python 検索エンジン

# データ
documents = [
    "Python is a popular programming language.",
    "It is known for its simplicity and readability.",
    "Python has a large community of developers.",
    "Python is widely used in web development."
]

# データの前処理
def preprocess(text):
    return text.lower()

# インデックスの作成
def create_index(documents):
    index = {}
    for doc_id, doc in enumerate(documents):
        doc = preprocess(doc)
        words = doc.split()
        for word in words:
            if word not in index:
                index[word] = []
            index[word].append(doc_id)
    return index

# クエリ処理
def search(query, index):
    query = preprocess(query)
    words = query.split()
    result = set(range(len(documents)))

    for word in words:
        if word in index:
            result &= set(index[word])

    return [documents[i] for i in result]

# メイン
if __name__ == "__main__":
    index = create_index(documents)

    while True:
        query = input("検索クエリを入力してください (終了するには 'exit' を入力): ")
        if query == 'exit':
            break
        results = search(query, index)
        if results:
            for i, result in enumerate(results, start=1):
                print(f"{i}. {result}")
        else:
            print("一致する文書はありません。")