On Sat, Aug 09, 2008 at 03:55:10PM +0200, Mario Lang wrote: > > I agree with you. The normalization should only happen just before the > > output. > > I.e., in a table, which sort of brings us back to Dave's equivalence idea.
Perhaps an option in the table could be provided to turn normalization on or off. Is there a good reason why normalization shouldn't be in effect all the time? If the characters involved are defined in Unicode to be equivalent, as with combining characters and the accented letters, then I can't think of a good reason to distinguish them at the level of the table. For other equivalences - the dashes or, potentially, mathematical symbols, there may be circumstances in which they need to be distinguished. If you are checking a typeset document, for example, the distinctions among the dashes could be vital. Dave's equivalence idea is a good one, except that somebody would have to go through the Unicode character set to find all of the equivalences that should be active by default, and write out a table accordingly. If there is a library function that can carry out the normalization and which is designed for this purpose, I expect it would be desirable to be able to use it, instead of having to re-create it in a table. _______________________________________________ This message was sent via the BRLTTY mailing list. To post a message, send an e-mail to: BRLTTY@mielke.cc For general information, go to: http://mielke.cc/mailman/listinfo/brltty
