cwicr-multilingual

CWICR の多言語サポート

概要

CWICR データベースは、統一された作業項目コードで 26 言語に対応しています。このスキルにより、言語をまたいだ作業項目の照合、翻訳、地域別価格の比較が可能になります。

対応言語

コード	言語	地域	通貨
AR	アラビア語	ドバイ	AED
DE	ドイツ語	ベルリン	EUR
EN	英語	トロント	CAD
ES	スペイン語	バルセロナ	EUR
FR	フランス語	パリ	EUR
HI	ヒンディー語	ムンバイ	INR
PT	ポルトガル語	サンパウロ	BRL
RU	ロシア語	サンクトペテルブルク	RUB
ZH	中国語	上海	CNY

Python による実装

import pandas as pd
from typing import Dict, Any, List, Optional, Tuple
from dataclasses import dataclass
from enum import Enum


class CWICRLanguage(Enum):
    """Supported CWICR languages."""
    ARABIC = ("ar", "Arabic", "AED", "Dubai")
    GERMAN = ("de", "German", "EUR", "Berlin")
    ENGLISH = ("en", "English", "CAD", "Toronto")
    SPANISH = ("es", "Spanish", "EUR", "Barcelona")
    FRENCH = ("fr", "French", "EUR", "Paris")
    HINDI = ("hi", "Hindi", "INR", "Mumbai")
    PORTUGUESE = ("pt", "Portuguese", "BRL", "São Paulo")
    RUSSIAN = ("ru", "Russian", "RUB", "St. Petersburg")
    CHINESE = ("zh", "Chinese", "CNY", "Shanghai")

    @property
    def code(self) -> str:
        return self.value[0]

    @property
    def name(self) -> str:
        return self.value[1]

    @property
    def currency(self) -> str:
        return self.value[2]

    @property
    def region(self) -> str:
        return self.value[3]


@dataclass
class MultilingualWorkItem:
    """Work item with translations."""
    work_item_code: str
    translations: Dict[str, str]  # language_code -> description
    prices: Dict[str, float]      # language_code -> unit_price
    unit: str


class CWICRMultilingual:
    """Work with CWICR across languages."""

    # Exchange rates to USD (approximate)
    EXCHANGE_RATES = {
        'AED': 0.27,
        'EUR': 1.08,
        'CAD': 0.74,
        'INR': 0.012,
        'BRL': 0.20,
        'RUB': 0.011,
        'CNY': 0.14,
        'USD': 1.0
    }

    def __init__(self, databases: Dict[str, pd.DataFrame] = None):
        """Initialize with language databases."""
        self.databases = databases or {}
        self._index_databases()

    def _index_databases(self):
        """Create code-based index for each database."""
        self.indexes = {}
        for lang, df in self.databases.items():
            if 'work_item_code' in df.columns:
                self.indexes[lang] = df.set_index('work_item_code')

    def load_database(self, language: CWICRLanguage,
                      file_path: str):
        """Load database for specific language."""
        # Detect format and load
        if file_path.endswith('.parquet'):
            df = pd.read_parquet(file_path)
        elif file_path.endswith('.xlsx'):
            df = pd.read_excel(file_path)
        elif file_path.endswith('.csv'):
            df = pd.read_csv(file_path)
        else:
            raise ValueError(f"Unsupported format: {file_path}")

        self.databases[language.code] = df
        if 'work_item_code' in df.columns:
            self.indexes[language.code] = df.set_index('work_item_code')

    def get_item_translations(self, work_item_code: str) -> MultilingualWorkItem:
        """Get all translations for a work item."""

        translations = {}
        prices = {}
        unit = ""

        for lang, index in self.indexes.items():
            if work_item_code in index.index:
                row = index.loc[work_item_code]
                translations[lang] = str(row.get('description', ''))
                prices[lang] = float(row.get('unit_price', 0))
                if not unit:
                    unit = str(row.get('unit', ''))

        return MultilingualWorkItem(
            work_item_code=work_item_code,
            translations=translations,
            prices=prices,
            unit=unit
        )

    def translate(self, work_item_code: str,
                  from_lang: str,
                  to_lang: str) -> Optional[str]:
        """Translate work item description."""

        if to_lang not in self.indexes:
            return None

        if work_item_code in self.indexes[to_lang].index:
            return str(self.indexes[to_lang].loc[work_item_code].get('description', ''))

        return None

    def compare_prices(self, work_item_code: str,
                       normalize_to_usd: bool = True) -> Dict[str, float]:
        """Compare prices across regions."""

        prices = {}

        for lang, index in self.indexes.items():
            if work_item_code in index.index:
                price = float(index.loc[work_item_code].get('unit_price', 0))

                if normalize_to_usd:
                    # Get currency for this language
                    currency = self._get_currency(lang)
                    rate = self.EXCHANGE_RATES.get(currency, 1.0)
                    price = price * rate

                prices[lang] = round(price, 2)

        return prices

    def _get_currency(self, lang_code: str) -> str:
        """Get currency for language code."""
        for lang in CWICRLanguage:
            if lang.code == lang_code:
                return lang.currency
        return 'USD'

    def find_cheapest_region(self, work_item_code: str) -> Tuple[str, float]:
        """Find region with lowest price (USD normalized)."""

        prices = self.compare_prices(work_item_code, normalize_to_usd=True)

        if not prices:
            return ('', 0)

        cheapest = min(prices.items(), key=lambda x: x[1])
        return cheapest

    def find_most_expensive_region(self, work_item_code: str) -> Tuple[str, float]:
        """Find region with highest price (USD normalized)."""

        prices = self.compare_prices(work_item_code, normalize_to_usd=True)

        if not prices:
            return ('', 0)

        expensive = max(prices.items(), key=lambda x: x[1])
        return expensive

    def cross_language_search(self, query: str,
                              source_lang: str) -> Dict[str, List[str]]:
        """Search in one language, get results in all languages."""

        if source_lang not in self.databases:
            return {}

        source_df = self.databases[source_lang]

        # Find matching codes
        matches = source_df[
            source_df['description'].str.contains(query, case=False, na=False)
        ]['work_item_code'].tolist()

        # Get translations for matches
        results = {}
        for code in matches[:10]:  # Limit to 10
            item = self.get_item_translations(code)
            results[code] = item.translations

        return results

    def price_comparison_report(self, work_item_codes: List[str]) -> pd.DataFrame:
        """Generate price comparison report across regions."""

        rows = []
        for code in work_item_codes:
            item = self.get_item_translations(code)
            prices_usd = self.compare_prices(code, normalize_to_usd=True)

            row = {
                'code': code,
                'description': item.translations.get('en', list(item.translations.values())[0] if item.translations else ''),
                'unit': item.unit
            }

            for lang, price in prices_usd.items():
                row[f'price_{lang}_usd'] = price

            if prices_usd:
                row['min_price'] = min(prices_usd.values())
                row['max_price'] = max(prices_usd.values())
                row['price_variance'] = row['max_price'] - row['min_price']

            rows.append(row)

        return pd.DataFrame(rows)


class LanguageDetector:
    """Detect language of construction text."""

    # Common construction terms by language
    KEYWORDS = {
        'en': ['concrete', 'wall', 'floor', 'door', 'window', 'steel', 'brick'],
        'de': ['beton', 'wand', 'boden', 'tür', 'fenster', 'stahl', 'ziegel'],
        'es': ['hormigón', 'pared', 'piso', 'puerta', 'ventana', 'acero', 'ladrillo'],
        'fr': ['béton', 'mur', 'plancher', 'porte', 'fenêtre', 'acier', 'brique'],
        'ru': ['бетон', 'стена', 'пол', 'дверь', 'окно', 'сталь', 'кирпич'],
        'zh': ['混凝土', '墙', '地板', '门', '窗', '钢', '砖'],
        'pt': ['concreto', 'parede', 'piso', 'porta', 'janela', 'aço', 'tijolo'],
        'ar': ['خرسانة', 'جدار', 'أرضية', 'باب', 'نافذة', 'فولاذ', 'طوب'],
        'hi': ['कंक्रीट', 'दीवार', 'फर्श', 'दरवाजा', 'खिड़की', 'इस्पात', 'ईंट']
    }

    @staticmethod
    def detect(text: str) -> str:
        """Detect language of text."""
        text_lower = text.lower()

        scores = {}
        for lang, keywords in LanguageDetector.KEYWORDS.items():
            score = sum(1 for kw in keywords if kw in text_lower)
            if score > 0:
                scores[lang] = score

        if scores:
            return max(scores.items(), key=lambda x: x[1])[0]

        return 'en'  # Default to English

クイックスタート

# Initialize multilingual support
multi = CWICRMultilingual()

# Load databases
multi.load_database(CWICRLanguage.ENGLISH, "cwicr_en.parquet")
multi.load_database(CWICRLanguage.GERMAN, "cwicr_de.parquet")
multi.load_database(CWICRLanguage.SPANISH, "cwicr_es.parquet")

# Get translations
item = multi.get_item_translations("CONC-001")
print(f"EN: {item.translations.get('en')}")
print(f"DE: {item.translations.get('de')}")

価格比較

# Compare concrete prices across regions
prices = multi.compare_prices("CONC-001", normalize_to_usd=True)
print(prices)

# Find cheapest region
region, price = multi.find_cheapest_region("CONC-001")
print(f"Cheapest: {region} at ${price}")

リソース

• DDC Book：第 2.2 章 - オープンデータ統合
• CWICR Database：26 言語、8 つの国別基盤 + 30 の市場