The failure to parse this test (Malformed string) is due to this line:
test subst-3.2 {backslash substitutions with utf chars} {
# 'j' is just a char that doesn't mean anything, and \344 is 'ä'
# that also doesn't mean anything, but is multi-byte in UTF-8.
list [subst \j] [subst \\j] [subst \\344] [subst \\\344]
} "j j \344 \344"
This is compiled to PIR (in part) as:
$P83 = ascii:"\n # 'j' is just a char that doesn't mean anything,
and \\344 is '\x{e4}'\n # that also doesn't mean anything, but is
multi-byte in UTF-8.\n list [subst \\j] [subst \\\\j] [subst \\\
\344] [subst \\\\\\344]\n"
the "\x{e4}" makes this non ascii, so parrot hits this line and blows
up.
--
Will "Coke" Coleda
wi...@coleda.com