Created attachment 15629 [details] [review]
Patch that fixes this issue

Hello,

I use fribidi library in BiDiTeX project --
support of Hebrew for LaTeX documents, and I found
that when I work with documents in cp1255
LRM/RLM marks are not converted correctly
and the code does not work.

There is a bug in fribidi library, or
more exactly in fribidi_cp1255_to_unicode_c()
and fribidi_unicode_to_cp1255_c() functions.

It does not convert LRM/RLM marks to UNICODE symbols
as required (253, 254 <--> 0x200E,0x200F).

The same happens in reverse direction
(unicode to cp1255)

The easiest way to reproduce the problem is following:
Take text ("a"," ","א","!",RLM," ","b") in utf-8
and run through fribidi you'll see "a !א b" as it should be
Same text encoded with cp1255 run with fribidi -c cp1255
you'll see "a א! b" -- incorrect one.

The patch that fixes this problem is attached,
it was build against version 0.10.9

Artyom