[tlbuild] etex serious bug

Vafa Khalighi simurgh12 at gmail.com
Fri Apr 20 14:14:04 CEST 2012


>
>
>
> Why do you want that the displayed equation thinks the paragraph is
> left-to-right when it is in fact right-to-left?
>
>
>
Thanks for the reply. This is the whole story:

In Persian typography, texts are written from right to left but mathematics
is written from left to right and equation numbers have the same meaning as
they have in LTR. i.e. \eqno means put equation number on the right hand
side and \leqno means put equation on the left hand side. Currently when we
make paragraph RTL in etex, etex places  equation numbers in the opposite
direction, i.e.  \eqno means put equation number on the left hand side and
\leqno means put equation number on the right hand side (exactly opposite
of what it supposed to be in Persian). So to fix this, I thought to put
\beginL before $$ and \endL after $$ but noticed that it introduces extra
vertical space, then I read etex manual and noticed \predisplaydirection
primitive is available. I asked about it on Xetex mailing list, please see

http://tug.org/mailman/htdig/xetex/2010-April/016260.html

http://tug.org/mailman/htdig/xetex/2010-April/016261.html

and the next messages until

http://tug.org/mailman/htdig/xetex/2010-April/016265.html

http://tug.org/mailman/htdig/xetex/2010-April/016266.html


 So to make \eqno and \leqno have their usual meaning,
bidi<http://ctan.org/pkg/bidi>package uses:

\everydisplay\expandafter{\the\everydisplay\predisplaydirection=0 }

but it has the side effect that the vertical space above equation and the
vertical space below equation are not the same.

So if this is not the bug, there remains only one question, how should one
fix this unequal vertical skip before and after $$? If \predisplaydirection
supposed to produce this incorrect result, why the primitive exists anyway?

Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tug.org/pipermail/tlbuild/attachments/20120420/39689867/attachment.html>


More information about the tlbuild mailing list