Hi Dave,
I think it makes perfect sense to have a context opcode also in the backward
direction. Back-translation can be very tricky, so, I think it will be of
use. Also, it gives a nice symmetry to the system.
I assume you mean that multi-pass opcodes without direction implies noback,
i.e. forward direction. An alternative is to make an automatic
search&replace in all tables, provided that only the name of the opcode has
changed. This has already been done in other situations, e.g. when capsign
changed to capsletter.
Bue
-----Oprindelig meddelelse-----
Fra: liblouis-liblouisxml-bounce@xxxxxxxxxxxxx
[mailto:liblouis-liblouisxml-bounce@xxxxxxxxxxxxx] På vegne af Dave Mielke
Sendt: 17. januar 2017 03:50
Til: liblouis-liblouisxml@xxxxxxxxxxxxx
Emne: [liblouis-liblouisxml] Multipass opcode suggestions.
With the multipass rules now separate (assuming the merge of
dm/simlify-passes), would it make sense to make context work for back
translation. This'd mean that "nofor context" would mean dots in the test
and characters in the action.
To try to be backward compatible, would it make sense for all five multipass
opcodes (correct, context, pass2, pass3, pass4), if unqualified, to imply
nofor? Would this, to any of your collective knowledges, break any existing
back translation(s)?
Right now, the _ (underscroe) test is silently increased to 0 if it goes
negative. This, to me, makes it nondeterministic, and, therefore, it makes
the result of the test unpredictable and potentially wrong. What do you
think about changing it so that the test fails if it goes negative since
characters before the start can't be matched?
--
Dave Mielke | 2213 Fox Crescent | The Bible is the very Word of
God.
Phone: 1-613-726-0014 | Ottawa, Ontario | http://Mielke.cc/bible/
EMail: Dave@xxxxxxxxx | Canada K2A 1H7 | http://FamilyRadio.org/
For a description of the software, to download it and links to
project pages go to http://liblouis.org
For a description of the software, to download it and links to
project pages go to http://liblouis.org
