']} {row['exercise_id']} {row['id'][:8]}",
+        )
+        return True
+    except Exception as e:
+        logger.error(f"HF upload failed: {e}")
+        return False
+
+# ---------------------------------------------------------------------------
+# Core actions
+# ---------------------------------------------------------------------------
+
+def run_code(code: str, lang: str, stdin: str = "") -> str:
+    """Translate native code to English Python and execute it safely."""
+    lang = normalize_lang(lang)
+    if not code.strip():
+        return "Write some code first."
+    if not LEGESHER_OK:
+        return "Legesher is not available in this environment. Code execution disabled."
+
+    try:
+        translator = _get_native_to_en(lang)
+        english_code = translator.translate_code(code)
+    except Exception as e:
+        return f"Translation error: {e}"
+
+    try:
+        with tempfile.NamedTemporaryFile(
+            mode="w", suffix=".py", delete=False, encoding="utf-8"
+        ) as tmp:
+            tmp.write(english_code)
+            tmp_path = tmp.name
+
+        result = subprocess.run(
+            [sys.executable, "-X", "utf8", tmp_path],
+            input=stdin if stdin.strip() else None,
+            capture_output=True,
+            text=True,
+            timeout=10,
+            encoding="utf-8",
+            env={**os.environ, "PYTHONIOENCODING": "utf-8", "PYTHONUTF8": "1"},
+        )
+        os.unlink(tmp_path)
+
+        out = result.stdout.strip()
+        err = result.stderr.strip()
+        if err:
+            out = (out + "\n\n⚠️ Errors:\n" + err).strip()
+        return out or "(no output — did you call your function at the bottom?)"
+
+    except subprocess.TimeoutExpired:
+        return "Timed out after 10 seconds. Check for infinite loops or missing input."
+    except Exception as e:
+        return f"Execution error: {e}"
+
+
+def submit_code(
+    code: str,
+    lang: str,  # normalized below
+    tier_idx: int,
+    exercise_title: str,
+    time_spent: int,
+    consent: bool,
+) -> str:
+    lang = normalize_lang(lang)
+    if not consent:
+        return "Please tick the consent checkbox before submitting."
+    if not code.strip():
+        return "Code is empty — nothing to submit."
+    if not exercise_title:
+        return "Select an exercise before submitting."
+
+    ex = get_exercise(lang, tier_idx, exercise_title)
+    exercise_id = ex["id"] if ex else f"{lang}-unknown"
+
+    row = {
+        "id": str(uuid.uuid4()),
+        "code": code,
+        "language": lang,
+        "exercise_id": exercise_id,
+        "exercise_title": exercise_title,
+        "tier": tier_idx + 1,
+        "time_spent_minutes": time_spent,
+        "submitted_at": datetime.now(timezone.utc).isoformat(),
+        "consent_given": True,
+        "legesher_version": "0.7.3",
+    }
+
+    saved = push_submission(row)
+    if saved:
+        return f"Submitted! Your contribution ID: `{row['id'][:8]}`. Thank you for contributing to Language, Decoded."
+    else:
+        # Fallback: save locally so nothing is lost
+        fallback_dir = Path(__file__).parent / "local_submissions"
+        fallback_dir.mkdir(exist_ok=True)
+        (fallback_dir / f"{row['id']}.json").write_text(
+            json.dumps(row, ensure_ascii=False, indent=2), encoding="utf-8"
+        )
+        return f"Saved locally (HF upload unavailable). ID: `{row['id'][:8]}`."
+
+# ---------------------------------------------------------------------------
+# UI helpers
+# ---------------------------------------------------------------------------
+
+def update_exercise_dropdown(lang: str, tier_idx: int):
+    choices = get_exercise_choices(lang, tier_idx)
+    value = choices[0] if choices else None
+    return gr.update(choices=choices, value=value)
+
+def update_prompt(lang: str, tier_idx: int, exercise_title: str) -> str:
+    if not exercise_title:
+        return ""
+    ex = get_exercise(lang, tier_idx, exercise_title)
+    if not ex:
+        return ""
+    time_note = f"*Estimated time: ~{ex['time_estimate_min']} minutes*\n\n"
+    return time_note + ex["prompt"]
+
+def update_on_lang_change(lang: str, tier_idx: int):
+    choices = get_exercise_choices(lang, tier_idx)
+    value = choices[0] if choices else None
+    prompt = update_prompt(lang, tier_idx, value) if value else ""
+    return gr.Dropdown(choices=choices, value=value), prompt
+
+# ---------------------------------------------------------------------------
+# Gradio UI
+# ---------------------------------------------------------------------------
+
+HEADER_MD = """
+# Native Code Contribution Portal
+### Language, Decoded — Expedition Tiny Aya
+
+Write Python code in your native language using [Legesher](https://legesher.io) and contribute to our research on native-language programming.
+
+**How it works:**
+1. Select your language and an exercise
+2. Write your solution using Legesher (native keywords + native variable names)
+3. Click **Run** to verify your code works
+4. Fill in the metadata and click **Submit**
+
+---
+"""
+
+CONSENT_TEXT = (
+    "I give permission for my submitted code to be used in the Language, Decoded research project "
+    "and to be published as part of an open dataset (Apache 2.0 / CC-BY-4.0). "
+    "No personal information is collected — submissions are anonymous."
+)
+
+SIDEBAR_MD = """
+### What counts as native code?
+
+- Variables, functions, and classes named in your language
+- Written from scratch — not translated from an English solution
+- Reflects how *you* would naturally think about the problem
+- Uses Legesher keywords throughout
+
+### What doesn't qualify?
+
+- Word-for-word translations of English code
+- AI-generated code
+- English variable names with only keywords swapped
+
+Need help? Read the full [criteria document](https://linear.app/legesher-research/document/native-code-criteria-qualification-disqualification-and-quality-rubric-5705fc83d6e8).
