Locale-Sensitive Number Formatting in Java

Locale-Sensitive Number Formatting in Java

While working on a piece of code, I encountered an issue that I thought was worth sharing. The task seemed straightforward: generate a file path that includes an integer ID. Here’s a simplified version of the code:

// id - arbitrary integer
int id = 123;
// Expected output: /tmp/app-files/123
String path = String.format(Locale.getDefault(), "/tmp/app-files/%d", id);

At first glance, this looks simple and works fine—as long as your locale uses Western Arabic Numerals (Unicode/ASCII 0x30 to 0x39). However, things take an unexpected turn when your locale uses a different numeral system, such as Bengali.

For instance, the Bengali numerals are:

০	১	২	৩	৪	৫	৬	৭	৮	৯

If your locale is set to Bengali, the generated path would look like:

/tmp/app-files/১২৩

This output might not be what you expected, especially if you require the path to strictly use Western Arabic numerals.

The Solution: Use Locale.US

To avoid this issue and ensure consistent formatting across different locales, it’s better to use Locale.US when formatting numbers for paths or similar cases where the output should remain consistent regardless of the user’s locale:

String path = String.format(Locale.US, "/tmp/app-files/%d", id);

Now, regardless of the locale, the path will be correctly formatted as:

/tmp/app-files/123

This approach ensures that your code behaves predictably in environments with different locale settings.

Thanks for reading!