I have a team who is getting lots of errors in the marker check: “Footnote caller is not the expected value (*)”.
I think I have solved the problem as I was writing this message, but I think it is worth still sending this in case others come across the problem.
The errors show only verses which have valid footnotes. That is, of the format
\f + \fr 2.23 \ft …\f*.
I confirmed with RegEx that the + was the only character used in valid footnotes. I used RegExPal to give me a count of the unique matches (\f . \fr). This confirmed that + was the only character used in valid footnotes.
The settings.xml has the +
+
Then just before sending this message I did another search in RegExPal (\f . .) and discovered that there were several hundred invalid footnotes in the New Testament that was done over a decade ago. These had manually defined callers (a b c …) but with no verse reference.
I still don’t know why Paratext was expecting * as the footnote caller, nor why only the valid footnotes were being reported as errors but at least I have a better understanding of the problem and assume the solution will be to correct the invalid ones so that it will accept the valid ones.
I think their second error “Most notes of this type do not have an origin reference.” is related. Once again it only lists what looks like valid footnotes. However, with 213 valid out of 700+ total footnotes, Paratext is correct - most of the footnotes don’t have an origin references, but they should! The valid ones are not the errors. In this case the majority are the errors with only a minority of correct ones.