[Blindmath] Translating documents with MathML into Nemeth (CORRECTION)
Susan Jolly
easjolly at ix.netcom.com
Mon Jul 18 19:17:06 UTC 2011
Neil,
Thanks for pointing out my mistake.
I'm still not positive I understand what MathPlayer 3 will be doing. My
specific question is whether MathPlayer 3 will convert just the math islands
(math expressions in MathML) or the entire document. If the former, then I
understand why the screenreader needs the capability to accept mixed
content, i.e. print text it needs to translateto braille plus braille math
(or braille anything really) that has already been translated and needs to
be passed through unchanged.
However, if MathPlayer is only translating the math islands I have a
concern. Nemeth is actually a complete braille system with its own rules for
formatting and for translating non-math text as well as math. The rules for
translating non-math text are similar but not identical to those for
standard contracted English braille (EBAE). The reasons for these
differences in the translation of text, many of which differences depend on
context, are necessary in order to avoid ambiguity.
Also, Nemeth treats embedded and displayed math somewhat differently, just
as MathML and LaTeX do.
BTW, I've discovered a potential source of a variety of tests of Nemeth
translation capability.
The BANA June 2001 UEBC Sampler 2 can be downloaded from this page.
http://www.iceb.org/ubcbhdr.html
The ubcb2brf.zip file contains the braille math translations. You'll have
to do a bit of sorting as the braille on the lefthand pages is Nemeth while
that on the righthand pages is an older version of UEBC.
The ubcb2wpd.zip file contains WordPerfect files. I discovered that if I
changed the file extension to .doc and opened the file in Word 2007 (under
Vista) that the print source appears in text boxes. I imagine that it
wouldn't take too much effort for a sighted person who is facile with
MathPlayer to convert at least some representative examples to electronic
print math.
Susan
More information about the BlindMath
mailing list