Comment 1 by jteh on 2010-01-24 22:23
Needs to be fixed in liblouis. Also, there is no "standard" table; each table needs to be considered separately. Can you be more specific?

Comment 2 by bramd (in reply to comment 1) on 2010-01-24 22:33
Replying to jteh:

Needs to be fixed in liblouis. Also, there is no "standard" table; each table needs to be considered separately. Can you be more specific?

This is tested with the US English 8 dot computer braille table, which is set as the default table in NVDA. I'll test with other tables and file relevant bugs with liblouis if needed. However it might be handy to show a "?" or some other indication if there is a character that's not in the used braille table. The current way of showing \x is quite annoying to read. Not sure if this is NVDA or liblouis behavior though.

I'll update this ticket when I have more info.

Comment 3 by jteh on 2010-01-24 22:43
It is liblouis behaviour. Furthermore, it breaks the mapping between input and output positions; see #499. A question mark or similar might be better, but note that this would be different for every table, so we might need some way to define a character to be inserted for "unknown" characters.

Comment 4 by Bernd on 2010-02-02 15:44
I tested this with the de-de-comp8.ctb table.

after sign \x00ff there aren't any signs any more. Unicode chars often begin with \x2 which aren't included here.

For example NVDA shows \0x20.. if you are in a table in Microsoft word. or for some quotes.

Comment 5 by MHameed on 2012-09-11 07:53
The tables have undergone some changes since this ticket was opened, no doubt there are still some characters showing up as \xhhhh.

Could you please make a list of the characters that are still undefined for your table, and we can patch the relevant table.

Thanks.
Changes:
Milestone changed from None to near-term

Comment 6 by bramd (in reply to comment 5) on 2012-09-11 09:44
Replying to MHameed:

The tables have undergone some changes since this ticket was opened, no doubt there are still some characters showing up as \xhhhh.

Do the latest snapshots contain the latest liblouis tables for testing?

Comment 7 by nvdakor on 2012-11-21 08:58
Hi,
Writing a few comments as it relates to a major project going on within LibLouis community:
Latest snapshots and tables: no. Latest dev snapshots doesn't include latest tables. It would have been nice if we did have a branch just for testing new tables (including the next gen UEB table).
As for hex code: as Mesar said, could you attach a UTF-8 txt file containing the actual Unicode char, char name and its hex value? Also, in the text file, could you specify which table this problem occurs so LibLouis people can go over them?
A good format of the file would be:
#Table name where the problem occurs e.g. German Grade 0.
Unicode char, comma, space, name of the char, comma, space, hex value, comma, space, dot pattern (if any)
For example:
#UEB grade 1/grade 2:
lowercase Alpha (insert the actual alpha char here), Greek alpha, x3b1, 46-1
That way, others can track the problem char and come up with a fix. Thanks.

If there are specific characters missing in specific tables that you believe can/should be mapped, please report issues to liblouis. Friendlier presentation of unknown characters will be addressed in #2949.