Follow-up Comment #11, bug #63176 (group groff): [comment #9 comment #9:] > this might be because a trap macro has forgotten to invoke > a request with the no-break control character.
Happily for debugging, -me sets a very limited number of traps, and they
involve only four macros.
$ egrep '\.\s*(wh|dt|it|blm|lsm|em)\b' tmac/e.tmac
. wh 0
.wh -\\n(_bu @f
.wh -\\n(_Bu @r
.wh 0 @h \" reset header
.wh 0 @h \" set header
.em @z
$ egrep 'de\s+@[frhz]\b' tmac/e.tmac
.de @z \" --- end macro
.de @h \" --- header
.de @f \" --- footer
.de @r \" --- reprocess overflow footnotes
I think we can rule out the header macro as the culprit, and your -r@2 output
in comment #3 shows that the @r macro is never invoked (which is to be
expected, as bad-footnote.me produces no overflow footnotes). Thus, if your
hypothesis is correct, this leaves only @f and @z as suspects.
However, given that the input produces one page with one footnote, I think
it's weird that the trace shows @f, the footnote macro, being called twice,
the second time _after_ @z, documented as the "end macro" and planted with the
.em request.
_______________________________________________________
Reply to this item at:
<https://savannah.gnu.org/bugs/?63176>
_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/
signature.asc
Description: PGP signature
