# Fortran90 reading in a formatted input file

• Fortran
Hi All,

Let me preface by saying this is my first post, but I'm very glad to have found and joined what looks like a great community here. Also I have no Fortran experience, so please bear with my greenness:

I'm trying to resurrect and use a code written by another engineer in our organization some years ago. I have the following code for opening and reading an input file, now my question is how the input file should look to feed into this correctly.

print*,'Opening / Reading Input File: ...'
print*
open(10,file='input',status='unknown',form='formatted')
do i=1,ntimes
enddo
close(10)

I'm sorry if this is very basic; I'm very new at this!

Thanks!

Hi All,

Let me preface by saying this is my first post, but I'm very glad to have found and joined what looks like a great community here. Also I have no Fortran experience, so please bear with my greenness:

I'm trying to resurrect and use a code written by another engineer in our organization some years ago. I have the following code for opening and reading an input file, now my question is how the input file should look to feed into this correctly.

print*,'Opening / Reading Input File: ...'
print*
open(10,file='input',status='unknown',form='formatted')
do i=1,ntimes
enddo
close(10)

I'm sorry if this is very basic; I'm very new at this!

Thanks!

What you can try is switching your read statements leading up to your DO/ENDDO loop with write statements and seeing what the file looks like. It would look like

Code:
open(10,file='test.txt')
write(10,*)a
write(10,*)b
write(10,*)c,d
write(10,*)e,f,g,h
write(10,*)i
write(10,*)j
write(10,*)ntimes
close(10)

With arbitrary values assigned to a-j and ntimes.

Just an idea.

Thanks for that; it did show me how basic of an input I could use.

I am having some trouble with this in debugging though. I get a ".exe triggered a breakpoint" error at the line of the first read statement. I'll do some more headscratching and then return with more details if I can't get it.

Thanks again.

Mark44
Mentor
"Triggering a breakpoint" isn't an error. It seems that you set a breakpoint at that statement, which causes the debugger to stop.

rcgldr
Homework Helper
"Triggering a breakpoint" isn't an error.
It could be an error, if the compiler generates breakpoint instructions (INT 3) or uses an invalid address to access memory in a failure case.

It's possible that the open statement failed. Is there anyway to check the status of the open statement?

rcgldr:

I think you might be close to home w/ suspecting the open statement. In Visual Studio 2005 I get a little green arrow pointing at the first read statement. When I hover over that for detail, I get the following message:

"This code has called into another function. When that function is finished, this is the next statement that will be executed."

And when I go to disassembly, I've got a little yellow arrow pointing at the line just after a line that reads:

"00431340 int 3"

I don't really know yet how to check the status of the open statement.

Thanks to all for the support.

Mark44
Mentor
rcgldr:

I think you might be close to home w/ suspecting the open statement. In Visual Studio 2005 I get a little green arrow pointing at the first read statement. When I hover over that for detail, I get the following message:

"This code has called into another function. When that function is finished, this is the next statement that will be executed."

And when I go to disassembly, I've got a little yellow arrow pointing at the line just after a line that reads:

"00431340 int 3"
This is the breakpoint instruction that rcgldr mentioned.

Have you set a breakpoint? If so, it will show up as a red circle in the left margin in Visual Studio. You can delete all breakpoints by clicking Delete All Breakpoints in the the Debug menu in VS. This menu item doesn't appear if there are no breakpoints set.
I don't really know yet how to check the status of the open statement.

Thanks to all for the support.

I think the compiler is automatically setting a breakpoint or something...I got a disassembly display and a call stack that mean nothing to me and I think I've hit a wall with self-diagnosis. If anyone can suggest a new lead or what I could post here to clue us in, I'll gladly oblige.

Mark44
Mentor
AFAIK, the compiler doesn't automatically set breakpoints.

Whatever, your open statement doesn't look right to me, especially that form = "..." part. Here's a link to a page that has several examples of open statements. Try modifying your open statement and see if that makes a difference.

http://www.livephysics.com/computational-physics/fortran/fortran-file-handling.html

For additional examples, you can do what I did, which was enter "fortran open" in the browser.

AFAIK, the compiler doesn't automatically set breakpoints.

Whatever, your open statement doesn't look right to me, especially that form = "..." part. Here's a link to a page that has several examples of open statements. Try modifying your open statement and see if that makes a difference.

http://www.livephysics.com/computational-physics/fortran/fortran-file-handling.html

For additional examples, you can do what I did, which was enter "fortran open" in the browser.

In addition to the link Mark44 provided, I have also found http://www.math.hawaii.edu/~hile/fortran/fortmain.htm" [Broken] to be infinitely helpful.

Last edited by a moderator:
Mark44
Mentor
Here's another link to a page with more than you want to know about the OPEN statement - http://www.hpc.unimelb.edu.au/doc/f90lrm/lrm0545.htm [Broken].

This page describes the FORM parameter, which I didn't remember using (it's been a long while since I did any Fortran programming...).

I don't know if you noticed it, but there's an extra space in the code you showed (in 'format ted'). If that's actually in your code, it's probably what's causing the compiler to complain.
Rowdy5000 said:
Code:
open(10,file='input',status='unknown',form='format ted')

Last edited by a moderator:
Thanks to everyone for the useful links and support with this. I actually resolved this finally by toggling status in the open statement to 'old'. It didn't really make sense to me based upon what I'd read about the status character, but I tried this and it worked.