English
English
When developing multilingual applications, we often involve string comparison issues. PHP provides multiple ways to compare strings, and the strcoll() function is particularly interesting because it determines the result of the comparison based on the current locale. This article will explore the performance differences of strcoll() under different locale settings and illustrate it through specific code examples.
strcoll() is a built-in function in PHP that compares two strings based on locale settings. It returns a result similar to strcmp() :
Returns 0 to indicate that the two strings are equal in the current locale;
Return less than 0 means that the first string is ranked before the second in the sort;
Returns greater than 0 to indicate that the first string is after the second in the sort.
Unlike strcmp() , strcoll() will consider locale rules, such as the sorting method of characters, case sensitivity, and the processing of some special characters.
In PHP, you can use the setlocale() function to set the current locale setting. For example:
setlocale(LC_COLLATE, 'en_US.UTF-8');
LC_COLLATE is a category specifically used to influence string comparison and sorting. Other categories such as LC_TIME , LC_MONETARY , etc. affect time, currency and other formats.
Let’s take the two locale, German and English as examples, and take a look at the performance differences of strcoll() .
setlocale(LC_COLLATE, 'en_US.UTF-8');
echo strcoll("z", "?"); // Output result A
setlocale(LC_COLLATE, 'de_DE.UTF-8');
echo strcoll("z", "?"); // Output result B
In English, "z" is before "?", while in German, since "?" is regarded as a phonic letter, it may be after "z" or even near "a". Therefore, the output results A and B may be different.
Suppose we have a set of names with accents that we want to sort according to the user's language preferences. The code is as follows:
$names = ["Zoe", "?nne", "Anna", "émile"];
setlocale(LC_COLLATE, 'en_US.UTF-8');
usort($names, function($a, $b) {
return strcoll($a, $b);
});
print_r($names);
Under en_US.UTF-8 , the sort may be:
Array
(
[0] => Anna
[1] => émile
[2] => Zoe
[3] => ?nne
)
If replaced with de_DE.UTF-8 :
setlocale(LC_COLLATE, 'de_DE.UTF-8');
Then you may get:
Array
(
[0] => Anna
[1] => ?nne
[2] => émile
[3] => Zoe
)
In some systems, the available locale may be limited. You can view it by running the following command on the command line:
locale -a
Or, try to set locale in PHP and use the return value of setlocale() to determine whether it is successful.
Always check the return value of setlocale() to ensure that locale is set correctly;
If you need to sort the user input language sensitively, be sure to use strcoll() instead of strcmp() ;
For cross-platform consistency, it is recommended to clearly specify the required locale in the application and ensure that these settings are supported in the server configuration;
If the sorting results of strcoll() are used for front-end display (such as contact list, country name, etc.), please simulate different locales in the test to ensure that the sorting logic meets expectations.
You can try the sorting effect of different locales using the following address:
https://gitbox.net/locale-strcoll-demo.php
The page supports selecting different locales and entering string pairs to compare them, so that you can intuitively understand the performance of strcoll() under different locales.
strcoll() is a very useful but often overlooked function. By setting locale reasonably, it can help us implement string comparison logic that is more in line with user language habits. Make good use of strcoll() in multilingual projects can significantly improve the user experience.