+"""
+
+def build_app() -> gr.Blocks:
+    with gr.Blocks(title="Native Code Submission — Language, Decoded") as app:
+
+        gr.Markdown(HEADER_MD)
+
+        with gr.Row():
+            # ---- Left column: config + exercise ----
+            with gr.Column(scale=1):
+                lang_dropdown = gr.Dropdown(
+                    label="Your language",
+                    choices=[(v, k) for k, v in LANGUAGES.items()],
+                    value="zh",
+                )
+
+                tier_radio = gr.Radio(
+                    label="Exercise tier",
+                    choices=TIER_LABELS,
+                    value=TIER_LABELS[0],
+                )
+
+                exercise_dropdown = gr.Dropdown(
+                    label="Exercise",
+                    choices=get_exercise_choices("zh", 0),
+                    value=get_exercise_choices("zh", 0)[0],
+                )
+
+                exercise_prompt = gr.Markdown(
+                    value=update_prompt("zh", 0, get_exercise_choices("zh", 0)[0]),
+                    label="Exercise prompt",
+                )
+
+                gr.Markdown(SIDEBAR_MD)
+
+                with gr.Accordion("Legesher keyword reference", open=False):
+                    cheatsheet = gr.Markdown(
+                        value=load_reference("zh"),
+                    )
+
+            # ---- Right column: editor + run + submit ----
+            with gr.Column(scale=2):
+                code_editor = gr.Code(
+                    label="Your Legesher code",
+                    language="python",
+                    lines=20,
+                    value="# Write your solution here using Legesher\n# Use native-language variable and function names\n",
+                )
+
+                with gr.Row():
+                    run_btn = gr.Button("Run code", variant="secondary")
+                    clear_btn = gr.Button("Clear", variant="stop")
+
+                stdin_input = gr.Textbox(
+                    label="Program input (stdin) — one value per line",
+                    placeholder="e.g. 15\n(leave blank if your code needs no input)",
+                    lines=2,
+                )
+
+                run_output = gr.Textbox(
+                    label="Output",
+                    lines=6,
+                    placeholder="Run your code to see output here...",
+                    interactive=False,
+                )
+
+                gr.Markdown("---")
+                gr.Markdown("### Submit your solution")
+
+                time_slider = gr.Slider(
+                    label="Time spent (minutes)",
+                    minimum=1,
+                    maximum=120,
+                    step=1,
+                    value=15,
+                )
+
+                consent_checkbox = gr.Checkbox(
+                    label=CONSENT_TEXT,
+                    value=False,
+                )
+
+                submit_btn = gr.Button("Submit contribution", variant="primary")
+
+                submit_status = gr.Textbox(
+                    label="Submission status",
+                    interactive=False,
+                )
+
+        # ---- State ----
+        tier_idx_state = gr.State(value=0)
+        lang_state = gr.State(value="zh")
+
+        # ---- Event wiring ----
+
+        def tier_to_idx(tier_label: str) -> int:
+            return TIER_LABELS.index(tier_label)
+
+        def on_tier_change(tier_label: str, lang: str):
+            try:
+                idx = TIER_LABELS.index(tier_label) if tier_label in TIER_LABELS else 0
+                choices = get_exercise_choices(lang, idx)
+                value = choices[0] if choices else None
+                prompt = update_prompt(lang, idx, value) if value else ""
+                return idx, gr.update(choices=choices, value=value), prompt
+            except Exception as e:
+                logger.error(f"on_tier_change error: {e!r} | tier_label={tier_label!r} lang={lang!r}")
+                return 0, gr.update(), ""
+
+        def on_lang_change(lang_raw: str, tier_label: str):
+            try:
+                lang_code = normalize_lang(lang_raw)
+                logger.info(f"on_lang_change: raw={lang_raw!r} -> code={lang_code!r} tier={tier_label!r}")
+                idx = TIER_LABELS.index(tier_label) if tier_label in TIER_LABELS else 0
+                choices = get_exercise_choices(lang_code, idx)
+                value = choices[0] if choices else None
+                prompt = update_prompt(lang_code, idx, value) if value else ""
+                sheet = load_reference(lang_code)
+                logger.info(f"on_lang_change OK: choices={choices}")
+                return lang_code, gr.update(choices=choices, value=value), prompt, sheet
+            except Exception as e:
+                logger.error(f"on_lang_change error: {e!r} | lang_raw={lang_raw!r}")
+                return "zh", gr.update(), "", load_reference("zh")
+
+        def on_exercise_change(exercise_title: str, lang_raw: str, tier_idx: int):
+            # Use lang_dropdown value directly (always current), not lang_state (can be stale)
+            try:
+                lang_code = normalize_lang(lang_raw)
+                return update_prompt(lang_code, tier_idx, exercise_title)
+            except Exception as e:
+                logger.error(f"on_exercise_change error: {e!r}")
+                return ""
+
+        lang_dropdown.change(
+            on_lang_change,
+            inputs=[lang_dropdown, tier_radio],
+            outputs=[lang_state, exercise_dropdown, exercise_prompt, cheatsheet],
+        )
+
+        tier_radio.change(
+            on_tier_change,
+            inputs=[tier_radio, lang_state],
+            outputs=[tier_idx_state, exercise_dropdown, exercise_prompt],
+        )
+
+        exercise_dropdown.change(
+            on_exercise_change,
+            inputs=[exercise_dropdown, lang_dropdown, tier_idx_state],
+            outputs=[exercise_prompt],
+        )
+
+        run_btn.click(
+            fn=run_code,
+            inputs=[code_editor, lang_state, stdin_input],
+            outputs=[run_output],
+        )
+
+        clear_btn.click(
+            fn=lambda: ("", "", ""),
+            outputs=[code_editor, stdin_input, run_output],
+        )
+
+        submit_btn.click(
+            fn=submit_code,
+            inputs=[
+                code_editor,
+                lang_state,
+                tier_idx_state,
