Forward translation will always take priority, but there are now users who are concerned about back-translation. There are rules for translating the occasional mathematical expressions within literary texts. Strictly speaking, they should be translated with the Nemeth Code. The liblouis tables have been tweaked to give a reasonable translation. The same is true of other translators. As might be expected, such tweaks can produce odd results in back-translation. To get an accurate translation of math expressions, even simple ones, they would have to be expressed in MathML and parsed by liblouisxml. John On Wed, Feb 25, 2009 at 09:25:35PM -0800, John Gardner wrote: > Back translation of braille will always have ambiguities. I do not believe > that forward translation should be compromized in order to make back > translation occasionally better. > > I have another question. Why should something not within math tags be > translated as math braille? I was under the impression that expressions > containing non-literary symbols such as = or + should be translated as > computer braille. Is there a rule? Mike? > > -----Original Message----- > From: liblouis-liblouisxml-bounce@xxxxxxxxxxxxx > [mailto:liblouis-liblouisxml-bounce@xxxxxxxxxxxxx] On Behalf Of John J. > Boyer > Sent: Wednesday, February 25, 2009 9:09 PM > To: liblouis-liblouisxml@xxxxxxxxxxxxx > Subject: [liblouis-liblouisxml] Re: Contractions near a slash > > After it has parsed and interpreted a MathML expression liblouisxml calls > liblouis with a special math table to do the final translation. > For Nemeth, this table is nemeth.ctb. It uses the include opcode to read > another table called nemethdefs.cti. This table then includes the > character-definitions table chardefs.cti. This may be one source of > confusion. Perhaps nemethdefs.cti should contain all character definitions > and leave the literary character-definitions table, chardefs.cti, to go its > own way. However, the line between literary and math can be blurry. For > example, a popular magazine article might contain the equation pi=3.14159, > not enclosed in MathML but simply part of qe text. So the literary > character-definition table must define symbols like = as math to get good > translation and back-translation. > > John > > > On Wed, Feb 25, 2009 at 03:04:53PM -0800, John Gardner wrote: > > Am I confused or is something wrong with this philosophy? Math is > > math, and literary is literary. Liblouis should regard anything not > > enclosed in math tags as literary, shouldn't it? So the intermediate > > representation needs to distinguish between symbols with math meanings > > and the same symbols with their literary meaning. If it doesn't, then > > all kinds of nasty things can happen. > > > > -----Original Message----- > > From: liblouis-liblouisxml-bounce@xxxxxxxxxxxxx > > [mailto:liblouis-liblouisxml-bounce@xxxxxxxxxxxxx] On Behalf Of John J. > > Boyer > > Sent: Wednesday, February 25, 2009 2:51 PM > > To: liblouis-liblouisxml@xxxxxxxxxxxxx > > Subject: [liblouis-liblouisxml] Re: Contractions near a slash > > > > Translation of math is a joint effort of liblouisxml and liblouis. > > liblouisxml makes a sort of intermediate representation of math. > > liblouis has to know how to translate that. > > > > John > > > > On Wed, Feb 25, 2009 at 02:38:46PM -0800, Mike Sivill wrote: > > > But with LibLouisXML handling MathML why do we need to define any > > characters as math? Won't LouisXML know it's math when it encounters > > the begin math tag? Isn't everything else handled as literary text? > > > Mike > > > > > > > > > -----Original Message----- > > > From: liblouis-liblouisxml-bounce@xxxxxxxxxxxxx > > > [mailto:liblouis-liblouisxml-bounce@xxxxxxxxxxxxx] On Behalf Of John > > > J. Boyer > > > Sent: Wednesday, February 25, 2009 2:24 PM > > > To: liblouis-liblouisxml@xxxxxxxxxxxxx > > > Subject: [liblouis-liblouisxml] Re: Contractions near a slash > > > > > > Lars, > > > > > > The slash is actually ambiguous. It can be either a mathematical > > > symbol or a punctuation mark. The former is more accurate. It is > > > used a lot between words in English also, but usually doesn't cause > > > problems. The effects could be more severe in other languages. The > > > solution is not obvious. Any change can cause worse problems, as you > > discovered. > > > > > > John > > > > > > On Wed, Feb 25, 2009 at 01:22:14PM +0100, Lars Bj rndal wrote: > > > > "John J. Boyer" <johnjboyer@xxxxxxxxxxxxx> writes: > > > > > > > > > Lars, > > > > > > > > > > I have been thinking about this problem. One solution would be > > > > > to define "han/hun" with its own table entry, like defining > > > > > contractions that cover more than one word. This would be like > > > > > either/or in English, where the table entry would be word > > > > > either/or 15-24-456-34-123-1235 Note that this problem does not > > > > > occur with and/or, because "and" is always contracted. There may > > > > > need to be a change in the code, but I'm not sure what. A change > > > > > might introduce worse problems than we have now. > > > > > > > > Thank you. The problem with defining expressions like han/hun, is > > > > that people may put whatever words they want infront and behind > > > > the slash sign. So it will break a lot of situations. > > > > > > > > Lars > > > > > On Tue, Feb 24, 2009 at 09:21:51PM +0100, Lars Bj rndal wrote: > > > > >> Hello! > > > > >> > > > > >> While still working on the Norwegian liblouis tables, I found > > > > >> an error, that I need some help to fix: > > > > >> > > > > >> The example is as follows: "han/hun" (he slash she). When using > > > > >> xml2brl, the words "han" and "hun" is not contracted when the > > > > >> slash is inbetween. In the table No-No-g0.utb, the slash > > > > >> character is defined as math, which I suppose is correct. So > > > > >> what's the trick here to get "han/hun" correct through liblouis? > > > > >> > > > > >> Lars > > > > For a description of the software and to download it go to > > > > http://www.jjb-software.com > > > > > > -- > > > My websites: > > > http://www.godtouches.org > > > http://www.jjb-software.com > > > Location: Madison, WI, USA > > > > > > For a description of the software and to download it go to > > > http://www.jjb-software.com > > > > > > For a description of the software and to download it go to > > > http://www.jjb-software.com > > > > -- > > John J. boyer; President, Chief Software Developer JJB Software, Inc. > > http://www.jjb-software.com > > Madison, WI USA > > Developing software for people with disabilities > > > > For a description of the software and to download it go to > > http://www.jjb-software.com > > > > For a description of the software and to download it go to > > http://www.jjb-software.com > > -- > John J. Boyer, Executive Director > GodTouches Digital Ministry, Inc. > http://www.godtouches.org > Madison, Wisconsin, USA > Peace, Love, Service > > For a description of the software and to download it go to > http://www.jjb-software.com > > For a description of the software and to download it go to > http://www.jjb-software.com -- John J. boyer; President, Chief Software Developer JJB Software, Inc. http://www.jjb-software.com Madison, WI USA Developing software for people with disabilities For a description of the software and to download it go to http://www.jjb-software.com