 |
forums.silverfrost.com
Welcome to the Silverfrost forums
|
| View previous topic :: View next topic |
| Author |
Message |
christyleomin
Joined: 08 Apr 2011
Posts: 155
|
 Posted: Tue Apr 09, 2013 8:25 pm Post subject: Converting a FORTRAN code than doing manual changes |
 |
|
I have got a FORTRAN code which has several .F files.
I need to do some fixed changes to the code. I mean, there are certain common changes that I need to do in all the .F files.
If I do this manually in each file, it is a very time consuming process.
Can I write a code that reads each line from each .f FILE and correspondingly does the necessary changes I intend?
Is there any better way? |
|
| Back to top |
|
 |
PaulLaidler
Site Admin
Joined: 21 Feb 2005
Posts: 8210
Location: Salford, UK
|
| Posted: Wed Apr 10, 2013 8:44 am Post subject: |
|
|
Yes you can do this but usually it only helps if your task would take more than a few hours to do by hand.
An intermediate approach is to use an editor with a macro facility. It is often possible to record macros for repetitive tasks. I have found that learning how to do this has saved me a lot of time in the longer term.
Both Plato and Visual Studio have macro facilities as do other popular editors. |
|
| Back to top |
|
|
jalih
Joined: 30 Jul 2012
Posts: 196
|
| Posted: Wed Apr 10, 2013 9:24 am Post subject: Re: |
|
|
| PaulLaidler wrote: |
An intermediate approach is to use an editor with a macro facility. It is often possible to record macros for repetitive tasks. I have found that learning how to do this has saved me a lot of time in the longer term.
|
Same here...
Could you give us more information about what is needed to be done? |
|
| Back to top |
|
|
IanLambley
Joined: 17 Dec 2006
Posts: 506
Location: Sunderland
|
| Posted: Wed Apr 10, 2013 12:52 pm Post subject: |
|
|
I wrote a quick and dirty editor, originally for converting VAX Fortran with "Structure" to "Record" type data.
It takes a file called "edlist" which contains a string of up to 40 characters on the odd numbered lines and replaces them with the upto 40 character string on the line below. It also allows you to enter a 40 character "end string" for replacement, so that if it finds the first 40 character string and it is followed somewhere on the line, both strings are swapped for the new strings on the next edlist line. It is also recursive and once something has been swapped by one of the matched strings, it tries again with the other swapping specifications, but only carries out each swap once for each line.
Example of file "edlist": | Code: | structure / /
type
end structure
end type
record / /
type ( )
mystuff.
mystuff% |
and the program "Quickedit.for" is as follows: |
|
| Back to top |
|
|
IanLambley
Joined: 17 Dec 2006
Posts: 506
Location: Sunderland
|
| Posted: Wed Apr 10, 2013 12:55 pm Post subject: |
|
|
| Code: | character*40 swap_lead(10000),swap_tail(10000)
character*40 swap_lead_new(10000),swap_tail_new(10000)
character*1000 linein,lineout,line_test,filein
character*3 new_extension
logical*2 result
integer*4 isl(10000),ist(10000),isln(10000),istn(10000)
integer*4 already_swapped(10000)
call cmnam(filein)
call cmnam(new_extension)
Print *,'Processing ',trim(filein)
if(filein .eq. ' ')stop
open(unit=10,file='edlist',status='old')
ios = 0
nswap = 0
nchange = 0
do while(ios .eq. 0)
read(10,1000,iostat=ios,end=2222)swap_lead(nswap+1),
& swap_tail(nswap+1)
read(10,1000,iostat=ios,end=2222)swap_lead_new(nswap+1),
& swap_tail_new(nswap+1)
nswap = nswap + 1
isl(nswap) = leng(swap_lead(nswap))
ist(nswap) = leng(swap_tail(nswap))
isln(nswap) = leng(swap_lead_new(nswap))
istn(nswap) = leng(swap_tail_new(nswap))
call upcase@(swap_lead(nswap))
call upcase@(swap_tail(nswap))
if(isl(nswap) .eq. 0)nswap = nswap - 1
1000 format(2a)
2222 continue
enddo
close(unit=10)
|
|
|
| Back to top |
|
|
IanLambley
Joined: 17 Dec 2006
Posts: 506
Location: Sunderland
|
| Posted: Wed Apr 10, 2013 12:56 pm Post subject: |
|
|
Continued:
| Code: |
c print *,nswap
c read(5,*)
open(unit=10,file=filein,status='old')
open(unit=11,file='change.out',status='unknown')
ios = 0
do while(ios .eq. 0)
read(10,1000,iostat=ios,end=3333)linein
lineout = linein
line_test = linein
call upcase@(line_test)
already_swapped = 0
1111 continue
do i=1,nswap
ipos_lead = 0
ipos_tail = -1
ipos_lead = index(line_test,swap_lead(i)(1:isl(i)))
ipos_tail = ipos_lead
in_bracket=0
if(ipos_lead .gt. 0 .and.