+                exercise_dropdown,
+                time_slider,
+                consent_checkbox,
+            ],
+            outputs=[submit_status],
+        )
+
+    return app
+
+
+if __name__ == "__main__":
+    app = build_app()
+    app.launch(theme=gr.themes.Soft(primary_hue="blue"))

exercises.json

@@ -0,0 +1,170 @@
+{
+  "zh": {
+    "tier1": [
+      {
+        "id": "zh-t1-fizzbuzz",
+        "title": "FizzBuzz",
+        "tier": 1,
+        "prompt": "写一个程序，打印从 1 到 30 的数字。\n\n规则：\n- 如果数字能被 3 整除，打印 \"嘶嘶\" 而不是数字\n- 如果数字能被 5 整除，打印 \"嗡嗡\" 而不是数字\n- 如果数字既能被 3 又能被 5 整除，打印 \"嘶嘶嗡嗡\"\n\n预期输出示例：1, 2, 嘶嘶, 4, 嗡嗡, 嘶嘶, 7, ...",
+        "expected_output": "1\n2\n嘶嘶\n4\n嗡嗡\n嘶嘶\n7\n8\n嘶嘶\n嗡嗡\n11\n嘶嘶\n13\n14\n嘶嘶嗡嗡",
+        "time_estimate_min": 10
+      },
+      {
+        "id": "zh-t1-factorial",
+        "title": "阶乘",
+        "tier": 1,
+        "prompt": "写一个函数，计算一个正整数的阶乘。\n\n例如：\n- 阶乘(5) → 120\n- 阶乘(0) → 1\n- 阶乘(7) → 5040\n\n提示：阶乘(n) = n × (n-1) × ... × 1",
+        "expected_output": "120\n1\n5040",
+        "time_estimate_min": 10
+      },
+      {
+        "id": "zh-t1-palindrome",
+        "title": "回文检测",
+        "tier": 1,
+        "prompt": "写一个函数，检测一个字符串是否是回文（正读反读都一样）。\n\n例如：\n- \"racecar\" → 是回文\n- \"hello\" → 不是回文\n- \"上海自来水来自海上\" → 是回文\n\n忽略大小写和空格。",
+        "expected_output": "是\n否\n是",
+        "time_estimate_min": 15
+      }
+    ],
+    "tier2": [
+      {
+        "id": "zh-t2-calculator",
+        "title": "计算器",
+        "tier": 2,
+        "prompt": "写一个简单的命令行计算器，支持以下功能：\n\n- 加法、减法、乘法、除法\n- 处理除以零的错误\n- 循环接受用户输入，直到用户输入 '退出'\n\n输入格式：\"数字 运算符 数字\"（例如：\"10 + 5\"）",
+        "expected_output": "计算器已启动（输入'退出'结束）\n>>> 10 + 5 = 15",
+        "time_estimate_min": 25
+      },
+      {
+        "id": "zh-t2-word-count",
+        "title": "词频统计",
+        "tier": 2,
+        "prompt": "写一个程序，统计一段文字中每个词出现的频率，并按频率从高到低排序输出前 10 个词。\n\n输入：一段中文或英文文字（可以硬编码在程序中）\n输出：词语及其出现次数，按频率排序",
+        "expected_output": "词频统计结果：\n\"的\": 15次\n\"是\": 12次\n...",
+        "time_estimate_min": 30
+      }
+    ],
+    "tier3": [
+      {
+        "id": "zh-t3-data-analysis",
+        "title": "数据分析脚本",
+        "tier": 3,
+        "prompt": "写一个数据分析脚本，分析一组学生成绩数据。\n\n功能要求：\n1. 定义一组学生数据（姓名 + 成绩列表，至少 10 名学生）\n2. 计算班级平均分、最高分、最低分\n3. 按成绩分组（优秀 90+、良好 75-89、及格 60-74、不及格 <60）\n4. 找出成绩高于平均分的学生\n5. 输出完整报告",
+        "expected_output": "班级成绩分析报告\n=================\n平均分：78.5\n最高分：98（李明）\n最低分：45（王芳）\n...",
+        "time_estimate_min": 45
+      }
+    ]
+  },
+  "es": {
+    "tier1": [
+      {
+        "id": "es-t1-fizzbuzz",
+        "title": "FizzBuzz",
+        "tier": 1,
+        "prompt": "Escribe un programa que imprima los números del 1 al 30.\n\nReglas:\n- Si el número es divisible entre 3, imprime \"Fizz\" en lugar del número\n- Si el número es divisible entre 5, imprime \"Buzz\" en lugar del número\n- Si es divisible entre ambos, imprime \"FizzBuzz\"\n\nEjemplo de salida: 1, 2, Fizz, 4, Buzz, Fizz, 7, ...",
+        "expected_output": "1\n2\nFizz\n4\nBuzz\nFizz\n7\n8\nFizz\nBuzz\n11\nFizz\n13\n14\nFizzBuzz",
+        "time_estimate_min": 10
+      },
+      {
+        "id": "es-t1-factorial",
+        "title": "Factorial",
+        "tier": 1,
+        "prompt": "Escribe una función que calcule el factorial de un número entero positivo.\n\nEjemplos:\n- factorial(5) → 120\n- factorial(0) → 1\n- factorial(7) → 5040\n\nPista: factorial(n) = n × (n-1) × ... × 1",
+        "expected_output": "120\n1\n5040",
+        "time_estimate_min": 10
+      },
+      {
+        "id": "es-t1-palindrome",
+        "title": "Palíndromo",
+        "tier": 1,
+        "prompt": "Escribe una función que verifique si una cadena de texto es un palíndromo (se lee igual de izquierda a derecha que de derecha a izquierda).\n\nEjemplos:\n- \"reconocer\" → es palíndromo\n- \"hola\" → no es palíndromo\n- \"A man a plan a canal Panama\" → es palíndromo\n\nIgnora mayúsculas/minúsculas y espacios.",
+        "expected_output": "Sí\nNo\nSí",
+        "time_estimate_min": 15
+      }
+    ],
+    "tier2": [
+      {
+        "id": "es-t2-calculator",
+        "title": "Calculadora",
+        "tier": 2,
+        "prompt": "Escribe una calculadora de línea de comandos que soporte:\n\n- Suma, resta, multiplicación y división\n- Manejo del error de división por cero\n- Ciclo que acepta entradas hasta que el usuario escriba 'salir'\n\nFormato de entrada: \"número operador número\" (por ejemplo: \"10 + 5\")",
+        "expected_output": "Calculadora iniciada (escribe 'salir' para terminar)\n>>> 10 + 5 = 15",
+        "time_estimate_min": 25
+      },
+      {
+        "id": "es-t2-word-count",
+        "title": "Contador de palabras",
+        "tier": 2,
+        "prompt": "Escribe un programa que cuente la frecuencia de cada palabra en un texto y muestre las 10 más frecuentes ordenadas de mayor a menor.\n\nEntrada: Un texto (puede estar codificado directamente en el programa)\nSalida: Palabras y su conteo, ordenadas por frecuencia",
+        "expected_output": "Resultado del conteo:\n\"el\": 15 veces\n\"la\": 12 veces\n...",
+        "time_estimate_min": 30
+      }
+    ],
+    "tier3": [
+      {
+        "id": "es-t3-data-analysis",
+        "title": "Análisis de datos",
+        "tier": 3,
+        "prompt": "Escribe un script de análisis de datos para calificaciones de estudiantes.\n\nRequisitos:\n1. Define un conjunto de datos de al menos 10 estudiantes (nombre + calificación)\n2. Calcula el promedio, la calificación más alta y la más baja\n3. Agrupa por desempeño (Excelente 90+, Bueno 75-89, Suficiente 60-74, Insuficiente <60)\n4. Identifica estudiantes con calificación superior al promedio\n5. Genera un informe completo",
+        "expected_output": "Informe de calificaciones\n========================\nPromedio: 78.5\nMás alta: 98 (María García)\nMás baja: 45 (Juan López)\n...",
+        "time_estimate_min": 45
+      }
+    ]
+  },
+  "ur": {
+    "tier1": [
+      {
+        "id": "ur-t1-fizzbuzz",
+        "title": "فز بز",
+        "tier": 1,
+        "prompt": "ایک پروگرام لکھیں جو 1 سے 30 تک کے اعداد پرنٹ کرے۔\n\nقواعد:\n- اگر عدد 3 پر پوری طرح تقسیم ہو تو \"فز\" پرنٹ کریں\n- اگر عدد 5 پر پوری طرح تقسیم ہو تو \"بز\" پرنٹ کریں\n- اگر دونوں پر تقسیم ہو تو \"فز بز\" پرنٹ کریں\n\nمثال: 1، 2، فز، 4، بز، فز، 7، ...",
+        "expected_output": "1\n2\nفز\n4\nبز\nفز\n7\n8\nفز\nبز\n11\nفز\n13\n14\nفز بز",
+        "time_estimate_min": 10
+      },
+      {
+        "id": "ur-t1-factorial",
+        "title": "فیکٹوریل",
