# Refactoring code into a subroutine **URL:** \ **Category:** Uncategorized\ **Created:** [January 12, 2021, 1:15am UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566 "2021-01-12T01:15:31Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![Beliavsky](https://avatars.discourse-cdn.com/v4/letter/b/ba8739/32.png) [@Beliavsky](https://fortran-lang.discourse.group/u/Beliavsky)\ **Post date:** [January 12, 2021, 1:15am UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566/1 "2021-01-12T01:15:31Z")
I often create a subroutine by - pasting a block of code into an empty subroutine - compiling with IMPLICIT NONE - copying declarations from the source file the code was copied from. Sometimes what the compiler thinks is an undeclared variable is actually a function which needs to be imported from a module. - making some variables INTENT(IN) or INTENT(out) arguments I wonder if there are any tools to automate this. Something that did 80% of the job would still be useful. A Windows batch file that uses Unix tools to list undeclared variables is ``` @ echo off :: list undeclared variables in Fortran source file, compiled with gfortran :: usage: gfundec.bat foo.f90 setlocal set gfout=temp_gf.txt gfortran -c -Wall %1 2> %gfout% echo undeclared in %1: grep IMPLICIT %gfout% | tr -d "'" | cut -d ' ' -f 3 ``` When run on the code ``` module foo_mod implicit none contains subroutine bar() ! code below copied from another source file ndays = size(xx,1) nret = ndays - 1 nsym = size(xx,2) allocate (profits(nret,nsym),xpos(ndays)) do nma2=nma2_min,nma2_max write (*,"(/,a,1x,i0)") "nma2",nma2 xpos = merge(1.0_dp,-0.5_dp,rollmean_ratio(xx(:,1),1,nma2) > 1.0_dp) profits = spread(xpos(:nret),2,nsym) * xret if (.false.) call print_stats(cstats_ret,profits,sym,title="profits") call print_tangent_port(profits,sym,cstats_ret=cstats_ret,print_uncon=print_tangent_portfolio, & print_long_only=print_tangent_port_long_only) end do end subroutine bar ``` it says ``` undeclared in temp_mac_tangent.f90: cstats_ret ndays nma2 nma2_max nma2_min nret nsym print_tangent_port_long_only print_tangent_portfolio profits sym xx ``` ---
**Author:** ![Arjen](https://avatars.discourse-cdn.com/v4/letter/a/b9bd4f/32.png) [@Arjen](https://fortran-lang.discourse.group/u/Arjen)\ **Post date:** [January 12, 2021, 9:08am UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566/2 "2021-01-12T09:08:52Z")
You could pass the report through a small, almost trivial, Fortran program to generate default declarations: - Variable names that start with one of i … n are probably meant to be integers (given that they are not declared makes that likely) - Variable names that start with something else are probably reals The program could also make a copy of the code and insert these declarations and do the other boilerplating stuff (module … end module, etc.) Running it through the compiler again will alert you to the flagrant mistakes: taking a substring from a real variable for instance. ---
**Author:** ![billlong](https://avatars.discourse-cdn.com/v4/letter/b/71e660/32.png) [@billlong](https://fortran-lang.discourse.group/u/billlong)\ **Post date:** [January 12, 2021, 3:52pm UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566/3 "2021-01-12T15:52:58Z")
Instead of IMPLICIT NONE you might consider IMPLICIT NONE(TYPE,EXTERNAL) that would require declaration of functions as well. Note that an IMPLICIT NONE in the specification part of a module also applies to any contained subprograms in the module, which explains the error messages you got for subroutine bar. ---
**Author:** ![andrew\_4619](https://avatars.discourse-cdn.com/v4/letter/a/e68b1a/32.png) [@andrew\_4619](https://fortran-lang.discourse.group/u/andrew_4619)\ **Post date:** [January 13, 2021, 5:38pm UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566/4 "2021-01-13T17:38:21Z")
> [@billlong](#): > > IMPLICIT NONE(TYPE,EXTERNAL) What compiler supports that? Ifort still chokes . ---
**Author:** ![billlong](https://avatars.discourse-cdn.com/v4/letter/b/71e660/32.png) [@billlong](https://fortran-lang.discourse.group/u/billlong)\ **Post date:** [January 13, 2021, 6:03pm UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566/5 "2021-01-13T18:03:38Z")
The Cray compiler supports it (along with most of F2018). Intel claims ifort supports 2018. Are you using an old version? A newish Intel compiler seems to work: % /opt/intel/oneapi/compiler/2021.1.1/mac/bin/intel64/ifort -c test.f90 % cat test.f90 implicit none (type,external) end program % ---
**Author:** ![epagone](https://yyz2.discourse-cdn.com/free1/user_avatar/fortran-lang.discourse.group/epagone/32/252_2.png) [@epagone](https://fortran-lang.discourse.group/u/epagone)\ **Post date:** [January 13, 2021, 6:27pm UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566/6 "2021-01-13T18:27:30Z")
The Eclipse IDE photran plug-in has a [refactoring set of functionalities](https://help.eclipse.org/2020-12/topic/org.eclipse.photran.doc.user/html/advanced/Refactoring.html) that do what you are after and more. ---
**Author:** ![andrew\_4619](https://avatars.discourse-cdn.com/v4/letter/a/e68b1a/32.png) [@andrew\_4619](https://fortran-lang.discourse.group/u/andrew_4619)\ **Post date:** [January 13, 2021, 11:42pm UTC](https://fortran-lang.discourse.group/t/refactoring-code-into-a-subroutine/566/7 "2021-01-13T23:42:28Z")
Latest ifort. It “supports” it but it does not work.