Fix handling of both UTF_16_LE and UTF_16_BE. Add an actual test for
authorstevenknight <stevenknight@fdb21ef1-2011-0410-befe-b5e4ea1792b1>
Mon, 2 Mar 2009 19:11:09 +0000 (19:11 +0000)
committerstevenknight <stevenknight@fdb21ef1-2011-0410-befe-b5e4ea1792b1>
Mon, 2 Mar 2009 19:11:09 +0000 (19:11 +0000)
commit1007e8518133efcab25d7ad713262a7660c5bf95
tree5ee6852eebcbda0dcb8ff137037c83c5dc377472
parentc3af1409f97624632420d75b3cba6fc414fef2d5
Fix handling of both UTF_16_LE and UTF_16_BE.  Add an actual test for
scanning Unicode files for implicit dependencies.  Clean up how we handle
decoding, and wrap it for earlier Python versions.

git-svn-id: http://scons.tigris.org/svn/scons/trunk@4061 fdb21ef1-2011-0410-befe-b5e4ea1792b1
src/CHANGES.txt
src/engine/SCons/Node/FS.py
test/Scanner/unicode.py [new file with mode: 0644]