sbsdev / dtbook2sbsform

Convert DTBook to SBSForm, an intermediate format to produce formatted braille
GNU Lesser General Public License v3.0
1 stars 0 forks source link

Use of dtbook:abbr #13

Closed egli closed 4 years ago

egli commented 5 years ago

Currently dtbook:abbr is used to mark up text that not only happens to be an abbr but also the following white space that is to be dropped for Braille.

This is a problem for some of the other converters as the abbr element was not meant for white space:

The current practice has been there for a long time and is encouraged by the preptools who seem to include whitespace inside the abbr element.

I guess we have 3 options:

  1. Live with this and try to cope with it by fixing the xml before sending it to converters other than braille
    • This might work in the current converter landscape
    • not sure if this will work if we create multiple renditions for example
  2. Fix all other converters to cope with this
  3. Use the abbr element correctly. This would imply
    • changing the preptools
    • training for the transcribers and most likely
    • another way to markup space to be dropped in Braille
egli commented 5 years ago

See http://redmine.sbszh.ch/issues/1414 and sbsdev/dtbook-preptools@9cb348fde843e1f7bd1059cb0889d01568c4cbf4

egli commented 5 years ago

If we'd change the usage of abbr we might have to enforce it possibly via some schematron

egli commented 5 years ago

Well we really have many more options:

  1. "Fix" the xml before sending it to converters other than braille

  2. Fix all other converters

  3. Use the abbr element correctly and use an entity for the space following it

    • this space can then be dropped in Braille
    • and rendered as non-breaking space in Large Print
    • this requires a (relatively simple) change in the preptools
  4. Use the abbr element correctly and no longer collapse spaces between some abbrs and numbers

  5. isn't such a great solution, 4. will not be accepted by the internal Braille standards group. 3. looks like it could be a workable solution.

As for the entity to choose, we want something that has non-breaking characteristics in the case of large print and collapsing or zero-width in the case of braille. Figure space ( ) might come pretty close.

egli commented 5 years ago

Use &csp; (collapsible/conditional space)