[sword-devel] OSIS example

Karl Kleinpaste karl at kleinpaste.org
Sun Mar 4 17:09:20 MST 2012

Daniel Owens <dhowens at pmbx.net> writes:
> But BibleTime, Xiphos, and BibleCS simply do not display the text of
> Canonical psalm titles. That is a severe bug.

Xiphos obtains headings using the canonical reference
mod.getEntryAttributes()["Heading"]["Preverse"]["0"] as passed through
mod.RenderText().  Period.

Troy's recent thread about pre-verse divs came out of a discussion he
and I had in #xiphos that morning, over the fact that Xiphos would not
display KJV Psalm pre-verse headers.  Turned out that the underlying
problem was merely that 2.3 displays fine but I had been using 2.4 out
of CrossWire Beta.  The 2.4 markup is different (newer, more modern)
exactly as regards pre-verse divs, and apparently reflects markup not
yet handled by the filters.  Perhaps an update using yet-unsupported
capability should not have gotten as far as Beta.

Convince RenderText() not to destroy the content found via
getEntryAttributes(), and all will be sweetness and light.

> One last note, variants do not work correctly in any front-end, which
> tells me the markup is not right. If anyone can provide a correction
> to the markup, please let me know.

In principle, Xiphos supports variants just fine, as I've seen it
deployed for ThML modules; I know this factually because 6 of my modules
use variants.  Cf. TischMorph in Luke 3:35; Luke 24:2; 2Cor 12:11; Rev
12:5; Rev 14:18.

A check of my mods.d/*.conf shows available modules using variants are:
2TGreek (ThML)
Byz (OSIS)
Tisch (OSIS)
TischMorph (ThML)

Ummmm...  In a fit of freakishness I had not anticipated at all, I just
discovered 2 minutes ago that Xiphos has never offered variant support
for OSIS at all; the conditional to find variants looked only for ThML.
This has apparently been an oversight since The Dawn Of Net.Time.  I've
fixed it and am running the updated build, so now I can select primary
-vs- secondary again from Xiphos' context menu.  But doing so has no
effect on the content, which looks...odd.  See screenshot from Tisch:


The secondary, which places a leading accent on Ραγαῦ, is displayed
/along with/ the primary, as though it was a sort of lemma.  This is
very wrong, and is very specifically a filter problem.

More information about the sword-devel mailing list