Created attachment 146 [details]
the bug document

The language id of the text that has an emphasis mark set determines what kind
of emphasis mark is shown.

e.g. The code for a dot under a chinese peoples republic (vr) character is the
same as that for a dot over a chinese taiwan character, only the language is
different. Select the text in word and change its language to see the mark
change as well.

I also see that word renders some symbols which are the same for different
languages such as the dot over chinese peoples republic and japanese which are
the same symbol, but is rendered smaller in japanese and is located higher above
the character.

For the import and export I should be able to get the currently set language id
for the text and choose the correct emphasis mark from that.

But this will be a bit tricky for import because at the time we process the
emphasis mark the language might not have been set. So we will have to search
for the three sprms that can set the character language, and it the language is
not found then get it from the default style of the character.

Fix checked in.

Good in 630c except for one mark, it turns out that the diacrital over simple
and singapore chinese is actually a comma not a dot, word just clips out most of
the mark at a large font size!

Fix checked in.

All emphasis marks looking correct in 631b. It appears that the asian language
setting is the one that determines what mark will be shown. 

So even with a westerntext the asian selection for the language is the one that
determines what gets shown, the languages which affect it should just be
japanese, korean, and the two chinese ones, traditional and simplified, with
taiwan and macau being traditional, and the peoples republic and singapore being
simplified. I do not think there are any other languages which matter, but its
something to watch out for.

Looks good in src633a.

Closed, because fixed.