+        "tier": 1,
+        "prompt": "ایک فنکشن لکھیں جو کسی مثبت پورے عدد کا فیکٹوریل حساب کرے۔\n\nمثالیں:\n- فیکٹوریل(5) ← 120\n- فیکٹوریل(0) ← 1\n- فیکٹوریل(7) ← 5040\n\nیاد رہے: فیکٹوریل(ن) = ن × (ن-1) × ... × 1",
+        "expected_output": "120\n1\n5040",
+        "time_estimate_min": 10
+      },
+      {
+        "id": "ur-t1-palindrome",
+        "title": "پیلنڈروم جانچ",
+        "tier": 1,
+        "prompt": "ایک فنکشن لکھیں جو یہ بتائے کہ کوئی لفظ یا جملہ پیلنڈروم ہے یا نہیں (یعنی آگے پیچھے سے ایک جیسا پڑھا جائے)۔\n\nمثالیں:\n- \"ملم\" ← پیلنڈروم ہے\n- \"سلام\" ← پیلنڈروم نہیں\n- \"racecar\" ← پیلنڈروم ہے\n\nبڑے اور چھوٹے حروف اور خالی جگہوں کو نظرانداز کریں۔",
+        "expected_output": "ہاں\nنہیں\nہاں",
+        "time_estimate_min": 15
+      }
+    ],
+    "tier2": [
+      {
+        "id": "ur-t2-calculator",
+        "title": "حساب کتاب",
+        "tier": 2,
+        "prompt": "ایک کمانڈ لائن کیلکولیٹر لکھیں جو یہ کام کرے:\n\n- جمع، تفریق، ضرب اور تقسیم\n- صفر سے تقسیم کی غلطی کو سنبھالنا\n- لوپ جو 'خروج' لکھنے تک جاری رہے\n\nاندراج کی شکل: \"عدد آپریٹر عدد\" (مثلاً: \"10 + 5\")",
+        "expected_output": "کیلکولیٹر شروع ہو گیا ('خروج' لکھیں بند کرنے کے لیے)\n>>> 10 + 5 = 15",
+        "time_estimate_min": 25
+      },
+      {
+        "id": "ur-t2-word-count",
+        "title": "الفاظ شمار",
+        "tier": 2,
+        "prompt": "ایک پروگرام لکھیں جو کسی متن میں ہر لفظ کتنی بار آیا یہ گنے اور سب سے زیادہ استعمال ہونے والے 10 الفاظ دکھائے۔\n\nاندراج: کوئی بھی متن (پروگرام میں لکھا جا سکتا ہے)\nنتیجہ: الفاظ اور ان کی تعداد، زیادہ سے کم ترتیب میں",
+        "expected_output": "الفاظ شماری کا نتیجہ:\n\"کا\": 15 بار\n\"اور\": 12 بار\n...",
+        "time_estimate_min": 30
+      }
+    ],
+    "tier3": [
+      {
+        "id": "ur-t3-data-analysis",
+        "title": "ڈیٹا تجزیہ",
+        "tier": 3,
+        "prompt": "طلبہ کے نمبروں کے ڈیٹا کا تجزیہ کرنے والا اسکرپٹ لکھیں۔\n\nضروریات:\n1. کم از کم 10 طلبہ کا ڈیٹا تیار کریں (نام + نمبر)\n2. اوسط، سب سے زیادہ اور سب سے کم نمبر حساب کریں\n3. کارکردگی کے مطابق گروپ بنائیں (اعلیٰ 90+، اچھا 75-89، ٹھیک 60-74، کمزور <60)\n4. اوسط سے زیادہ نمبر والے طلبہ کی فہرست بنائیں\n5. مکمل رپورٹ تیار کریں",
+        "expected_output": "طلبہ کی کارکردگی رپورٹ\n====================\nاوسط: 78.5\nسب سے زیادہ: 98 (احمد علی)\nسب سے کم: 45 (سارہ خان)\n...",
+        "time_estimate_min": 45
+      }
+    ]
+  }
+}

references/es.md

@@ -0,0 +1,291 @@
+# Spanish (Español) - Python Quick Reference
+
+![Legesher](https://img.shields.io/badge/Legesher-es-blue?style=for-the-badge&logo=python&logoColor=white)
+![Script](https://img.shields.io/badge/Script-Latin-green?style=for-the-badge)
+![Status](https://img.shields.io/badge/Status-Experimental-yellow?style=for-the-badge)
+
+> Write Python code in Spanish. This guide shows the Spanish keywords
+> to use for Legesher to understand your code.
+
+## Language Details
+
+| Attribute | Value |
+| --------- | ----- |
+| **Language** | Spanish |
+| **Native Name** | Español |
+| **Script** | Latin |
+| **ISO 639-1** | `es` |
+| **ISO 639-2** | `spa` |
+| **BCP 47 Tag** | `es` |
+| **Text Direction** | Left-to-Right (LTR) |
+| **Status** | Experimental |
+
+## Keywords (39 terms)
+
+These are Python's reserved words translated to Spanish.
+
+| English | Spanish |
+| ------- | ------ |
+| `and` | `y` |
+| `as` | `como` |
+| `assert` | `afirmar` |
+| `async` | `asincrónico` |
+| `await` | `esperar` |
+| `break` | `romper` |
+| `case` | `caso` |
+| `class` | `clase` |
+| `continue` | `continuar` |
+| `def` | `definir` |
+| `del` | `eliminar` |
+| `elif` | `si_no` |
+| `else` | `sino` |
+| `except` | `excepto` |
+| `except*` | `excepto*` |
+| `False` | `falso` |
+| `finally` | `finalmente` |
+| `for` | `por` |
+| `from` | `desde` |
+| `global` | `declara_variable_global` |
+| `if` | `si` |
+| `import` | `importar` |
+| `in` | `en` |
+| `is` | `ser` |
+| `lambda` | `funcion_anonima` |
+| `match` | `coincidir` |
+| `None` | `nulo` |
+| `nonlocal` | `nolocal` |
+| `not` | `no` |
+| `or` | `o` |
+| `pass` | `pasar` |
+| `raise` | `levantar` |
+| `return` | `devolver` |
+| `True` | `verdadero` |
+| `try` | `probar` |
+| `type` | `tipo` |
+| `while` | `mientras` |
+| `with` | `con` |
+| `yield` | `dar` |
+
+## Built-in Functions (72 terms)
+
+Python's built-in functions and types in Spanish.
+
+| English | Spanish |
+| ------- | ------ |
+| `abs` | `valorabsoluto` |
+| `aiter` | `iterador_asincrónico` |
+| `all` | `todo` |
+| `anext` | `próximo` |
+| `any` | `cualquier` |
+| `ascii` | `asciiificar` |
+| `bin` | `binario` |
+| `bool` | `booleano` |
+| `breakpoint` | `puntodeinterrupción` |
+| `bytearray` | `conjuntoDeBytes` |
+| `bytes` | `bytes_inmutables` |
+| `callable` | `llamable` |
+| `chr` | `carácter` |
+| `classmethod` | `métododeclase` |
+| `compile` | `compilar` |
+| `complex` | `complejo` |
+| `delattr` | `eliminar_atributo` |
+| `dict` | `diccionario` |
+| `dir` | `directorio` |
+| `divmod` | `division_con_resta` |
+| `ellipsis` | `elipsis` |
+| `enumerate` | `elenumerar` |
+| `eval` | `evalúo` |
+| `exec` | `ejecutar` |
+| `filter` | `filtrar` |
+| `float` | `flotante` |
+| `format` | `formatear` |
+| `frozenset` | `conjuntocongelado` |
+| `getattr` | `obtenerelatributo` |
+| `globals` | `globales` |
+| `hasattr` | `teneratributo` |
+| `hash` | `resumen_hash` |
+| `help` | `ayuda` |
+| `hex` | `hexadecimal` |
+| `id` | `identificador` |
+| `input` | `entrada` |
+| `int` | `entero` |
+| `isinstance` | `ser_de_la_clase` |
+| `issubclass` | `subclasede` |
+| `iter` | `iterar` |
+| `len` | `longitud` |
+| `list` | `lista` |
+| `locals` | `loslocales` |
+| `map` | `mapear` |
+| `max` | `máximo` |
+| `memoryview` | `vistadememoria` |
+| `min` | `minimo` |
+| `next` | `siguiente` |
+| `notimplemented` | `noimplementado` |
+| `object` | `objeto` |
+| `oct` | `octal` |
+| `open` | `abrir` |
+| `ord` | `ordinal` |
+| `pow` | `potencia` |
+| `print` | `imprimir` |
+| `property` | `propiedad` |
+| `range` | `rango` |
+| `repr` | `representación` |
+| `reversed` | `revesado` |
+| `round` | `redondear` |
+| `set` | `conjunto` |
+| `setattr` | `asignar` |
+| `slice` | `rebanada` |
+| `sorted` | `ordenado` |
+| `staticmethod` | `estático` |
+| `str` | `cadena` |
+| `sum` | `sumar` |
+| `super` | `superClase` |
+| `tuple` | `tupla` |
+| `type` | `tipo` |
+| `vars` | `variables` |
+| `zip` | `acopiar` |
+
+## Exceptions (69 terms)
+
+Exception classes for error handling.
