STET: Allow fully specifying bolding of words in certain languages

[linearcombination]

One of the STET source languages that’s been labored on over the past
several months is Tok Pisin. Sorry if I’ve already asked about this…

With Tok Pisin, there are fewer total words in the language than in
most languages. At least one or two of maybe two dozen central words
(some are bilong, gutpela, long, man, mekim, nogat, pasin, samting,
and wanpela) seem to be in every single sentence.

So it won’t be as easy as just giving you an input file with the 4th
column showing which words should show up in bold anytime they appear
within those verses, because the same simple word or two-word phrase
can have so many different meanings depending on how they’re combined
with the other simple words around them.

There might also be a whole phrase that’s needed to convey one
spiritual term, but then one word or two words in the middle might not
belong to that phrase and instead be part of the meaning of a
different part of the sentence!

Or words from that same little phrase might show up later in the verse
and not have anything to do with the term.

In this kind of case, can you customize super-exactly to each verse
what shows up as bold in the source text?

This is made up, but let’s say these two verses have the bolded words
as meaning “adopted.”

Rom 8:23
I no ol dispela samting tasol. Nogat. Yumi tu i bin i gat namba wan
pikinini kaikai bilong Spirit yumi tu i save krai insait long bel na
wet i stap long God bai kisim mipela olsem pikinini bilong em na kisim
bek dispela bodi bilong bagarap.
Rom 9:4
Em ol lain Israel. Ol i gat pasin bilong papa i kisim ol bek, na ol i
gat glori na ol i gat kontrak na ol i gat presen bilong lo na ol i gat
pasin bilong lotu long God na ol i gat ol promis.

So you see that there’s a big gap between the words in the first
verse.

And in the second verse, some words are bold but then they appear
later in the verse in the same sequence but don’t have anything to do
with the term when they appear that second time.

Can you think of a way to handle this kind of thing - maybe by a
differently treated input file with very specifically bolded words in
the source text column?

[linearcombination]

Here is the now implemented spec for this feature:

Process to prepare a STET input document for a language, e.g., tpi,
that needs bolding to be fully specified:

1. The third column of a STET input document is normally where just a
   reference and any '(*)' comments are stored. However, for languages
   where the bolding needs to be fully specified with no ambiguity the
   third column needs to hold both the reference and the passage text
   encoded in the following manner:

   • The bible reference, e.g., Rom 4:23, will be delimited by
     opening and closing, and respectively, xml-like markers.

   • The passage text will be delimited by opening and closing,
     and respectively, xml-like markers.

   • Within the passage text, i.e., between and , can be n
     number of non-overlapping bolding indicators which will be
     delimited by opening and closing, and respectively,
     xml-like markers.

   • Thus a full specification for a particular bible reference and
     passage text looks like:

     Rom 3:23Passage text goes here. There can multiple
     bolding spans in one passage. For instance here is another
     one and another.

   • There can be n number of these full passage specifications in the
     3rd column for a particular row, e.g.,

     Rom 3:23Passage text goes here. There can multiple
     bolding spans in one passage. For instance here is another
     one and another
     Rom 9:12-15Another passage here with some bold
     words and some more bold words.

   • It looks best for readers of the STET input document if each
     fully specified passage beginning occurs on a new line within the
     3rd column for a particular row (as above example shows).

Please note that fully specifying passages in the 3rd column is only
necessary for languages that need this level of specificity in
bolding. Languages that don't should still only provide the reference
in the 3rd column (plus any of the comment nomenclature as per usual),
and bolding, if any, in an optional 4th column (as per usual).

STET is smart enough to detect when a STET input document uses the
fully specified passage format in the 3rd column. Any language's STET
input document that uses the fully specified verse format in the 3rd
column will be treated in this way. As before, STET is smart enough to
detect languages that provided bolding in the 4th column.

If the STET input document uses the fully specified passage format in
column 3, it should not also provide a 4th column in which bolded
words are specified.

Tips for preparation of the STET input document, e.g., tpi, in which
the fully specified passage format will be used:

• Use STET to generate the output document for the language in
  question. This will provide the localized (when possible) references
  and the passage text. You can use this to update the new STET input
  document for the language in question, e.g., tpi, to fill the 3rd
  column:

  • Replace the non-localized reference with the localized referenced
    you just copied and enclose it in . Then type a and
    directly after it paste the passage text for that passage
    reference and then type the closing delimiter .

  • Repeat this for all references in the 3rd column.

  • Then go back to each fully specified passage in the 3rd column and
    indicate what spans of words you want bolded by enclosing such
    spans in and .

  • Provide the updated STET input document to be added/updated in
    STET by the developer and on a subsequent update this feature will
    then be supported by the STET system.