[sword-devel] Av11n modules - was Deuterocanonical books / Apocrypha
dmsmith at crosswire.org
Mon Feb 3 15:19:41 MST 2014
Maybe I misunderstood what Chris said/meant (reading his post now, apparently I did). The osisID book names can be found here: http://crosswire.org/wiki/OSIS_Book_Abbreviations
As to numbering, I just know how the v11n work in SWORD and JSword. They assume that chapters start at 1 and verses start at one.
I suppose that it is possible to give 0 as the number of verses in a chapter in the verse per chapter array.
Regarding mid-verse titles, I'm not sure if SWORD handle it or not. That's a different issue from v11ns.
Regarding the plethora of v11ns that translators use, that is a hard one to solve well. Do note, that all the text in an input to osis2mod should appear in the module. Nothing should be thrown away. Hopefully it will be appended to the end of a verse that precedes it.
Verse ranges should be handled properly. This has been part of osis2mod for a very long time. As a comment, OSIS does not allow the - in an osisID but does in an osisRef. So Genesis 1:1-3 should be written osisID="Gen.1.1 Gen.1.2 Gen.1.3". That said, I think osis2mod is lenient allowing an osisRef where an osisID is expected.
The handling of a missing verse in the WEB as verse containing only a footnote is good. Other works do the same. So is dropping the verse altogether from the numbering. Shorter chapters pose no problem for a v11n that allows for more verses.
Longer chapters cause verses to be appended to the right location.
There's a lot more to be done to solve *all* the problems.
In His Service,
On Feb 3, 2014, at 4:38 PM, Kahunapule Michael Johnson <Kahunapule at eBible.org> wrote:
> To the best of my knowledge, there are no Bible book name errors in the OSIS files I generate. There might be some ambiguity in the interpretation of book names for Daniel and Esther from the Greek Septuagint vs. the fragments that only contain the parts that the Hebrew Bible lacks. If I'm mistaken (which is entirely possible), please tell me which books are in error, and what the current and proper spellings are. Then I can fix them.
> Chapter numbering of books in the Apocrypha is a bit of an enigma. I can renumber Psalm 151 to be chapter 1 of the book AddPs. How are chapter numbers in the fragments of Daniel from the Septuagint (Bel & the Dragon, etc.) that aren't in the traditional Hebrew Bible handled in Sword? What about Esther? What is your recommended best practice for these? In real life, these don't start at chapter 1, either. Or, perhaps it may be better to skip these and just include the whole books of Daniel and Esther from the Greek (i.e., essentially those from the LXX2012 as a first draft)?
> I've been contemplating stripping noncanonical section titles from the OSIS files that I generate, because I get the feeling that SWORD can't properly handle these if they happen midverse (and the frequently do), and that it isn't likely to change in the near future. What do you think?
> As for versification variations, those are a fact of life. The number of versification systems we will encounter in real world Bible will always be greater than any number you hard code. If all that is necessary is to pick a system that is close, and have no fear of problems resulting from a few intentionally missing or appended verses, or from verse bridges (i.e. verse 1-3 as one unit), then that isn't a major problem. While I agree that versification systems are not of as great a value if they aren't standardized, there is no underestimating or restraining the creativity of Bible translators in about 7,000 languages and cultures and who knows how many church traditions. There is no way that a small group of programmers (or a large group, for that matter) is likely to change much in the way of these traditions or translational decisions. That said, the World English Bible's versification was affected slightly by The Sword Project at an early stage, back when the Sword Project was KJV-only. The versification of the WEB NT is based on the Greek Majority Text NT, which has a few differences between that and the KJV. For those spots where verses are "missing", a verse is there containing nothing but a footnote explaining the difference. When push comes to shove, though, translators will stick by their versification decisions and expect the programmers to adapt.
> <MichaelHead64.jpg> Your partner in electronic Bible publishing,
> MICHAEL JOHNSON
> 1215 S KIHEI RD STE O # 728
> KIHEI HI 96753-5225
> Verizon Wireless Mobile: +1 808-333-6921
> Skype: kahunapule or +1 719-387-7238 eBible.org
> On 02/03/2014 08:06 AM, DM Smith wrote:
>> It may not be obvious, but SWORD and osis2mod is really picky when it comes to Bible book names.
>> Chris notes that the WEB OSIS file has errors in the Bible book names. These have to be fixed for osis2mod to not complain about something it does not know.
>> Also, the chapter and verse numbers for each book starts at 1 and goes up from there. So, the note that AddPs.151 is AddPs.1 in SWORD means that AddPs starts at chapter 1. There is no chapter 151.
>> Also, Chris is probably correct about the seg fault. I have encountered a seg fault when osis2mod encounters a book name it does not recognize. I've also encountered seg faults for chapters not in the av11n. Actually, SWORD library seg faulted.
>> With those fixed, the NSRVA might be very close.
>> Hope this helps.
>> In His Service,
>> On Feb 2, 2014, at 6:22 PM, Chris Little <chrislit at crosswire.org> wrote:
>>> On 02/02/2014 06:57 AM, Barry Drake wrote:
>>>> On 02/02/14 11:20, Barry Drake wrote:
>>>>> Correct me if I'm wrong, but currently the only modern English
>>>>> Language version I can see that contains the deutero-canon is the CPDV
>>>>> version. I'm not at all sure about this version, partly because it is
>>>>> a bit ponderous to read, and also because it is the work of only one
>>>>> person rather than a group of translators. Michael mentions the WEB
>>>>> version as having the extra books - the Sword module I can find
>>>>> contains only the 66 book canon. Is there anything else out there
>>>> I thought I'd have a play. I downloaded the osis file from
>>>> http://ebible.org/ and using the command: osis2mod ./
>>>> eng-webbe_osis.xml -z -v KJVA and the following as a .conf file:
>>>> LCSH=Bible. English.
>>>> DistributionLicense=Public Domain
>>>> Description=World English Bible with Apocrypha
>>>> I made a working module. The osis2mod error messages show that neither
>>>> KJVA nor any of the other currently available 'with'apocrypha'
>>>> versifications are correct. As an example, 4 Maccabees is ignored
>>>> completely .... there are a host of other unsupported verses. I wonder
>>>> if there is any chance someone could add a WEBA versification (or
>>>> explain to me very gently how to make one?) Chris Little maybe?
>>>> God bless, Barry.
>>> There is no possibility that I would consider adding a versification system for the WEB. It fits none of the criteria warranting a new versification system.
>>> NRSVA should work fairly well, but there are some inconsistencies here & there. For example Ps 151 is called AddPs.151 in the WEB but AddPs.1 in Sword, and EpJer is called EpJer.6 in the WEB but Bar.6 in Sword. I'm not sure what's causing you a segfault, but I think it is sometimes caused by unrecognized book names. I'm currently working on improving osis2mod, so I'll keep an eye out for that issue.
>>> Alternately, Kostya made a suggestion a while back that might serve translations with quirky versifications like this. I'll comment on that suggestion when I get an opportunity to bump it for visibility.
>>> sword-devel mailing list: sword-devel at crosswire.org
>>> Instructions to unsubscribe/change your settings at above page
>> sword-devel mailing list: sword-devel at crosswire.org
>> Instructions to unsubscribe/change your settings at above page
> sword-devel mailing list: sword-devel at crosswire.org
> Instructions to unsubscribe/change your settings at above page
-------------- next part --------------
An HTML attachment was scrubbed...
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 4145 bytes
Desc: not available
More information about the sword-devel