+
+| English | Spanish |
+| ------- | ------ |
+| `arithmeticerror` | `erroraritmético` |
+| `assertionerror` | `errordeaserción` |
+| `attributeerror` | `Errordeatributo` |
+| `baseexception` | `excepciónbase` |
+| `baseexceptiongroup` | `grupodeexcepcionesbase` |
+| `basegeneratorexit` | `generarsalida` |
+| `blockingioerror` | `errordeESbloqueante` |
+| `brokenpipeerror` | `errordetuberíarota` |
+| `buffererror` | `errordebúfer` |
+| `byteswarning` | `advertenciadebytes` |
+| `childprocesserror` | `errordeprocesosecundario` |
+| `connectionabortederror` | `errordeabortarconexión` |
+| `connectionerror` | `errordeconexión` |
+| `connectionrefusederror` | `errorderechazodeconexión` |
+| `connectionreseterror` | `resetearconexión` |
+| `deprecationwarning` | `advertenciadedepreciación` |
+| `encodingwarning` | `advertenciadecodificación` |
+| `environmenterror` | `errordeentorno` |
+| `eoferror` | `errordeentradasalida` |
+| `exception` | `excepción` |
+| `exceptiongroup` | `grupodeexcepciones` |
+| `fileexistserror` | `errordeexistenciadearchivo` |
+| `filenotfounderror` | `Errordenoencontrarelarchivo` |
+| `floatingpointerror` | `errordepunteroflotante` |
+| `futurewarning` | `advertenciafutura` |
+| `generatorexit` | `salidadegenerador` |
+| `importerror` | `errordeimportación` |
+| `importwarning` | `advertenciadeimportación` |
+| `indentationerror` | `errordesangría` |
+| `indexerror` | `Errordeíndice` |
+| `interruptederror` | `errordeinterrupción` |
+| `ioerror` | `errordeES` |
+| `isadirectoryerror` | `errordedirectorio` |
+| `keyboardinterrupt` | `interrupcióndelteclado` |
+| `keyerror` | `errordeclave` |
+| `lookuperror` | `errordebúsqueda` |
+| `memoryerror` | `errordememoria` |
+| `modulenotfounderror` | `Errordemódulonoencontrado` |
+| `nameerror` | `Errordenombre` |
+| `notadirectoryerror` | `errordenoserundirectorio` |
+| `notimplementederror` | `errordenoimplementación` |
+| `oserror` | `ErrorDeOS` |
+| `overflowerror` | `errordedesbordamiento` |
+| `pendingdeprecationwarning` | `advertenciadeadvertenciadedepreciaciónpendiente` |
+| `permissionerror` | `errordepermiso` |
+| `processlookuperror` | `errordebúsquedadeproceso` |
+| `recursionerror` | `errorderecursión` |
+| `referenceerror` | `errordereferencia` |
+| `resourcewarning` | `advertenciaderecursos` |
+| `runtimeerror` | `errorentiempodeejecución` |
+| `runtimewarning` | `advertenciadetiempodeejecución` |
+| `stopasynciteration` | `detenerlaiteraciónasincrónica` |
+| `stopiteration` | `detener_iteración` |
+| `syntaxerror` | `errordesintaxis` |
+| `syntaxwarning` | `advertenciadesintaxis` |
+| `systemexit` | `salirdelsistema` |
+| `taberror` | `errordetabulación` |
+| `timeouterror` | `errordetiempodeespera` |
+| `typeerror` | `errordetipo` |
+| `unboundlocalerror` | `errordereferencialocalnoenlazada` |
+| `unicodedecodeerror` | `errordedecodificacióndeUnicode` |
+| `unicodeencodeerror` | `errordecodificaciónunicode` |
+| `unicodeerror` | `errordeunicode` |
+| `unicodetranslateerror` | `errordetraduccióndeUnicode` |
+| `unicodewarning` | `advertenciadeunicode` |
+| `userwarning` | `advertenciadelusuario` |
+| `valueerror` | `errordevalor` |
+| `warning` | `advertencia` |
+| `zerodivisionerror` | `errordedivisiónporcero` |
+
+## Example Code
+
+Here's an example showing Spanish Python code with Legesher:
+
+```python
+# Ejemplo: Función factorial con manejo de errores
+
+desde typing importar Optional
+
+definir factorial(n):
+    """Calcular el factorial de n."""
+    si n < 0:
+        levantar Errordevalor("Los números negativos no están permitidos")
+    si_no n == 0 o n == 1:
+        devolver 1
+    sino:
+        devolver n * factorial(n - 1)
+
+definir principal():
+    numeros = [entero(x) por x en rango(6)]
+
+    por numero en numeros:
+        resultado = factorial(numero)
+        imprimir(f"{numero}! = {resultado}")
+
+    # Demostrar bucle while
+    contador = 0
+    mientras contador < 3:
+        imprimir(f"Contador: {contador}")
+        contador += 1
+
+    # Demostrar try/except
+    probar:
+        factorial(-1)
+    excepto Errordevalor como error:
+        imprimir(f"Error: {error}")
+    finalmente:
+        imprimir("¡Listo!")
+
+si __name__ == "__main__":
+    principal()
+```
+
+## Usage
+
+```python
+from legesher_i18n import LanguagePackLoader
+
+# Load Spanish pack for Python 3.12
+loader = LanguagePackLoader()
+pack = loader.load("es", "python", "3.12")
+
+# Translate code
+from legesher import translate
+english_code = translate(your_es_code, from_lang="es", to_lang="en")
+```
+
+## Contributing
+
+Found a better translation? We welcome contributions from native speakers!
+
+1. Edit `legesher_i18n_python_es/packs/<version>.yml` for the Python version where the term was introduced:
+   - `3.10.yml` - Base translations (most terms)
+   - `3.11.yml` - Terms added in Python 3.11 (e.g., `except*`)
+   - `3.12.yml` - Terms added in Python 3.12 (e.g., `type`)
+   - `3.13.yml` / `3.14.yml` - Future additions
+2. Submit a pull request with your reasoning
+3. Translations should be natural, commonly used in CS education, and easy to type

references/ur.md

@@ -0,0 +1,294 @@
+# Urdu (اردو) - Python Quick Reference
+
+![Legesher](https://img.shields.io/badge/Legesher-ur-blue?style=for-the-badge&logo=python&logoColor=white)
+![Script](https://img.shields.io/badge/Script-Arabic-green?style=for-the-badge)
+![Status](https://img.shields.io/badge/Status-Experimental-yellow?style=for-the-badge)
+![RTL](https://img.shields.io/badge/Direction-RTL-orange?style=for-the-badge)
+
+> Write Python code in Urdu. This guide shows the Urdu keywords
+> to use for Legesher to understand your code.
+>
+> **Note:** This is a right-to-left (RTL) language. RTL support is currently a work in progress — we welcome your feedback!
+
+## Language Details
+
+| Attribute | Value |
+| --------- | ----- |
+| **Language** | Urdu |
+| **Native Name** | اردو |
+| **Script** | Arabic |
+| **ISO 639-1** | `ur` |
+| **ISO 639-2** | `urd` |
+| **BCP 47 Tag** | `ur` |
+| **Text Direction** | Right-to-Left (RTL) |
+| **Status** | Experimental |
+
+## Keywords (39 terms)
+
+These are Python's reserved words translated to Urdu.
