Digit Shapes

Platform SDK: International Features

Digit Shapes

Arabic, Thai, and Indic languages have classical shapes for numbers that are different from the conventional western digits most often used on computers. To avoid ambiguity in naming these shapes, this document uses the following names from the Unicode standard.

Countries	Unicode name
Europe, America, and many others	European digits
Arabic countries (although many use European digits)	Arabic-Indic digits
Thailand (although many use European digits)	Thai digits

Unicode provides separate code points for each digit shape. Thus, to access special language–digit shapes, you can use the relevant Unicode character codes for the digits above U+0030 through U+0039. These are always displayed with the appropriate shape (subject to the available fonts).

The Unicode digits U+0030 through U+0039 are nominally the European digits of 0 through 9, but their digit shape can be altered. The system displays the digits U+0030 through U+0039 in the following ways:

If the national digits control code (NADS) precedes the output, the system uses digit shapes based on the user's selected locale.
If the nominal digits control code (NODS) precedes the output, the system uses European digits.
If neither NADS nor NODS precedes the output, the system uses the value of LOCALE_IDIGITSUBSTITUTION to determine digit shapes.

Native Digits

Native digits are the digit shapes chosen by the user in the Regional Settings program on the Number property sheet. To find the user's preferred digit presentation, use the GetLocaleInfo function with the LOCALE_SNATIVEDIGITS flag. Typically, Unicode digit codes are generated in run-time system routines such as _itoa, so common run-time systems must be upgraded for the application to inspect this variable appropriately.

Digit Substitution

Digit substitution tells the system how to print digits U+0030 through U+0039 in the absence of the NADS or NODS control codes. The LOCALE_IDIGITSUBSTITUTION constant, which controls this operation, can have one of the following values.

Value	Meaning
0	Context-based substitution. Digits are displayed based on the previous text in the same output—that is, European digits follow Latin scripts, Arabic-Indic digits follow Arabic text, and Thai digits follow Thai text. When there is no preceding text, the locale and the displayed reading order determine digit substitution, as shown in the following.
	Locale	Reading order	Digits used
	Arabic	Right-to-left	Arabic-Indic
	Thai	Left-to-right	Thai
	All others	Any	No substitution is used.
1	No substitution is used. This gives full Unicode compatibility.
2	Native digit substitution. Digits are displayed according to LOCALE_SNATIVEDIGITS.

Digit Shaping for a Single API

The ExtTextOut, GetCharacterPlacement, and GCP_RESULTS APIs have flags that govern the substitution of Unicode characters U+0030 through U+0039 for the duration of the function call. That is, these flags override regional settings in Control Panel, but do not reset the settings. Also, they do not override the Unicode control characters NADS and NODS. The following flags are available.

Flags	Digits used
ETO_NUMERICSLATIN	European digits.
ETO_NUMERICSLOCAL	Digits appropriate to the locale.
GCP_NUMERICSLATIN	European digits.
GCP_NUMERICSLOCAL	Digits appropriate to the locale.
GCPCLASS_LATINNUMBER	European digits.
GCPCLASS_LOCALNUMBER	Digits appropriate to the locale.