Conversation
@mcave - thanks for the report, could you send the file to thomas.robitaille@gmail.com? |
Thanks for the prompt answer. I had to create a link because the file is too big: |
@mcave - I had a first look at this but could not find any easy solution. In order to help with the debugging, could you let me know what kind of variable is in the file? Could you reduce it to the simplest example that produces a failure? |
I think the problem are the pointers:
I send you by email the file. Marco |
@mcave - thanks for sending this over! Just to help me debug, how do you view the content the pointer is pointing to in IDL? I tried:
|
This is just null pointer (http://www.exelisvis.com/docs/IDL_Pointers.html). It's dirty, I know, but it is how IDL works. Would it be possibly to assign simply an empty numpy array for this cases? Thank you |
Ah, I see! I'll try and understand how IDL is encoding that information to see how to deal with it. Maybe simply using the Python |
You are right, it sounds better. Thank you again |
@mcave - I think I figured it out, but just to test, can you send me a new file that includes this variable, followed by a variable named 'check' that is set to the value of 5? (integer) |
Sent! |
@mcave - thanks for the file, and sorry for the delay! The code attached to this issue now fixes this. I'll wait to see if the tests pass. I'll also fix the same issue in scipy.io.idl (which is the same as idlsave). |
Exception: VARSTART is not 7
@mcave - this should now be fixed. Try updating to the latest developer version of idlsave - could you let me know if your files work fine now? |
It does work! Thank you fow your help |
@mcave - and thanks for your patience! |
I did encountered again same problem as another issue. I'm trying to read a structure.
Thanks