Finer grained character boundaries in computing SMS fragment lengths

The standard Java character iterator has potentially unbounded
distance between character boundaries, meaning that when breaking an
SMS message into fragments it's not safe to assume that the fragment
will contain such a boundary. This patch special-cases flags (pairs of
Regional Indicator Symbols) and also guarantees some progress in the
case of no boundary found.

Bug: 26210851
Change-Id: Ic16637a834af0dfe40197fbcdb3471c1bf5cd6d2
1 file changed
tree: a5eb15594078a78b47c66641234f35b1f0f8b197
  1. src/
  2. tests/
  3. tools/
  6. README.txt