& swap_lead(i)(isl(i):isl(i)) .eq. '(' .and.
& swap_tail(i)(1:1) .eq. ')' )then
c
c bracket match search required
in_bracket = 1
endif
if(ipos_lead .gt. 0)then
if(in_bracket .eq. 0)then
c
c normal
if(swap_tail(i) .ne. ' ')then
ipos_tail = index(line_test(ipos_lead+isl(i):),
& swap_tail(i)(1:ist(i)))
if(ipos_tail .gt. 0)then
ipos_tail = ipos_tail+ipos_lead+isl(i)-1
endif
endif
else
|
|
|
| Back to top |
|
|
IanLambley
Joined: 17 Dec 2006
Posts: 506
Location: Sunderland
|
| Posted: Wed Apr 10, 2013 12:57 pm Post subject: |
|
|
Continued:
| Code: |
ipos_tail = ipos_lead
do ii=ipos_lead+isl(i),leng(line_test)
if(line_test(ii:ii) .eq. '(')in_bracket=in_bracket+1
if(line_test(ii:ii) .eq. ')')in_bracket=in_bracket-1
if(in_bracket .eq. 0)then
c
c reached terminating string
ipos_tail = ii
goto 3344
endif
enddo
3344 continue
endif
endif
c print *,ipos_lead,ipos_tail,i,already_swapped(i)
c print *,trim(linein)
if(already_swapped(i) .eq. 0)then
if(ipos_lead .gt. 0 .and. ipos_tail .gt. 0)then
if(ipos_tail .ge. ipos_lead+isl(i) )then
c
c found a candidate for a change
lineout = linein(:ipos_lead-1)//
& swap_lead_new(i)(:isln(i))//
& linein(ipos_lead+isl(i):ipos_tail-1)//
& swap_tail_new(i)(:istn(i))//
& linein(ipos_tail+ist(i):)
linein=lineout
line_test = linein
call upcase@(line_test)
nchange = nchange + 1
already_swapped(i) = 1
goto 1111
else
lineout = linein(:ipos_lead-1)//
& swap_lead_new(i)(:isln(i))//
& linein(ipos_lead+isl(i):)
linein=lineout
line_test = linein
call upcase@(line_test)
nchange = nchange + 1
already_swapped(i) = 1
goto 1111
endif
endif
endif
enddo
write(11,1000)trim(lineout)
enddo
3333 continue
close(unit=10)
|
|
|
| Back to top |
|
|
IanLambley
Joined: 17 Dec 2006
Posts: 506
Location: Sunderland
|
| Posted: Wed Apr 10, 2013 1:01 pm Post subject: |
|
|
Continued:
| Code: |
if(new_extension .ne. ' ')then
call set_suffix@(filein,new_extension,result)
endif
open(unit=10,file=filein,status='unknown')
rewind(unit=11)
ios = 0
do while(ios .eq. 0)
read(11,1000,iostat=ios,end=4444)linein
c il = leng(linein)
c if(il .gt. 0)then
write(10,1000)trim(linein)
c else
c write(10,1020)
c 1020 format()
c endif
4444 continue
enddo
close(unit=10)
close(unit=11,status='delete')
print *,nchange,' changes'
end
|
Run from the command line as:
quickedit oldfile.f for
This will output a new file with the extension ".for".
Original file "oldfile.f" will be read and the edits made. output will be "oldfile.for"
If the new file extension is omitted then the original file will be overwritten.
No responsibility taken for any destruction and havoc it may wreak!
Ian |
|
| Back to top |
|
|
davidb
Joined: 17 Jul 2009
Posts: 560
Location: UK
|
| Posted: Wed Apr 10, 2013 5:20 pm Post subject: |
|
|
I would just use VIM, you can do pretty much anything in VIM once you have learnt it.
If I had many files to convert I would probably do something in Python.
_________________
Programmer in: Fortran 77/95/2003/2008, C, C++ (& OpenMP), java, Python, Perl |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
