This is the technical description of a DBT Translation table. If you want more general information about languages and template choices, please see the list of templates.

Initially, the language table for braille translation is determined by the selected template, and may be changed using the Document / Translation Tables menu. Using those menus does not require use of the table designator. However, to switch to a different translation table partway through a file, one must enter a DBT code and the designator for the table to switch to. For switching secondary languages within a base language table, see the [lng~X] command. For switching from one base language to another, see the [lnb~...] command.

The French tables support print-to-braille translation of French-language literary text and mathematical and computer notation, following the codes and customs established by the agreement reached in July 2005 at the 3rd International Forum at INJA in Paris, by an expert group comprising delegates from France and other French-speaking countries, as interpreted for Duxbury Systems by the Association Valentin Hauy (AVH), Paris.

Braille-to-print translation is supported for this language. However, braille-to-print translation may not be perfect, therefore beware that errors can occur. If you find errors or have suggestions, please send both the *.dxb and *.dxp files along with an explanation to: languages@duxsys.com. Please be sure to include sample files!

True braille-to-print translation is supported within French text, but not in technical notation. This means that any mathematics or computer-notation portions of a braille file will not translate correctly to print. It also means that the "translated line" will typically contain gibberish when the cursor is in mathematics or computer notation.

Uncontracted English and Spanish are supported as secondary languages.

In addition, it is possible to switch to any of the available translation tables listed in DBT. (See the [lnb~...] code below.)

The mathematics-oriented technical notation code is defined in "Notation Mathematique," published by AVH. The computer-notation code, "Informatique," is also defined by AVH.

In addition, it is possible to switch to any of the available translation tables listed in DBT (see the [lnb~...]code below), many of which do support various technical codes, such as for mathematics or computer notation, or which support “unified” treatment of technical notation as well as literary text in the base language associated with the table.

The following DBT translation codes are available when using the French (Unified) table.Any other translation codes used will be ignored, or indeed may cause unexpected results.If using an alternative translation table, i.e.when switching to another base language table by means of the [lnb~...] code, please refer to the relevant topic and available codes for that table.

[/]

[ab] is equivalent to [g2]

[bar]

[beN] where N is a number in the range 0 through 4 is a special variant of [be] to end a subscript with "forced" treatment; see [bsN] below.

[be]

[bline] -- ignored.

[bsN] where N is a number in the range 0 through 4 is a special variant form of the [bs] code that forces treatment of a subscript as (0) simple, (1) medium, (2) high, (3) framed by parentheses, or (4) framed by brackets. (If treatment is not forced, the simple treatment is used if it can be determined that such treatment is OK and otherwise parenthesis framing is used. Note that these forcing codes are subject to change in future versions.)

[bs]

[caplvN] N must be either 2 or 3:

N = 2 means that capitals are indicated in "basic" form, i.e.at most one dots 46 indicator on a word whether the full word or just the initial letter is capitalized.

N = 3 means that full capitalization indicators are used.

[cb] is a synonym for [iq], which see below.

[cbi] is a synonym for [iq], which see below.

[cbn] is a synonym for [iq], except that no indicator is generated.

[cz]

[cz;f] is like [cz]except that the ASCII characters correspond to braille according to French conventions (as of 2008).

[e]

[feN] where N is 0 or 1 is a special variant form of the [fe] code used when forcing the denominator to have (0) "simple" or (2) "framed" treatment of a fraction; see [fsN] and [flMN] below.

[fe]

[flMN] where M and N are each 0 or 1 is a special variant form of the [fl] code used when forcing both the numerator and denominator to have simple or framed treatment. For example,

[fs0]numerator[fl01]denominator[fe1] would force the numerator to have simple treatment and the denominator to have framed treatment. (If forcing is not used, simple treatment is used if it can be determined that such treatment is clearly OK, otherwise the framed treatment is used. Note that these forcing codes are subject to change in future versions.)

[ftlvN] N must be either 1 or 3:

N = 1 means that all emphasis markings ([fts~...] ... [fte~...] codes) are ignored.

N = 3 means that all emphasis markings are honored, i.e.indicated in the braille.

[fl]

[fsN] where N is 0 or 1 is a special variant form of the [fs] code used when forcing the numerator to have (0) "simple" or (2) "framed" treatment of a fraction; see [feN] and [flMN] above.

[fs]

[fte~b]

[fte~i]

[fte~u]

[fts~b]

[fts~i]

[fts~u]

[g1] switches to "grade 1" (uncontracted) braille.

[g1l] switches to "grade 1" (uncontracted) braille and "locks" that setting.

[g1u] undoes the "locking" effect of a prior [g1l], while leaving the contraction grade as "grade 1" (uncontracted) braille.

[g2] switches to "grade 2" (contracted) braille (which is the normal mode for this table).

[g2l] switches to "grade 2" (contracted) braille and "locks" that setting.

[g2u] undoes the "locking" effect of a prior [g2l], while leaving the contraction grade as "grade 2" (contracted) braille.

[i]

[ii]

[i0] (0 = numeric zero)suppresses the automatic addition of a letter sign and allows the following word to be contracted, or not, according to the grade currently in effect

[i1] (1 = numeric one)is equivalent to [i]

[in] is equivalent to [g1]

[iq] begins "informatique" (computer braille); see also [tx] below.

[ixrtdN] where N is 0, 1 or 2 is a special variant of [ixrtd] used for forcing (1) simple, (2) medium or (3) framed treatment of the radicand. (If forcing is not used, simple treatment is used if it can be determined that such treatment is clearly OK, otherwise the framed treatment is used. Note that these forcing codes are subject to change in future versions.)

