Re: LYNX-DEV minor display problem (?character 0xA2?)

[Klaus Weide]

On Sat, 3 May 1997, Hynek Med wrote:

> On Sat, 3 May 1997, Klaus Weide wrote:
> 
> > There is also a specific test for just this char 0x9B in src/GridText.c.
> > It is blocked in some cases, but not for IBM PC character set.
> > That may have also to be modified.
> 
> Well, what about changing this test to check all control characters in the
> upper area (0x80-0x9f, see the problem I described a few messages ago)? 

Done in 2.7.1ac-0.20.

Or rather, not done as you proposed (and I doubt your proposal was
serious), but each display character set has an associated
lowest_eightbit[i] which tells which is the lowest char with high bit set
which is allowed for that character set.  So normally this would be either
128 (for PC code pages, KOI8-R, etc.) or 160 (for ISO charsets).
That was already there, the only change made now is that characters not
allowed according to lowest_eightbit[i] are thrown away, in a late stage
of processing.  Complaints from people who now cannot see what they used
to be able to see will be directed your way. :)

Note that this doesn't solve Bela's problem, he wants to punch a hole of
just one char into the range of allowed byte values.

Throwing away characters this late in the game shouldn't be necessary,
it should all be handled at earlier stages.  But that isn't quite
finished, you are no doubt aware that chartrans doesn't handle ALT strings
correctly etc.  When I (try to) complete that, I may disable the
late-stage check again, at least for testing.

   Klaus

;
; To UNSUBSCRIBE:  Send a mail message to address@hidden
;                  with "unsubscribe lynx-dev" (without the
;                  quotation marks) on a line by itself.
;