[tyndale-devel] missing tags in ChiUns text

[David F. Haslam]

Hi David,

The only time to watch out for is as and when you use my described
method to rebuild a module from an exported (then edited) IMP file.
The IMP file should be searched to find if there are occurrences of a
verse [or successive verses] with /identical text/ to the previous verse.
/This task can easily be scripted with any programming language that you
prefer./

Whether or not the OSIS tags are properly formed for verse ranges is
related, but the nub of the issue is the /duplication/
which occurs with _all three Sword utilities_ that are capable of
exporting content from a module.

  * diatheke <http://crosswire.org/wiki/Frontends:Diatheke> (which is a
    useful front-end app in its own right, albeit a command line driven one)
  * *mod2imp* (now the main focus of this email thread)
  * *mod2osis* (which is 'sort of' deprecated, due to it being incapable
    to achieve a round trip with osis2mod)


Best regards,
David

On 2012-11-20 14:47, David Instone-Brewer wrote:
> (I don't know if you've already got this - I found it unsent in my Out
> box)
>
> Thanks David.
>
> Yes, we will be using lots of modules.
>
> On your link I found an example:
>
> <verse eID='Exod.22.2-Exod.22.4'/>
>
>
> So I searched a bunch of modules (about 130)  for RegEx   [0-9]-[A-Z]
> I found verse spans in lots of dictionaries & commentaries
> and in Bibles with notes & XRefs (eg ESV, LEB, Luther, NET, MHC)
> But I didn't find any using verse spans like that described.
>
> Looks like something we should watch out for, but it won't occur often.
>
> David IB
>
>
> At 13:28 20/11/2012, David F. Haslam wrote:
>> Hi David,
>>
>> The warning was mainly geared to modules other than ChiUn & ChiUns.
>> /I just thought you might have more stuff in the pipeline for Tyndale
>> STEP,
>> so I didn't wish to leave you imagining that there were no pitfalls
>> in using the IMP format method./
>>
>> When I referred to a verse range, this is best illustrated in USFM
>> format.
>> The extreme case I mentioned need not concern you, but in Paratext
>> the line began as follows:
>>
>> \v 12-83 blah blah .....
>>
>> When converted from USFM to OSIS this sort of thing becomes much more
>> complicated.
>> I already described the principles of linked verse ranges in this
>> section of a wiki talk page:
>> http://crosswire.org/wiki/Talk:OSIS_Bibles#Linked_verses
>> <http://crosswire.org/wiki/Talk:OSIS_Bibles#Linked_verses>
>> /Albeit the reason for doing so was to correct a different
>> misapprehension arising in another translation conversion project.
>> /
>> Paratext verse range tagging can be a nightmare! We've often seen
>> things like this:
>>
>> \v 2-6a followed by the text for verse 2 through the first half of
>> verse 6
>> \v 6b-11 followed by the text for the second half of verse 6 through
>> verse 11
>>
>> Converting such examples from USFM to OSIS is not so easy - and some
>> workarounds were usually required until now.
>> Hopefully, Chris Little's new python script will cover such issues
>> properly - once it's been fully tested.
>>
>> These issues are unrelated to Alternate Versification
>> <http://crosswire.org/wiki/Alternate_Versification>. Default v11n for
>> SWORD modules is the KJV.
>> Your reference to Num.17 was therefore not relevant - this chapter
>> has 13 verses in both the KJV and the ChiUns Bibles.
>> Admittedly the WLC module (Hebrew) has 28 verses in Num.17 - but none
>> of these are a linked verse range in the sense being discussed.
>> cf.  The WLC module is configured with Versification=Leningrad
>>
>>
>> Best regards,
>>
>> David
>>
>> On 2012-11-20 11:54, David Instone-Brewer wrote:
>>> Thanks for the warning!
>>> I guess the verse range is in Num.17 where there is a severe
>>> mismatch between the KJV and Hebrew numbering.
>>> But presumably you don't mean that alternate versification makes it
>>> fall over
>>> Could you give me an example of the coding for a verse range so I
>>> know what to look out for?
>>>
>>> David IB
>>>
>>>
>>>
>>> At 10:36 20/11/2012, David F. Haslam wrote:
>>>> On 2012-11-19 14:47, David Instone-Brewer wrote:
>>>>> David,
>>>>>
>>>>> This is GREAT. I was concerned about mod2osis so this is really
>>>>> helpful.
>>>>>
>>>>> David IB
>>>>
>>>> Hi David,
>>>>
>>>> _Caveat_:  Beware of this problem:
>>>>
>>>> http://www.crosswire.org/tracker/browse/MODTOOLS-31
>>>> <http://www.crosswire.org/tracker/browse/MODTOOLS-31>
>>>>
>>>> This would affect modules that have linked verse ranges.
>>>> Rebuilding using imp2vs would thus lead to some duplicated verses
>>>> in the module.
>>>> One would require a workaround processing step to avoid this outcome.
>>>>
>>>> For ChiUn and ChiUns, probably not a problem, as they follow the
>>>> KJV v11n.
>>>> Only three KJV verses are not tagged in these 2 modules. See
>>>> attached output from emptyvss.
>>>>
>>>> Most more modern translations frequently have several or even many
>>>> Biblical passages tagged to a verse range.
>>>> cf. The longest example I've come across to date was one in which
>>>> verses 12-83 were given one tag for a passage somewhere in Numbers.
>>>> :'(
>>>>
>>>> David
>>>
>>>     David IB 
>>>     Â  
>>>     ///Â Â  Dr David Instone-Brewer 
>>>     dib   Senior Research Fellow in Rabbinics and the New Testament 
>>>     Â ^Â Â Â  Tyndale House, 36 Selwyn Gardens, Cambridge, CB3 9BA, UK 
>>>     \=/Â Â  Rabbinics at Tyndale.cam.ac.uk
>>>     <mailto:Rabbinics at Tyndale.cam.ac.uk>Â Â Â Â Â 
>>>     www.TyndaleHouse.com <http://www.tyndalehouse.com/> 
>>>
>>     David IB 
>>       
>>     ///   Dr David Instone-Brewer 
>>     dib   Senior Research Fellow in Rabbinics and the New Testament 
>>      ^    Tyndale House, 36 Selwyn Gardens, Cambridge, CB3 9BA, UK 
>>     \=/   Rabbinics at Tyndale.cam.ac.uk      www.TyndaleHouse.com
>>     <http://www.tyndalehouse.com/> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.crosswire.org/pipermail/tyndale-devel/attachments/20121121/ffdf8d2a/attachment-0001.html>