codeql/java/ql/src/Performance/InefficientKeySetIterator.qhelp at main · github/codeql · GitHub

Repository navigation

47 lines (34 loc) · 1.62 KB

<!DOCTYPE qhelp PUBLIC

"-//Semmle//qhelp//EN"

"qhelp.dtd">

<qhelp>

Java's Collections Framework provides several different ways of iterating

the contents of a map. You can retrieve the set of keys, the collection of values, or the

set of "entries" (which are, in effect, key/value pairs).

The choice of iterator can affect performance. For example, it is considered bad practice to iterate

the key set of a map if the body of the loop performs a map lookup on each retrieved key

anyway.

</overview>

Evaluate the requirements of the loop body. If it does not actually need the key

apart from looking it up in the map, iterate the map's values (obtained by a call to

<code>values</code>) instead. If the loop genuinely needs both key and value for

each mapping in the map, iterate the entry set (obtained by a call to

<code>entrySet</code>) and retrieve the key and value from each entry. This saves a

more expensive map lookup each time.

</recommendation>

In the following example, the first version of the method <code>findId</code> iterates

the map <code>people</code> using the key set. This is inefficient because the body of the loop

needs to access the value for each key. In contrast, the second version iterates the map

using the entry set because the loop body needs both the key and the value for each mapping.

</example>

<li>

Java API Specification:

<a href="https://docs.oracle.com/en/java/javase/11/docs/api/java.base/java/util/Map.html#entrySet()">Map.entrySet()</a>.

</li>

</references>

</qhelp>