Re: FW: Why are the low surrogates numerically larger than the high surrogates?

From: Philippe Verdy <>
Date: Wed, 23 Jan 2013 22:47:10 +0100

2013/1/23 Markus Scherer <>:
> FYI: Because of this confusion, in ICU APIs and documentation we prefer to
> talk about "lead surrogates" and "trail surrogates", in analogy to MBCS lead
> & trail bytes.
You may also combine both expressions as "leading high surrogates" and
"trailing low surrogates" to make sure that you remember their correct
encoding order as well as their exprected value ranges. But when you
speaks about the correctly encoded pairs you don't need this
distrinction and just speaks about "surrogate pairs" (which is not a
pair of random surrogates for any ranges but really a pair contaning
the leading high surrogate followed by the trailing low surrogate.
If you just assume the conformant encoding you don't need to add these
extra precisions, you assume them and abbreviate the expressions. They
are equivalent in such *conformant* context.
Received on Wed Jan 23 2013 - 15:50:55 CST

This archive was generated by hypermail 2.2.0 : Wed Jan 23 2013 - 15:50:55 CST