[sword-devel] ERV issues in And Bible

Wed Mar 8 14:31:46 MST 2017

Hello, David.

It sounds like my previous reply did not reach you.

I understand that the OSIS file that you extract from my module is flawed. However, those flaws do not exist in the OSIS file I used to create the same module. The milestone IDs in the OSIS source document are perfect in their uniqueness, although slightly more verbose than they need to be. Note that none of my milestones start with the characters "pp". The milestone IDs you get from mod2imp are flawed. I don't maintain mod2imp, and it is not used in the module creation process. I use osis2mod. It is
entirely possible that osis2mod is causing the problem. I do use osis2mod in the module creation process. Is osis2mod causing the problem? If the copy of osis2mod in the Ubuntu distribution is flawed, then that needs to be fixed. In the mean time, is there a non-broken version of osis2mod I should use? Do I need to build it from source?

Again, just for clarity, I understand that the OSIS you derive from my module has milestone IDs that are duplicated, but the OSIS I built that module from does not have that problem. Something else (osis2mod or mod2imp) is introducing those errors. I don't maintain those, but I'm happy to rebuild the eBible.org repository with a corrected osis2mod if that will fix the problem.

On 03/07/2017 10:51 PM, David Haslam wrote:
> Hi Michael,
>
> Not sure why my analysis is not getting through.
>
> It's quite feasible for an OSIS XML file to validate to a schema and yet to
> remain 'unfit for purpose'.
>
> There are things that the schema does not test for but which are required
> for use in SWORD, etc.
>
> One thing a schema cannot test for is the bad use of milestones:
>
> * Milestone elements in the wrong order (end before start, or overlaps
> between verses, etc)
> * Milestone elements that are completely missing, but which are a needed
> half of a pair
> * Milestone element pairs that do not have the same ID number but which
> ought to
> * Milestone elements that are duplicated
>
> Your module has OSIS that has 503 milestoned *div* elements that are
> *duplicated*.
>
> This probably validates to your MOSIS schema, but it will cause serious
> issues with JSword (based on Martin's evidence) and possibly also for SWORD
> apps too.
>
> Finally, the /milestoned/ for of the *div* element is not part of OSIS 2.1.1
> nor in the latest CrossWire schema.
> https://crosswire.org/wiki/OSIS_211_CR
>
> That might or might not give other problems, but it's probably not the root
> cause of And Bible crashes.
>
> btw. Did you get my email?
>
> Just in case the attachment couldn't be easily retrieved from the message
> here in the mailing list, I sent it you also by private email. 
>
> The IMP file for the whole Bible was fed into a simple filter to extract and
> count all the *div* elements with *type="x-milestone"*. The count is the
> first field in the output file.
>
> Every item for which count=00002 is suspect, and there are 503 of them.
>
> NB. The sorted order in the file is largely irrelevant.
>
> Best regards,
>
> David
>
>
>
>
>
> --
> View this message in context: http://sword-dev.350566.n4.nabble.com/ERV-issues-in-And-Bible-tp4656923p4656931.html
> Sent from the SWORD Dev mailing list archive at Nabble.com.
>
> _______________________________________________
> sword-devel mailing list: sword-devel at crosswire.org
> http://www.crosswire.org/mailman/listinfo/sword-devel
> Instructions to unsubscribe/change your settings at above page

-- 
signature

Aloha,
*/Michael Johnson/**
PO BOX 881143 • PUKALANI HI 96788-1143*• USA
mljohnson.org <http://mljohnson.org> • Phone: +1 808-333-6921 • Skype: kahunapule