If the suggestion/underline covers more than one cohort, we need to use the casing from that whole string, not just the cohort of the reading with the error tag.
If input is
"<A.>"
"A" N Sg <NoSpaceAfterPunctMark> &no-space-after-punct-mark ID:1 R:RIGHT:2
"A" N Sg <NoSpaceAfterPunctMark> "<A. Eira>" &no-space-after-punct-mark &SUGGESTWF ID:1 R:RIGHT:2
"<Eira>"
"Eira" N Prop &LINK &no-space-after-punct-mark ID:2
we want &SUGGESTWF to produce A. Eira (as if it had the <fixedcase> tag).
gets called with inputCasing that's set by inputCasing = getCasing("A.") which is considered uppercase, should be called with getCasing("A.Eira") which is mIxed (and leads to no case change).
If the suggestion/underline covers more than one cohort, we need to use the casing from that whole string, not just the cohort of the reading with the error tag.
If input is
we want &SUGGESTWF to produce
A. Eira
(as if it had the<fixedcase>
tag).But
gets called with
inputCasing
that's set byinputCasing = getCasing("A.")
which is considered uppercase, should be called withgetCasing("A.Eira")
which ismIxed
(and leads to no case change).