I need a little help: according to this site, there is a limit on the maximum number of continuation lines.
So I decided to test this. I wrote a jumbo FUNCTION
where it computes colossal algebraic formula expressed in statement split into 17,146 continuation lines.
!test.f90 1.6 MB file
DOUBLE COMPLEX FUNCTION myfunction(a, b)
DOUBLE COMPLEX, INTENT(IN) :: a
DOUBLE COMPLEX, INTENT(IN) :: b
myfunction = gd0/16.d0)*a*b*((a+b)**2)*((32.d0*DCONJG(f(4)))+(12&
&8.d0*DCONJG(f(11)))+(160.d0*DCONJG(f(24)))+(64.d0*DCONJG(f(46)))+(32.d0*DCONJG(f(3)))+(256.d0*DCON&
&JG(f(10)))+(480.d0*DCONJG(f(23)))+(256.d0*DCONJG(f(45)))+(32.d0*DCONJG(g(9)))+(128.d0*DCONJG(f(&
&9)))+(480.d0*DCONJG(f(22)))+(384.d0*DCONJG(f(44)))+(96.d0*DCONJG(g(21)))+(160.d0*DCONJG(f(21)))+(2&
&56.d0*DCONJG(f(43)))+(64.d0*DCONJG(g(42)))+(64.d0*DCONJG(f(42)))+(64.d0*DCONJG(f(8)))+(192.d0*DCON&
& (64.d0*DCONJG(g(42)) ! and so on and so forth...
END FUNCTION
I compiled this abomination with gfortran -c test.f90
, and it returned an 11.4MB test.o
file after 5 minutes, without any errors or warnings. I ran it, and it returned correct results.
Why isn't gfortran
observing the maximum continuation lines rule?