formatError assumes string sequence

[Cory Dodt]

Here's a simple test case. This prints an AttributeError for lines = input.split('\n') inside formatError.

#--------------------------------------------------------------------
from pymeta.grammar import OMeta
g = """xyz ::= [ "xx":a "yy":b "zz":c ] => 'pass'"""
XYZ = OMeta.makeGrammar(g, {}, "XYZ")
input = ['xx', 'OOPS', 'zz']
try:
    print XYZ([input]).apply('xyz')[0]
except Exception, e:
    print e
    print e.formatError(input)