[ixrtd]

[ixrteN] where N is 0, 1 or 2 is a special variant of [ixrte] used for forcing (1) simple, (2) medium or (3) framed treatment of the radicand. See [ixrtdN], above.

[ixrte]

[ixrtsN] where N is 0, 1 or 2 is an allowed special variant form of [ixrts] that is always equivalent to [ixrts]; see [ixrtdN], above.

[ixrts]

[lnb]

[lnb~...] (for switching to another base [primary] language table)

[lng~en]

[lng~es]

[lng~fr] or [lng]

[oeN] where N is a number in the range 0 through 4 is a special variant of [oe] to end an "over" index with "forced" treatment; see [osN] below.

[oe]

[osN] where N is a number in the range 0 through 4 is a special variant form of the [os] code that forces treatment of an "over" index as (0) simple, (1) medium, (2) high, (3) framed by parentheses, or (4) framed by brackets. (If treatment is not forced, the simple treatment is used if it can be determined that such treatment is OK and otherwise parenthesis framing is used. Note that these forcing codes are subject to change in future versions.)

[os]

[peN] where N is a number in the range 0 through 4 is a special variant of [pe] to end a superscript with "forced" treatment; see [psN] below.

[pe]

[psN] where N is a number in the range 0 through 4 is a special variant form of the [ps] code that forces treatment of a superscript as (0) simple, (1) medium, (2) high, (3) framed by parentheses, or (4) framed by brackets. (If treatment is not forced, the simple treatment is used if it can be determined that such treatment is OK and otherwise parenthesis framing is used. Note that these forcing codes are subject to change in future versions.)

[ps]

[sqrteN] where N is 0, 1 or 2 is a special variant of [sqrte] used for forcing (1) simple, (2) medium or (3) framed treatment of the radicand. See [sqrtsN], below.

[sqrte]

[sqrtsN] where N is 0, 1 or 2 is a special variant of [sqrts] used for forcing (1) simple, (2) medium or (3) framed treatment of the radicand. (If forcing is not used, simple treatment is used if it can be determined that such treatment is clearly OK, otherwise the framed treatment is used. Note that these forcing codes are subject to change in future versions.)

[sqrts]

[tce] cancels the effect of [tcs], restoring literary treatment (which is the default).

[tcs] is required in this table to enable math (technical) context, i.e.for general treatment of technical material including the surrounding text. See the Topical How-To Guide, under Help, for general instructions.

[te] cancels the effect of [ts], restoring normal text mode (even though "technical context" may remain in effect).

[ts] initiates "technical notation," i.e."math mode" within technical context.

[tx] cancels "informatique" (computer braille -- see [iq]) or "no translate" (see [cz]), reverting to ordinary text.

[txn] is like [tx]) except that no indicator is generated.

[ueN] where N is a number in the range 0 through 4 is a special variant of [ue] to end an "under" index with "forced" treatment; see [usN] below.

[ue]

[usN] where N is a number in the range 0 through 4 is a special variant form of the [us] code that forces treatment of an "under" index as (0) simple, (1) medium, (2) high, (3) framed by parentheses, or (4) framed by brackets. (If treatment is not forced, the simple treatment is used if it can be determined that such treatment is OK and otherwise parenthesis framing is used. Note that these forcing codes are subject to change in future versions.)

[us]

[vrn]cancels the effect of [vrn~nuc], reverting to Antoine treatment of numbers (which is the initial and default treatment with this table)

[vrn~nuc] switches to "classical" treatment of numbers

[xcs]

The table is designed to work with the following groups of characters:

All ASCII printable characters

Accented characters and punctuation marks typical of French, German, Italian, and Spanish

British pound and other miscellaneous signs (DUSCI page D+ee...)

Mathematical signs, shapes, etc. (in "technical context" only) (DUSCI pages D+e2..., D+e5..., D+ef..., D+f0..., D+f1...)

Note that, for historical reasons and subject to change in future versions, certain printable characters are reserved for "forcing" symbols and indicators, namely:

D+60 (accent grave): "letter sign" (grade 1) indicator (dot 6) (note: if followed by space or [<], this forces a line break with dot 5 continuation indicator)

D+f528 (diaeresis): capital indicator (dots 46)

D+9b (cent sign): indicator for start of continuous (multi-word) capitals

D+9d (yen sign): indicator for one word in full capitals D+9f (small script f with hook): indicator for last word in continuous (multi-word) capitals D+f9 (middle dot): forces "abbreviation point" instead of full stop

The above is a general guide only (see "General Notes" section at the beginning of this document).

The development of the literary portions of these tables commenced in March 1987, under the sponsorship of the Association Valentin Hauy (AVH), Paris, and as a joint technical effort of Duxbury Systems, Inc. and AVH. The primary specification for French literary braille usage with contractions is "Abrege Orthographique Etendu," a publication of AVH."Table de Transcription pour la Production du Braille Abrege par Ordinateur," by Michel Jacquin and published by AVH (1986) was also instrumental in this early development.

An "informatique" (computer notation) code was also included in the early work, and revised considerably in 1999-2000 in response to a change in specifications.

The braille-to-print translation tables were added starting in November 1990, with the same sponsors and developers.

Mathematics translation facilities, following "Notation Mathematique" as prepared by AVH, were added in 1999-2000, with AVH again providing sponsorship and collaborating in the technical work with Duxbury Systems, Inc.

Starting in February 2006, with AVH's continued sponsorship and guidance, these tables were modified for the Unified French rules agreed to in July 2005 at the 3rd International Forum at INJA in Paris, by an expert group comprising delegates from France, French-speaking African countries, Belgium, Switzerland and Quebec.

Likewise, starting in April 2007, under the same auspices, Duxbury has been working to implement the revisions to the French math code.

(Documentation reviewed: July 2010.)