Performance Tip Optimizing Iteration: Difference between revisions
AutoSpider (talk | contribs) (Add "cleanup" tag) |
(Cleanup) |
||
(4 intermediate revisions by one other user not shown) | |||
Line 1: | Line 1: | ||
{{ | {{LangSwitch}} | ||
[[Category:Learning]] | [[Category:Learning]] | ||
[[Category:HowTo]] | [[Category:HowTo]] | ||
[[Category:Developing with Qt]] | [[Category:Developing with Qt]] | ||
Here are two tips to help produce code that makes the most out of Qt internal performance. | Here are two tips to help produce code that makes the most out of Qt internal performance. | ||
Line 14: | Line 9: | ||
Qt allows you to use both Java-style and STL-style iterators to step through your structures. Your selection of iterator style has no performance impact, but your selection of iterator operator does. | Qt allows you to use both Java-style and STL-style iterators to step through your structures. Your selection of iterator style has no performance impact, but your selection of iterator operator does. | ||
<code>QListIterator<int> i(list); | Java style iterator: | ||
<code> | |||
QListIterator<int> i(list); | |||
while (i.hasNext()) | while (i.hasNext()) | ||
process(i.next()); | |||
</code> | </code> | ||
STL (Standard Template Library) style iterator: | |||
<code> | |||
<code>QList<int>::iterator i; | QList<int>::iterator i; | ||
for (i = list.begin(); i != list.end(); ++i) | for (i = list.begin(); i != list.end(); ++i) | ||
process(*i); | |||
</code> | </code> | ||
When using STL-style iterators with a list containing complex items, execution is faster if you use the ++i operator instead of the i++ operator. The i++ will force your loop to work on a copy of the item i. | |||
When using STL-style iterators with a list containing complex items, execution is faster if you use the + | |||
== Careful with foreach == | |||
Benchmarking Qt applications indicates there is always a performance penalty to using a foreach loop as opposed to a for loop with an iterator. However, you can greatly reduce the performance penalty if you use a const iterator in your foreach loop. This can often make the performance penalty negligible, though it is never zero. | Benchmarking Qt applications indicates there is always a performance penalty to using a foreach loop as opposed to a for loop with an iterator. However, you can greatly reduce the performance penalty if you use a const iterator in your foreach loop. This can often make the performance penalty negligible, though it is never zero. | ||
<code> | |||
<code>foreach (const QString & | foreach (const QString &i, list) | ||
process(i); | |||
</code> | </code> | ||
== Additional reading == | |||
# Kai Köhne's blog, [https://blog.qt.io/blog/2009/01/23/iterating-efficiently/ Iterating efficiently] | |||
' |
Latest revision as of 21:18, 28 June 2015
En Ar Bg De El Es Fa Fi Fr Hi Hu It Ja Kn Ko Ms Nl Pl Pt Ru Sq Th Tr Uk Zh
Here are two tips to help produce code that makes the most out of Qt internal performance.
Iterator selection
Qt allows you to use both Java-style and STL-style iterators to step through your structures. Your selection of iterator style has no performance impact, but your selection of iterator operator does.
Java style iterator:
QListIterator<int> i(list);
while (i.hasNext())
process(i.next());
STL (Standard Template Library) style iterator:
QList<int>::iterator i;
for (i = list.begin(); i != list.end(); ++i)
process(*i);
When using STL-style iterators with a list containing complex items, execution is faster if you use the ++i operator instead of the i++ operator. The i++ will force your loop to work on a copy of the item i.
Careful with foreach
Benchmarking Qt applications indicates there is always a performance penalty to using a foreach loop as opposed to a for loop with an iterator. However, you can greatly reduce the performance penalty if you use a const iterator in your foreach loop. This can often make the performance penalty negligible, though it is never zero.
foreach (const QString &i, list)
process(i);
Additional reading
- Kai Köhne's blog, Iterating efficiently