Tadziu asked me to post my solution, so here goes. Essentially, I
searched sequentially for ", "; the successful result gave me their
position in the string. The four levels of search were brought about
by the fact that in the English names for the days of the week three
have 6 characters, one has 7, two have 8, and one has 9.
Note, the nonce variable FIND*COMMA must be redefined at the
commencement of each level of the search; I suppose I should have
deleted it at the end.
Thanks to all for your interest and suggestions,
Robert
.ds DATE*FULL Monday, 29 January 2018
.\" Do NOT amend the following by hand.
.\" The variables will be created automagically.
.\" The RHS (date) of the second line of the Header Continuation text.
. ds DATE*SHORT \*[DATE*FULL]
. ds FIND*COMMA \*[DATE*FULL]
. substring FIND*COMMA 6 7
. ie '\*[FIND*COMMA]', ' .substring DATE*SHORT 8
. el \{\
. ds FIND*COMMA \*[DATE*FULL]
. substring FIND*COMMA 7 8
. ie '\*[FIND*COMMA]', ' .substring DATE*SHORT 9
. el \{\
. ds FIND*COMMA \*[DATE*FULL]
. substring FIND*COMMA 8 9
. ie '\*[FIND*COMMA]', ' .substring DATE*SHORT 10
. el \{\
. ds FIND*COMMA \*[DATE*FULL]
. substring FIND*COMMA 9 10
. if '\*[FIND*COMMA]', ' .substring DATE*SHORT 11
. \}
. \}
. \}
