[sword-devel] KJV 2006 - 3rd beta

[DM Smith]

I have readied another beta. Please take a look at it for problems with 
these changes and prior ones. There is still more to be done (fixing 
punctuation and spelling mistakes) so no need to report these yet.

You can get the module from here: 
http://www.crosswire.org/~dmsmith/kjv2006 (The download links are at the 
bottom of the page.)

Changes from the last beta:
* Fixed 2 problems noted with the last beta, both regarding the merging 
of the definite article into the following <w> element.
* Paragraph markers are now positioned at the beginning of a verse.
* Whitespace has been fixed and normalized:
       A space follows each verse.
       <w> elements that contained leading punctuation and spacing have 
had it moved to immediately before it.
       <w> elements with trailing punctuation and space have had it 
moved to immediately after it.
       Empty <w> elements followed by spaces and punctuation have had it 
moved before the element.
       Newlines and tabs have been replaced with spaces.
       Spaces before certain punctuation marks have been removed. (e.g. 
? ! , : ; ....)
       Multiple spaces have been reduced to a single space.
* Apostrophe problems have been fixed.
       There were 17 different patterns of problems that were solved. 
Here are a few:
       Fixed problems when a word ended with ' and it should have ended 
with 's.
       Fixed problems when two words were run together and these were 
separated by a '.
       Fixed problems where the ' or 's was not inside the containing 
element (e.g. <w>dog</w>'s became <w>dog's</w>)
       Fixed problems where a space should have followed a ' but it did not.
      
Note: All of these changes were programmatic. There may be some that 
need to be undone. These will be identified when the text is compared 
with other etexts.

Next step:
Create verse per line etexts from different sources.
Compare them with wdiff.
Fix obvious errors.
Catalog other errors.
Release another beta.

In doing this effort I have discovered that a KJV that you buy today, 
probably, won't agree with the KJV 1796. This makes it problematic in 
producing an accurate 1796 version. I've been doing some research on the 
subject, and it appears that the "Old Scofield" or an early version of 
the Thompson's Chain Reference (version 1 -3) is probably the closest. 
As the latter is out of print, I will be purchasing the former and using 
it to do tie-breaking of reported errors. Another possible reference is 
http://www.interleafbible.com/ and it's etext at 
http://printkjv.ifbweb.com/.

So the next beta will not fix minor differences in spelling. That will 
be a following beta.

Many thanks to Lynn Allan, Tim Lanfear and Miklos Zsido who provided 
long lists of problem verses. These have been indispensable.