Robust distinction between IDENTIFIER and TYPE_NAME tokens. To distinguish the tokens we used to track the lexer context, which is fragile. This patch implements a better way - combine the two tokens into a common parser rule and let the parser context decide if the declaration is valid, which it already does by checking the naming conflicts. TRAC #23158 Signed-off-by: Nicolas Capens Merged-by: Jamie Madill Authored-by: alokp@chromium.org git-svn-url: https://code.google.com/p/angleproject/source/detail?r=2190 Review URL: https://codereview.appspot.com/8797044