<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">On 07/23/2012 01:12 PM, Peter von
Kaehne wrote:<br>
</div>
<blockquote cite="mid:20120723171256.133470@gmx.net" type="cite">
<pre wrap="">
</pre>
<blockquote type="cite">
<pre wrap="">Von: DM Smith
Mark titles with x-preverse.
</pre>
</blockquote>
<pre wrap="">
DM, this is exactly what we do at the moment - see svn/sword-tools/modules/misc_cleanup/titles_cleanup.pl (or some such path, sorry not at home) Not terribly successful, as it causes all kinds of problems with titles which are not preverse, but nevertheless. Without it titles behave totally arbitrary - some toggle, many do not.</pre>
</blockquote>
Which SVN revision are you using. I may be able to comment further.
What actually makes its way into the module is what matters.<br>
<br>
The SWORD engine recognizes two kinds of pre-verse markup. The
x-preverse on titles and the milestone div. The filter needs some
change in handling the milestone div properly. The latter is the
point of this thread.<br>
<br>
If a module solely uses x-preverse titles with something other than
the latest osis2mod, then everything (depending on which version of
osis2mod your using) that osis2mod deems as otherwise preverse
won't/mightn't be properly marked. And will be rendered after the
verse number. This is the cause of the orphaned verse number.<br>
<br>
The work around for your perl magic may be to figure where to split
between verses and put the markup into the title, if there is one
and remove it otherwise.<br>
<br>
<blockquote cite="mid:20120723171256.133470@gmx.net" type="cite">
<pre wrap="">
I appreciate the suggestion with -d 2 and re-running the import. This is new to me and might indeed be helpful. It is endlessly frustrating to get to the bottom of the whole pre-verse debacle. </pre>
</blockquote>
Note: you'll need to convert these from <milestone resp="v"
.../> to <verse ...> to reimport with osis2mod. Importing
with vpl2mod might have better success.<br>
<br>
The suggestion is at least helpful to know what is being put into
each slot.<br>
<br>
The milestone representing the verse start is a diagnostic help.
SWORD heading filter knows nothing about it and will ignore it if it
makes its way into a module. Likewise for JSword. However, if you
see something other than a title before it, realize that SWORD will
pull the title (marked x-preverse) out of it. The rest will be
treated as verse content. The front-end puts the verse number
between what SWORD identifies as the title (or preverse material)
and the verse content.<br>
<br>
If the milestoned preverse div is used to wrap the entirety (as with
the latest osis2mod in SVN), then you'll get different undesirable
behavior.<br>
<br>
If you have titles splitting a verse (as David outlined in another
response.) Note that it will be pulled out of its position and that
position is not remembered. It will be put into the array of
pre-verse titles. (I hope I'm wrong about this.)<br>
<br>
<blockquote cite="mid:20120723171256.133470@gmx.net" type="cite">
<pre wrap="">
As an aside, you seem to bemoan that all kinds of other material is toggled by the "title" switch. I thought this was desired and I certainly did not think this was unusual. In fact I always saw the toggle as one which switches between naked canonical text and text + "extras" - and considered the "title" name for the switch simply an anachronism.
</pre>
</blockquote>
Most BSP modules put structural markup (sections, paragraphs and
poetry) between verses. This is commonly part of the canonical
markup.<br>
<br>
Example (-- denotes the boundary between SWORD module verses as
determined by osis2mod):<br>
--<br>
<div type="section"><br>
<title>First section title</title><br>
<p canonical="false"><br>
Description of the section.<br>
</p><br>
<p><br>
verse 5<br>
--<br>
verse 6<br>
<meta http-equiv="content-type" content="text/html;
charset=ISO-8859-1">
--<br>
verse 7<br>
</p><br>
--<br>
<p><br>
verse 8<br>
--<br>
verse 9<br>
--<br>
verse 10<br>
--<br>
verse 11<br>
</p><br>
</div><br>
--<br>
In this case, only <br>
<title>First section title</title><br>
<p canonical="false"><br>
Description of the section.<br>
</p><br>
is non-canonical markup. But everything before the verse is
pre-verse.<br>
<br>
It is the text between verses that is (other than Psalm titles)
non-canonical. These may have their own structural markup and thus
that would also be non-canonical.<br>
<br>
So if all markup and content before the verses is remove, then the
whitespace (especially vertical) can change dramatically. In the
example the paragraph starts disappear.<br>
<br>
Hope this makes sense.<br>
<br>
In Him,<br>
DM<br>
</body>
</html>