Fwd: libfontenc bugs for iso8859-7 and iso8859-6

[Alan Coopersmith]

Forwarding a mail from someone who confused my personal address with that
of our bug tracker or mailing list, and has seemingly ignored my suggestion
to take it to one.   I'm not going to have time to fix this anytime soon,
but it doesn't seem that involved if someone else is interested.

	-alan-

-------- Original Message --------
Subject: libfontenc bugs for iso8859-7 and iso8859-6
Date: Sat, 31 Mar 2012 10:01:03 -0400 (EDT)
From: Norm Pierce
To: alan.coopersmith at oracle.com

I came across a couple of bugs in fontenc.c of libfontenc.

Function iso8859_7_to_unicode() maps cell 0xd2 to U+03A2.  However,
Unicode does not define code point U+03A2, nor does iso8859-7 define cell
0xd2.

Function iso8859_6_to_unicode() maps cell 0xc0 to U+0620.  Although
(beginning with version 6.0 in 2011) Unicode does define a new code point
U+0620, iso8859-6 does not define cell 0xc0, so that mapping is
inappropriate.

Because of these two bugs, mkfontscale believes there are glyphs missing
and so will never create entries in fonts.scale for iso8859-7 or iso8859-6
even though the re-encodable font file contains a complete set of glyphs
for these encodings.

Norm Pierce

References:
http://www.unicode.org/Public/MAPPINGS/ISO8859/8859-7.TXT
http://www.unicode.org/Public/MAPPINGS/ISO8859/8859-6.TXT
http://www.unicode.org/charts/PDF/U0370.pdf
http://www.unicode.org/charts/PDF/U0600.pdf