+
+| English | Urdu |
+| ------- | ------ |
+| `and` | `اور` |
+| `as` | `بطور` |
+| `assert` | `تصدیق` |
+| `async` | `غیرہمزمان` |
+| `await` | `انتظار` |
+| `break` | `توڑ` |
+| `case` | `صورت` |
+| `class` | `جماعت` |
+| `continue` | `جاری` |
+| `def` | `تعریف` |
+| `del` | `مٹاؤ` |
+| `elif` | `ورنہاگر` |
+| `else` | `ورنہ` |
+| `except` | `سوائے` |
+| `except*` | `سوائے*` |
+| `False` | `غلط` |
+| `finally` | `بالآخر` |
+| `for` | `لیے` |
+| `from` | `سے` |
+| `global` | `عالمی` |
+| `if` | `اگر` |
+| `import` | `درآمد` |
+| `in` | `میں` |
+| `is` | `ہے` |
+| `lambda` | `لیمڈا` |
+| `match` | `مماثلت` |
+| `None` | `کچھنہیں` |
+| `nonlocal` | `غیرمقامی` |
+| `not` | `نہیں` |
+| `or` | `یا` |
+| `pass` | `گزر` |
+| `raise` | `اٹھاؤ` |
+| `return` | `واپسی` |
+| `True` | `صحیح` |
+| `try` | `کوشش` |
+| `type` | `نوعیت` |
+| `while` | `جبتک` |
+| `with` | `ساتھ` |
+| `yield` | `حاصل` |
+
+## Built-in Functions (72 terms)
+
+Python's built-in functions and types in Urdu.
+
+| English | Urdu |
+| ------- | ------ |
+| `abs` | `مطلق` |
+| `aiter` | `غیرہمزمانتکرارکنندہ` |
+| `all` | `تمام` |
+| `anext` | `غیرہمزماناگلا` |
+| `any` | `کوئی` |
+| `ascii` | `آسکی` |
+| `bin` | `ثنائی` |
+| `bool` | `منطقی` |
+| `breakpoint` | `وقفہ` |
+| `bytearray` | `بائٹصف` |
+| `bytes` | `بائٹس` |
+| `callable` | `قابلِاستعمال` |
+| `chr` | `حرف` |
+| `classmethod` | `جماعتطریقہ` |
+| `compile` | `تالیف` |
+| `complex` | `مرکب` |
+| `delattr` | `خاصیت_مٹاؤ` |
+| `dict` | `لغت` |
+| `dir` | `فہرست` |
+| `divmod` | `تقسیمباقی` |
+| `ellipsis` | `حذف` |
+| `enumerate` | `شمار` |
+| `eval` | `جانچ` |
+| `exec` | `عمل` |
+| `filter` | `چھاننا` |
+| `float` | `اعشاری` |
+| `format` | `شکل` |
+| `frozenset` | `منجمدمجموعہ` |
+| `getattr` | `خاصیت_لاؤ` |
+| `globals` | `عالمیمتغیرات` |
+| `hasattr` | `خاصیت_ہے` |
+| `hash` | `ہیش` |
+| `help` | `مدد` |
+| `hex` | `سداسی` |
+| `id` | `شناخت` |
+| `input` | `داخلہ` |
+| `int` | `عددصحیح` |
+| `isinstance` | `مثال_ہے` |
+| `issubclass` | `ذیلی_جماعت_ہے` |
+| `iter` | `تکرارکنندہ` |
+| `len` | `لمبائی` |
+| `list` | `فہرستی` |
+| `locals` | `مقامیمتغیرات` |
+| `map` | `نقشہ` |
+| `max` | `زیادہ` |
+| `memoryview` | `یادداشتنظارہ` |
+| `min` | `کمترین` |
+| `next` | `اگلا` |
+| `notimplemented` | `غیرنافذ` |
+| `object` | `شے` |
+| `oct` | `ہشتی` |
+| `open` | `کھولو` |
+| `ord` | `ترتیب` |
+| `pow` | `قوت` |
+| `print` | `چھاپو` |
+| `property` | `خاصیت` |
+| `range` | `حدود` |
+| `repr` | `نمائندگی` |
+| `reversed` | `الٹا` |
+| `round` | `گول` |
+| `set` | `مجموعہ` |
+| `setattr` | `خاصیت_مقرر` |
+| `slice` | `ٹکڑا` |
+| `sorted` | `ترتیبشدہ` |
+| `staticmethod` | `جامدطریقہ` |
+| `str` | `متن` |
+| `sum` | `کل` |
+| `super` | `بالائی` |
+| `tuple` | `ٹپل` |
+| `type` | `نوعیت` |
+| `vars` | `متغیرات_لاؤ` |
+| `zip` | `جوڑا` |
+
+## Exceptions (69 terms)
+
+Exception classes for error handling.
+
+| English | Urdu |
+| ------- | ------ |
+| `arithmeticerror` | `حسابیغلطی` |
+| `assertionerror` | `تصدیقغلطی` |
+| `attributeerror` | `خاصیتغلطی` |
+| `baseexception` | `بنیادیاستثنا` |
+| `baseexceptiongroup` | `بنیادیاستثناگروہ` |
+| `basegeneratorexit` | `بنیادیجنریٹرخروج` |
+| `blockingioerror` | `مسدودداخلہخارجہغلطی` |
+| `brokenpipeerror` | `ٹوٹاپائپغلطی` |
+| `buffererror` | `بفرغلطی` |
+| `byteswarning` | `بائٹسانتباہ` |
+| `childprocesserror` | `ذیلیعملغلطی` |
+| `connectionabortederror` | `رابطہمنسوخغلطی` |
+| `connectionerror` | `رابطہغلطی` |
+| `connectionrefusederror` | `رابطہمستردغلطی` |
+| `connectionreseterror` | `رابطہدوبارہغلطی` |
+| `deprecationwarning` | `متروکانتباہ` |
+| `encodingwarning` | `ضابطہبندیانتباہ` |
+| `environmenterror` | `ماحولغلطی` |
+| `eoferror` | `فائلختمغلطی` |
+| `exception` | `استثنا` |
+| `exceptiongroup` | `استثناگروہ` |
+| `fileexistserror` | `فائلموجودغلطی` |
+| `filenotfounderror` | `فائلنہملیغلطی` |
+| `floatingpointerror` | `اعشاریاشارہغلطی` |
+| `futurewarning` | `مستقبلانتباہ` |
+| `generatorexit` | `جنریٹرخروج` |
+| `importerror` | `درآمدغلطی` |
+| `importwarning` | `درآمدانتباہ` |
+| `indentationerror` | `حاشیہغلطی` |
+| `indexerror` | `اشاریہغلطی` |
+| `interruptederror` | `رکاوٹغلطی` |
+| `ioerror` | `داخلہخارجہغلطی` |
+| `isadirectoryerror` | `ڈائریکٹریہےغلطی` |
+| `keyboardinterrupt` | `کلیدتختہرکاوٹ` |
+| `keyerror` | `کلیدغلطی` |
+| `lookuperror` | `تلاشغلطی` |
+| `memoryerror` | `یادداشتغلطی` |
+| `modulenotfounderror` | `ماڈیولنہملاغلطی` |
+| `nameerror` | `نامغلطی` |
+| `notadirectoryerror` | `ڈائریکٹرینہیںغلطی` |
+| `notimplementederror` | `غیرنافذغلطی` |
+| `oserror` | `آپریٹنگنظامغلطی` |
+| `overflowerror` | `اوورفلوغلطی` |
+| `pendingdeprecationwarning` | `زیرالتواءمتروکانتباہ` |
+| `permissionerror` | `اجازتغلطی` |
+| `processlookuperror` | `عملتلاشغلطی` |
+| `recursionerror` | `تکرارغلطی` |
+| `referenceerror` | `حوالہغلطی` |
+| `resourcewarning` | `وسائلانتباہ` |
+| `runtimeerror` | `وقتعملغلطی` |
+| `runtimewarning` | `وقتعملانتباہ` |
+| `stopasynciteration` | `غیرہمزمانتکراررُک` |
+| `stopiteration` | `تکراررُک` |
+| `syntaxerror` | `نحوغلطی` |
+| `syntaxwarning` | `نحوانتباہ` |
+| `systemexit` | `نظامخروج` |
+| `taberror` | `ٹیبغلطی` |
+| `timeouterror` | `وقتختمغلطی` |
+| `typeerror` | `قسم_غلطی` |
+| `unboundlocalerror` | `غیرمربوطمقامیغلطی` |
+| `unicodedecodeerror` | `یونیکوڈضابطہکشائیغلطی` |
+| `unicodeencodeerror` | `یونیکوڈضابطہبندیغلطی` |
+| `unicodeerror` | `یونیکوڈغلطی` |
