-+% Exchanged the Unicode equilavents for 0xE6 and 0xE7, to reflect
-+% the change in the mirroring behaviour of U+00AB and U+00BB
-+% between versions 2.1.5 and 2.1.8 of the Unicode Standard.
-+
- % I think the mnemonic for upper space (0xA0) should be different from
- % lower space (0x20). Because the semantics are different, although they
- % are mapped to a single Unicode character. This is specially important
-@@ -18,9 +27,9 @@
- % invented on the pattern of Unicode character names. --RP
- %
- % NOTE: Some entries were missing in the mnemonics.ds file; We invented our
--% own. Those are: <zwnj>, <zwj>, <!+>, <Rl>, <%+>, <.+>, <,f>, <(+>,
-+% own. Those are: <zwnj>, <zwj>, <!+>, <Rl>, <%+>, <.a>, <,f>, <(+>,
- % <)+>, <*X+>, <+f>, <-f>, <.f>, <//f>, <0f>..<9f>, <:f>, <<f>, <=f>,
--% </>f>, <kf>, <yf>, <)/>+>, <<(+>, <!)+>, <(!+>, <<<+>, </>/>+>, <*+>,
-+% </>f>, <yf>, <)/>+>, <<(+>, <!)+>, <(!+>, <<<+>, </>/>+>, <*+>,
- % <!!+>, and <////+>. Also, all character sequences
- % <NU>..<US>, <SP>, <DT> have counter-parts <NU+>..<US+>, <SP+>, <DT+>
-
++% Exchanged the Unicode equilavents for 0xE6 and 0xE7, to reflect
++% the change in the mirroring behaviour of U+00AB and U+00BB
++% between versions 2.1.5 and 2.1.8 of the Unicode Standard.
++
+ % Please note that the Unicode equivalents of the Persian part (0x80..0xFF)
+ % are for compatiblity purposes only. I have also changed the fourth
+ % column consisting of names to reflect the difference. The names are