[sword-devel] USFM conformance in usfm2osis.py

[Kahunapule Michael Johnson]

On 08/01/2012 09:17 PM, Chris Little wrote:
> End tags like \fqa* make little difference. Without the relaxed processing, the only difference my script makes is to not delete these end tags.

Simply deleting \fqa* might not have the intended effect. For example,
\f + xxxx could also be translated \fqa yyyy \fqa*or \fqa zzzz\fqa*.\f*
does not mean the same thing as
\f + xxxx could also be translated \fqa yyyy or \fqa zzzz.\f*
but does mean the same thing as
\f + xxxx could also be translated \fqa yyyy \ft or \fqa zzzz\ft .\f*

> But if someone is still using \ss to mark a subsection, I'd like the person running the converter to be cognizant of the fact that they're working with legacy markup and I want them to have to make a conscious decision to go on using a USFM converter on non-USFM docs. 

I agree that this merits some sort of nonfatal warning.