+| `unicodetranslateerror` | `یونیکوڈترجمہغلطی` |
+| `unicodewarning` | `یونیکوڈانتباہ` |
+| `userwarning` | `صارفاطلاع` |
+| `valueerror` | `قدرغلطی` |
+| `warning` | `انتباہ` |
+| `zerodivisionerror` | `صفرتقسیمغلطی` |
+
+## Example Code
+
+Here's an example showing Urdu Python code with Legesher:
+
+```python
+# مثال: غلطی کی ہینڈلنگ کے ساتھ فیکٹوریل فنکشن
+
+سے typing درآمد Optional
+
+تعریف فیکٹوریل(n):
+    """عدد ن کا فیکٹوریل حساب کریں۔"""
+    اگر n < 0:
+        اٹھاؤ قدرغلطی("منفی اعداد کی اجازت نہیں")
+    ورنہاگر n == 0 یا n == 1:
+        واپسی 1
+    ورنہ:
+        واپسی n * فیکٹوریل(n - 1)
+
+تعریف مرکزی():
+    اعداد = [عددصحیح(x) لیے x میں حدود(6)]
+
+    لیے عدد میں اعداد:
+        نتیجہ = فیکٹوریل(عدد)
+        چھاپو(f"{عدد}! = {نتیجہ}")
+
+    # جبتک لوپ کا مظاہرہ
+    گنتی = 0
+    جبتک گنتی < 3:
+        چھاپو(f"گنتی: {گنتی}")
+        گنتی += 1
+
+    # کوشش/سوائے کا مظاہرہ
+    کوشش:
+        فیکٹوریل(-1)
+    سوائے قدرغلطی بطور غلطی:
+        چھاپو(f"غلطی: {غلطی}")
+    بالآخر:
+        چھاپو("مکمل!")
+
+اگر __name__ == "__main__":
+    مرکزی()
+```
+
+## Usage
+
+```python
+from legesher_i18n import LanguagePackLoader
+
+# Load Urdu pack for Python 3.12
+loader = LanguagePackLoader()
+pack = loader.load("ur", "python", "3.12")
+
+# Translate code
+from legesher import translate
+english_code = translate(your_ur_code, from_lang="ur", to_lang="en")
+```
+
+## Contributing
+
+Found a better translation? We welcome contributions from native speakers!
+
+1. Edit `legesher_i18n_python_ur/packs/<version>.yml` for the Python version where the term was introduced:
+   - `3.10.yml` - Base translations (most terms)
+   - `3.11.yml` - Terms added in Python 3.11 (e.g., `except*`)
+   - `3.12.yml` - Terms added in Python 3.12 (e.g., `type`)
+   - `3.13.yml` / `3.14.yml` - Future additions
+2. Submit a pull request with your reasoning
+3. Translations should be natural, commonly used in CS education, and easy to type

references/zh.md

@@ -0,0 +1,291 @@
+# Chinese (中文) - Python Quick Reference
+
+![Legesher](https://img.shields.io/badge/Legesher-zh-blue?style=for-the-badge&logo=python&logoColor=white)
+![Script](https://img.shields.io/badge/Script-Simplified%20Chinese%20%28Han%29-green?style=for-the-badge)
+![Status](https://img.shields.io/badge/Status-Experimental-yellow?style=for-the-badge)
+
+> Write Python code in Chinese. This guide shows the Chinese keywords
+> to use for Legesher to understand your code.
+
+## Language Details
+
+| Attribute | Value |
+| --------- | ----- |
+| **Language** | Chinese |
+| **Native Name** | 中文 |
+| **Script** | Simplified Chinese (Han) |
+| **ISO 639-1** | `zh` |
+| **ISO 639-2** | `zho` |
+| **BCP 47 Tag** | `zh-Hans` |
+| **Text Direction** | Left-to-Right (LTR) |
+| **Status** | Experimental |
+
+## Keywords (39 terms)
+
+These are Python's reserved words translated to Chinese.
+
+| English | Chinese |
+| ------- | ------ |
+| `and` | `和` |
+| `as` | `作为` |
+| `assert` | `断言` |
+| `async` | `异步` |
+| `await` | `等待` |
+| `break` | `终止` |
+| `case` | `情况` |
+| `class` | `类` |
+| `continue` | `继续` |
+| `def` | `定义` |
+| `del` | `删除` |
+| `elif` | `否则如果` |
+| `else` | `否则` |
+| `except` | `除外` |
+| `except*` | `多重捕获*` |
+| `False` | `错误` |
+| `finally` | `最后` |
+| `for` | `对于` |
+| `from` | `从` |
+| `global` | `全球` |
+| `if` | `如果` |
+| `import` | `导入` |
+| `in` | `在` |
+| `is` | `是` |
+| `lambda` | `拉姆达` |
+| `match` | `匹配` |
+| `None` | `无` |
+| `nonlocal` | `非本地` |
+| `not` | `不` |
+| `or` | `或` |
+| `pass` | `过` |
+| `raise` | `提出` |
+| `return` | `返回` |
+| `True` | `真实` |
+| `try` | `尝试` |
+| `type` | `类型` |
+| `while` | `当` |
+| `with` | `跟` |
+| `yield` | `产生` |
+
+## Built-in Functions (72 terms)
+
+Python's built-in functions and types in Chinese.
+
+| English | Chinese |
+| ------- | ------ |
+| `abs` | `绝对值` |
+| `aiter` | `异步迭代器` |
+| `all` | `全` |
+| `anext` | `下一项` |
+| `any` | `任意` |
+| `ascii` | `ASCII` |
+| `bin` | `二进制` |
+| `bool` | `布尔` |
+| `breakpoint` | `断点` |
+| `bytearray` | `字节数组` |
+| `bytes` | `字节` |
+| `callable` | `可调用` |
+| `chr` | `字符` |
+| `classmethod` | `类方法` |
+| `compile` | `编译` |
+| `complex` | `复数` |
+| `delattr` | `删除属性` |
+| `dict` | `字典` |
+| `dir` | `列出属性和方法` |
+| `divmod` | `除模` |
+| `ellipsis` | `省略号` |
+| `enumerate` | `枚举` |
+| `eval` | `评估` |
+| `exec` | `执行` |
+| `filter` | `过滤` |
+| `float` | `浮点数` |
+| `format` | `格式化` |
+| `frozenset` | `冻结集` |
+| `getattr` | `获取属性` |
+| `globals` | `全局变量` |
+| `hasattr` | `具有` |
+| `hash` | `哈希` |
+| `help` | `帮助` |
+| `hex` | `十六进制` |
+| `id` | `标识` |
+| `input` | `输入` |
+| `int` | `整数` |
+| `isinstance` | `是实例` |
+| `issubclass` | `是子类` |
+| `iter` | `迭代` |
+| `len` | `长度` |
+| `list` | `列表` |
+| `locals` | `局部变量` |
+| `map` | `映射` |
+| `max` | `最大值` |
+| `memoryview` | `内存视图` |
+| `min` | `最小值` |
+| `next` | `下一个` |
+| `notimplemented` | `未实现` |
+| `object` | `对象` |
+| `oct` | `八进制` |
+| `open` | `打开` |
+| `ord` | `顺序` |
+| `pow` | `幂` |
+| `print` | `打印` |
+| `property` | `属性` |
+| `range` | `范围` |
+| `repr` | `表示` |
+| `reversed` | `反转` |
+| `round` | `四舍五入` |
+| `set` | `集合` |
+| `setattr` | `设置属性` |
+| `slice` | `切片` |
+| `sorted` | `排序` |
+| `staticmethod` | `静态方法` |
+| `str` | `字符串` |
+| `sum` | `求和` |
+| `super` | `超类` |
+| `tuple` | `元组` |
+| `type` | `类型` |
+| `vars` | `变量字典` |
+| `zip` | `压缩` |
+
+## Exceptions (69 terms)
+
+Exception classes for error handling.
