Opening binary files in Fortran vs Python

Question

I have code written in Fortran and I would like to convert it to Python. I understand the flow of the code but when I wrote the code in Python, the outputs of the codes did not match.

So, after debugging I found out that the difference is coming from the first part of the code and I don't understand it. I think the way in which the two codes open and read the file are different because I don't think there is any problem with the seek() command. My file is a binary file and it is too large to attach it here.

Here are the first 10 outputs I am getting from Python: [1]: https://i.stack.imgur.com/KGXlI.png

Here is the first 10 outputs from Fortran: [2]: https://i.stack.imgur.com/wvlui.png

The first part of the two codes are pasted below:

open(1,file='ca48_exp_010.mvmelst',ACCESS='STREAM')
call fseek(1,8,0)  ! offset 8 bytes from beginning
nsection=0
continueReading= .true.;

do while(continueReading)
    read(1) sectionHeader   
    print *, sectionHeader

Python version:

with open('ca48_exp_010.mvmelst',mode='rb') as f:
    f.seek(8,0)
    nsection = 0
    continueReading = True
     
    while continueReading:
        sectionHeader = f.read(4) # read 4 bytes at a time

        if not sectionHeader:
            continueReading = False     

        sectionHeader = int.from_bytes(sectionHeader, byteorder="big")
        print(sectionHeader)

Have a look here https://stackoverflow.com/a/60296072/2836621 — Mark Setchell, Apr 03 '23 at 15:00
@MarkSetchell would you say that this question is a duplicate of that one? If so, can you hammer it? I added the [tag:python] tag, so I can't. — MattDMo, Apr 03 '23 at 15:10
@MattDMo, certainly not _obvious_ duplicates. That linked question is about Fortran sequential access, this one is about Fortran stream access — francescalus, Apr 03 '23 at 15:13
We really do need to see more detail about the input and (erroneous against expected) output. For example, we have no real reason to believe that `read(1) sectionHeader` is doing what the comment claims. — francescalus, Apr 03 '23 at 15:15
Where do you update `continueReading` in the Fortran version? — Barmar, Apr 03 '23 at 15:30
Do not use `fseek` to jump in Fortran stream access files. There is the `pos=` specfier for that in `READ(unit, POS=something)`. You should really show how your variables are declared. See [mcve]. — Vladimir F Героям слава, Apr 03 '23 at 16:16
Do not show text output as pictures. Copy and paste the numbers as text. — Vladimir F Героям слава, Apr 03 '23 at 16:16
@francescalus section header is an 4-byte integer declared at the beginning like: integer*4 sectionHeader — cr_007, Apr 03 '23 at 16:18
@francescalus the input file is a hex dump that I am reading in python — cr_007, Apr 03 '23 at 16:25
Little endian `5426` is big endian `840237056`, which I guess is somewhat significant given `byteorder="big"` in your Python. — francescalus, Apr 03 '23 at 16:55

Opening binary files in Fortran vs Python

0 Answers0