Fix Native Digits Support

[tarekgh]

We always assumed the native digits one character. This was the case for NLS. Now in ICU, there are some locales which use native digits expressed as surrogate pairs. ccp-Cakm-BD locale is a good example of that which uses the native digits "\U0001E950", "\U0001E951", "\U0001E952", "\U0001E953", "\U0001E954", "\U0001E955", "\U0001E956", "\U0001E957", "\U0001E958", "\U0001E959". This change is to allow reading the native digits correctly from ICU.

[ghost]

Tagging subscribers to this area: @dotnet/area-system-globalization
See info in area-owners.md if you want to be subscribed.

Issue Details

---

We always assumed the native digits always one character. This was the case for NLS. Now in ICU, there are some locales which use native digits expressed as surrogate pairs. ccp-Cakm-BD locale is a good example of that which uses the native digits "\U0001E950", "\U0001E951", "\U0001E952", "\U0001E953", "\U0001E954", "\U0001E955", "\U0001E956", "\U0001E957", "\U0001E958", "\U0001E959". This change is to allow reading the native digits correctly from ICU.

Author:	tarekgh
Assignees:	-
Labels:	area-System.Globalization
Milestone:	-

[tarekgh]

CC @eerhardt@maryamariyan if you can help review it.

[danmoseley]

Does this affect treatment of \d in a regex pattern?

[stephentoub]

Does this affect treatment of \d in a regex pattern?

No

[tarekgh]

@eerhardt@stephentoub do you have any more comments?

[eerhardt]

LGTM. Thanks @tarekgh!

[stephentoub]

Should this be index == 10? result is exactly 10 items long, so if we tried to do result[10], it would have thrown.

[tarekgh]

Yes, we can do that. I am not sure if this is worth another PR though.