+
+| English | Chinese |
+| ------- | ------ |
+| `arithmeticerror` | `算术错误` |
+| `assertionerror` | `断言错误` |
+| `attributeerror` | `属性错误` |
+| `baseexception` | `基本异常` |
+| `baseexceptiongroup` | `基异常组` |
+| `basegeneratorexit` | `基底生成器退出` |
+| `blockingioerror` | `阻塞的IO错误` |
+| `brokenpipeerror` | `破管错误` |
+| `buffererror` | `缓冲区错误` |
+| `byteswarning` | `字节警告` |
+| `childprocesserror` | `子进程错误` |
+| `connectionabortederror` | `连接中断错误` |
+| `connectionerror` | `连接错误` |
+| `connectionrefusederror` | `连接拒绝错误` |
+| `connectionreseterror` | `连接重置错误` |
+| `deprecationwarning` | `弃用警告` |
+| `encodingwarning` | `编码警告` |
+| `environmenterror` | `环境错误` |
+| `eoferror` | `文件结束错误` |
+| `exception` | `异常` |
+| `exceptiongroup` | `异常组` |
+| `fileexistserror` | `文件已存在错误` |
+| `filenotfounderror` | `文件未找到错误` |
+| `floatingpointerror` | `浮点数错误` |
+| `futurewarning` | `未来警告` |
+| `generatorexit` | `生成器退出` |
+| `importerror` | `导入错误` |
+| `importwarning` | `导入警告` |
+| `indentationerror` | `缩进错误` |
+| `indexerror` | `越界访问错误` |
+| `interruptederror` | `中断错误` |
+| `ioerror` | `IO错误` |
+| `isadirectoryerror` | `是目录错误` |
+| `keyboardinterrupt` | `键盘中断` |
+| `keyerror` | `键错误` |
+| `lookuperror` | `查找错误` |
+| `memoryerror` | `内存错误` |
+| `modulenotfounderror` | `模块未找到错误` |
+| `nameerror` | `名称错误` |
+| `notadirectoryerror` | `非目录错误` |
+| `notimplementederror` | `未实现错误` |
+| `oserror` | `操作系统错误` |
+| `overflowerror` | `溢出错误` |
+| `pendingdeprecationwarning` | `待删除警告` |
+| `permissionerror` | `权限错误` |
+| `processlookuperror` | `进程查找错误` |
+| `recursionerror` | `递归错误` |
+| `referenceerror` | `引用错误` |
+| `resourcewarning` | `资源警告` |
+| `runtimeerror` | `运行时错误` |
+| `runtimewarning` | `运行时警告` |
+| `stopasynciteration` | `停止异步迭代` |
+| `stopiteration` | `停止迭代` |
+| `syntaxerror` | `语法错误` |
+| `syntaxwarning` | `语法警告` |
+| `systemexit` | `系统退出` |
+| `taberror` | `制表符错误` |
+| `timeouterror` | `超时错误` |
+| `typeerror` | `类型错误` |
+| `unboundlocalerror` | `未绑定局部错误` |
+| `unicodedecodeerror` | `统一码解码错误` |
+| `unicodeencodeerror` | `统一码编码错误` |
+| `unicodeerror` | `统一码错误` |
+| `unicodetranslateerror` | `统一码转换错误` |
+| `unicodewarning` | `统一码警告` |
+| `userwarning` | `用户警告` |
+| `valueerror` | `值错误` |
+| `warning` | `警告` |
+| `zerodivisionerror` | `零除数错误` |
+
+## Example Code
+
+Here's an example showing Chinese Python code with Legesher:
+
+```python
+# 示例：带错误处理的阶乘函数
+
+从 typing 导入 Optional
+
+定义 阶乘(n):
+    """计算n的阶乘。"""
+    如果 n < 0:
+        提出 值错误("不允许负数")
+    否则如果 n == 0 或 n == 1:
+        返回 1
+    否则:
+        返回 n * 阶乘(n - 1)
+
+定义 主函数():
+    数字列表 = [整数(x) 对于 x 在 范围(6)]
+
+    对于 数字 在 数字列表:
+        结果 = 阶乘(数字)
+        打印(f"{数字}! = {结果}")
+
+    # 演示while循环
+    计数 = 0
+    当 计数 < 3:
+        打印(f"计数： {计数}")
+        计数 += 1
+
+    # 演示try/except
+    尝试:
+        阶乘(-1)
+    除外 值错误 作为 错误:
+        打印(f"错误： {错误}")
+    最后:
+        打印("完成！")
+
+如果 __name__ == "__main__":
+    主函数()
+```
+
+## Usage
+
+```python
+from legesher_i18n import LanguagePackLoader
+
+# Load Chinese pack for Python 3.12
+loader = LanguagePackLoader()
+pack = loader.load("zh", "python", "3.12")
+
+# Translate code
+from legesher import translate
+english_code = translate(your_zh_code, from_lang="zh", to_lang="en")
+```
+
+## Contributing
+
+Found a better translation? We welcome contributions from native speakers!
+
+1. Edit `legesher_i18n_python_zh/packs/<version>.yml` for the Python version where the term was introduced:
+   - `3.10.yml` - Base translations (most terms)
+   - `3.11.yml` - Terms added in Python 3.11 (e.g., `except*`)
+   - `3.12.yml` - Terms added in Python 3.12 (e.g., `type`)
+   - `3.13.yml` / `3.14.yml` - Future additions
+2. Submit a pull request with your reasoning
+3. Translations should be natural, commonly used in CS education, and easy to type

requirements.txt

@@ -0,0 +1,14 @@
+gradio>=4.44.0
+huggingface_hub>=0.24.0
+datasets>=2.20.0
+tree-sitter>=0.20.4
+tree-sitter-python>=0.21.0
+
+# Legesher packages (bundled in wheels/)
+./wheels/legesher_core-0.7.3-py3-none-any.whl
+./wheels/legesher_i18n-0.7.3-py3-none-any.whl
+./wheels/legesher-0.7.3-py3-none-any.whl
+./wheels/legesher_i18n_python_en-0.0.2-py3-none-any.whl
+./wheels/legesher_i18n_python_zh-0.0.4-py3-none-any.whl
+./wheels/legesher_i18n_python_es-0.0.3-py3-none-any.whl
+./wheels/legesher_i18n_python_ur-0.0.2-py3-none-any.whl

wheels/legesher_i18n-0.7.3-py3-none-any.whl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:227588a0e1ce7a3ba9a049a8999debb18b214b70c1027ebb4f060ae6ac11f0c5
+size 1854709