https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107721
--- Comment #19 from anlauf at gcc dot gnu.org --- (In reply to Christopher Albert from comment #18) > Created attachment 62909 [details] > Next try > > How about this solution? Yes this appears to be very close... ;-) (And it is well readable.) I am sorry to torture you further, but here's another mean case with nested constructors which loses the typespec: print *, [ character(16) :: ['a','b'] ] print *, [ character(16) :: ['a','b'] ] // "|" print *, [[character(16) :: ['a','b'] ]] ! lost typespec print *, [[character(16) :: ['a','b'] ]] // "|" ! lost typespec print *, "|" // [[character(16) :: ['a','b'] ]] ! lost typespec end code: WRITE UNIT=6 FMT=-1 TRANSFER (/ 'a ' , 'b ' /) DT_END WRITE UNIT=6 FMT=-1 TRANSFER (/ 'a |' , 'b |' /) DT_END WRITE UNIT=6 FMT=-1 TRANSFER (/ 'a ' , 'b' /) DT_END WRITE UNIT=6 FMT=-1 TRANSFER (/ 'a |' , 'b|' /) DT_END WRITE UNIT=6 FMT=-1 TRANSFER (/ '|a ' , '|b' /) DT_END
