【Python】ジェネレータ式とリスト内包表記の使い分け｜設計判断とアンチパターン

Copy
# リスト内包表記
[式 for x in iterable if 条件]

# ジェネレータ式
(式 for x in iterable if 条件)

Copy
# 反復可能オブジェクト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# terget_userはジェネレータオブジェクト
terget_user = (user for user in users if user["user_id"]  == "1")

Copy
(式 for 変数 in iterable if 条件)

Copy
# ディクショナリリスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# ジェネレータ式でジェネレータオブジェクト生成
user_names = (user["name"] for user in users)

Copy
# ディクショナリリスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# ジェネレータ式でジェネレータオブジェクト生成
user_names = (user["name"] for user in users)

# forで遅延評価して値を取り出す
for name in user_names:
    print(name)  # ['Tujimura', 'mori', 'shimada', 'kyogoku']

Copy
# ディクショナリリスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# ジェネレータ式でジェネレータオブジェクト生成
user_names = (user["name"] for user in users)

# nextで遅延評価して値を取り出す
print(next(user_names)) # Tujimura
print(next(user_names)) # mori
print(next(user_names)) # shimada
print(next(user_names)) # kyogoku
print(next(user_names)) # StopIteration

Copy
# nextで遅延評価して値を取り出す
print(next(user_names)) # Tujimura
print(next(user_names)) # mori
print(next(user_names)) # shimada
print(next(user_names)) # kyogoku
print(next(user_names)) # StopIteration

Copy
# 辞書リスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "1", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# ジェネレータ式で抽出(2件とれる)
terget_user = (user for user in users if user["user_id"] == "1")

# 遅延評価で最初の1件取得
terget_user_val = next(terget_user)
print(terget_user_val) 
# 実行結果
# {'user_id': '1', 'name': 'Tujimura', 'age': 11}

Copy
# 辞書リスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# ジェネレータ式で1件を抽出
terget_user = (user for user in users if user["user_id"] == "1")

# 遅延評価（for）
for user in terget_user:
    print(user) 
# 実行結果
# {'user_id': '1', 'name': 'Tujimura', 'age': 11}

Copy
# 辞書リスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# ① user_id が "1" のユーザーを抽出
filtered = (user for user in users if user["user_id"] == "1")

# ② name だけに変換
names = (user["name"] for user in filtered)

# ③ 文字列長が5以上のものだけ通す（例）
validated = (name for name in names if len(name) >= 5)

# ④ 最終出力
for name in validated:
    print(name)

# 実行結果
# Tujimura

Copy
def filter_user(users):
    return (user for user in users if user["user_id"] == "1")

def extract_name(users):
    return (user["name"] for user in users)

def validate_name(names):
    return (name for name in names if len(name) >= 5)

pipeline = validate_name(extract_name(filter_user(users)))

for name in pipeline:
    print(name)

Copy
# 辞書リスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# パイプライン（遅延評価）
filtered = (user for user in users if user["age"] >= 20)
names = (user["name"] for user in filtered)
validated = (name for name in names if len(name) >= 5)

# ここで全件保持してしまう（遅延評価のメリットが消える）
result = list(validated)
print(result)
# 実行結果: ['shimada', 'kyogoku'] など

Copy
result = [u["name"] for u in users if u["age"] >= 20 and len(u["name"]) >= 5]
print(result)
# 実行結果: ['shimada', 'kyogoku'] など

Copy
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

validated = (
    user["name"]
    for user in users
    if user["age"] >= 20 and len(user["name"]) >= 5
)

# 1回目の走査（OK）
print(list(validated))  # ['shimada', 'kyogoku']

# 2回目の走査（空になる）
print(list(validated))  # []

Copy
validated_list = [
    user["name"]
    for user in users
    if user["age"] >= 20 and len(user["name"]) >= 5
]
print(validated_list)  # 何度でも使える

Copy
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

validated = (
    user["name"]
    for user in users
    if user["age"] >= 20
)

# 長さが欲しい（できない）
# print(len(validated))  # TypeError

# 3番目が欲しい（できない）
# print(validated[2])  # TypeError

Copy
validated_list = [user["name"] for user in users if user["age"] >= 20]

print(len(validated_list))     # 3
print(validated_list[1])       # 'shimada' など

Copy
# 辞書リスト
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

validated = (
    user["name"]
    for user in users
    if user["age"] >= 20
)

# デバッグのつもりで中身を確認（ここで全消費される）
print(list(validated))  # ['mori', 'shimada', 'kyogoku']

# その後に処理しようとしても、もう空
for name in validated:
    print(name)  # 何も出ない

Copy
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70}
]

# パイプライン（遅延評価）
gen = (user for user in users if user["age"] >= 20)

# ここで「確認のため」「使い回しのため」と list 化してしまう
tmp = list(gen)  # ← この時点で全件保持（遅延評価のメリットが消える）

# 以降は結局 list を処理しているだけ
names = (user["name"] for user in tmp)
print(list(names))  # ['mori', 'shimada', 'kyogoku']

Copy
users = [
    {"user_id": "1", "name": "Tujimura", "age": 11},
    {"user_id": "2", "name": "mori", "age": 20},
    {"user_id": "3", "name": "shimada", "age": 50},
    {"user_id": "4", "name": "kyogoku", "age": 70},
    {"user_id": "5", "name": None, "age": 30},  # 例外の原因を混ぜる
]

# どこで壊れたのか分かりにくい
pipeline = (
    name.strip().lower()
    for name in (u["name"] for u in users if u["age"] >= 20)
    if len(name) >= 5
)

for x in pipeline:
    print(x)
# 実行時に AttributeError: 'NoneType' object has no attribute 'strip' 等

Copy
def select_users(users):
    return (u for u in users if u["age"] >= 20)

def extract_name(users):
    return (u["name"] for u in users)

def normalize(names):
    # ここが落ちうる（Noneなど）
    for name in names:
        yield name.strip().lower()

def filter_len(names):
    return (name for name in names if len(name) >= 5)

u1 = select_users(users)
n1 = extract_name(u1)
n2 = normalize(n1)
n3 = filter_len(n2)

for x in n3:
    print(x)