[XeTeX] search arabic text in pdf using adobe reader 7.0

[sh]

I am using MiKTeX-XeTeX 2.7.2904 (0.997 svn 539) (MiKTeX 2.7) in
Microsoft Windows XP . I have been successful in creating pdf using
arabxetex and Scheherazade opentype font.

I am using adobe reader 7.0 to read the pdf. When I copy the arabic
characters from the pdf, I get garbage characters when I paste it to
MS Word (which is set to use the unicode Arial MT font). The
individual characters copy just fine, but the characters that are in
the intermediate form do not get copied.

What do I need to do to able to copy the characters from the pdf. The
pdf is encoded with identity-H/CID. I suspect I need to do something
with Cmap or mapping?