Readable Comprehensions

The previous lessons built new lists of scores and Boolean checks with ordinary loops. Those loops are clear, and they remain the right tool when the work has several steps.

When a loop performs one small, predictable transformation, Python offers a shorter form called a comprehension. Start with this familiar loop:

long_titles = []
for lesson in lessons:
    if lesson["minutes"] >= 10:
        long_titles.append(lesson["title"])

A list comprehension expresses the same work in one collection-shaped expression:

long_titles = [
    lesson["title"]
    for lesson in lessons
    if lesson["minutes"] >= 10
]

Read it as: collect each lesson title for each lesson if its duration is at least ten minutes.

Other comprehension results

A set comprehension keeps unique values:

chapters = {
    lesson["chapter"]
    for lesson in lessons
}

A dictionary comprehension creates key-value pairs:

minutes_by_title = {
    lesson["title"]: lesson["minutes"]
    for lesson in lessons
}

The braces look similar. A colon creates a dictionary comprehension; a single expression creates a set comprehension.

Prefer clarity over compression

Use a comprehension when it has:

• one clear output expression;

• one source collection;

• at most one simple filter.

Use a normal loop when the work needs several statements, multiple decisions, error handling, or explanatory intermediate names. A comprehension is a readability tool, not a contest to fit the most logic on one line.

Drop the brackets when nothing needs the list

A comprehension builds a whole collection. When that collection is fed straight into sum(), min(), max(), any(), or all(), nothing ever needs it to exist:

Same answer. The first line builds a two-item list, hands it to sum(), and throws it away. The second supplies the values one at a time without building an intermediate list.

Without the brackets it is called a generator expression, and inside a function call like this the parentheses of the call are enough. Prefer it whenever the collection is not the point:

sum(lesson["minutes"] for lesson in lessons)
any(lesson["minutes"] > 15 for lesson in lessons)

Keep the brackets when you actually want the collection: to store it, return it, loop over it twice, or ask for its length.

The saving here is two items and invisible. Over a large collection it is the difference between holding everything in memory and holding one item at a time. Chapter 11 returns to generator expressions and explains why they are lazy.

{lesson["chapter"] for lesson in lessons}

Build three useful views

Complete the function with one list, one set, and one dictionary comprehension.