[liblouis-liblouisxml] Re: Status ueb integration
- From: Bert Frees <bertfrees@xxxxxxxxx>
- To: liblouis-liblouisxml@xxxxxxxxxxxxx
- Date: Thu, 16 Jul 2015 11:46:49 +0200
Bert, is this in line with your ideas?
Yes.
I can see only one potential problem if decpoint is a shorthand for begnum and
numericmodechars:
If a number such as .06 needs to be translated to #.jf, that may not be
supported by numericmodechars and begnum (I don't know, I have to check). So in
this case we should probably change the behavior of one of these two opcodes in
order to fit the bill.
Davy Kager writes:
The question is if it is useful to have something that is a 'decpoint' but
that still ends a number
While we could achieve this behavior with a combination of numericmodechars
and
decpoint, I think we should try to keep supporting decpoint, e.g. by just
automatically adding decpoint characters to numericmodechars.
What I was thinking of is that decpoint is really a shorthand for begnum and
numericmodechars. So
decpoint . 46
is equivalent to
begnum . 46
numericmodechars .
There is no 'decpoint' character type, so really all the opcode does is
combine begnum and numericmodechars. And some tables may actually want to use
midnum instead of begnum, e.g. in Dutch a number like .06 isn't seen a lot.
So to summarize, you should use one of these two options:
1. decpoint, which is a shorthand for begnum and numericmodechars. This is
the old behavior. Note that this doesn't work with the UEB patches, but can
easily be fixed.
2. begnum/midnum and numericmodechars. This is the new and IMO much more
elegant solution.
Bert, is this in line with your ideas?
Davy
-----Oorspronkelijk bericht-----
Van: liblouis-liblouisxml-bounce@xxxxxxxxxxxxx
[mailto:liblouis-liblouisxml-bounce@xxxxxxxxxxxxx] Namens Bert Frees
Verzonden: donderdag 16 juli 2015 10:53
Aan: liblouis-liblouisxml@xxxxxxxxxxxxx
Onderwerp: [liblouis-liblouisxml] Re: Status ueb integration
Davy Kager writes:
The question is if it is useful to have something that is a 'decpoint' but
that still ends a number
AFAIK decpoint it has two purposes: it does not end a number, and it only
matches when it preceeds a digit.
While we could achieve this behavior with a combination of numericmodechars
and
decpoint, I think we should try to keep supporting decpoint, e.g. by just
automatically adding decpoint characters to numericmodechars.
For cases when a decpoint should end a number, begnum could be used instead.
Why are these only failing with the yaml test harness? Is the Python harness
not ran with make check?
Exactly.
For a description of the software, to download it and links to
project pages go to
http://liblouis.org
Other related posts:
