CODEWITHSUNDEEP
binaryasciinon-ascii-charactersidl-programming-language
EddyTheB
EddyTheB

Reputation: 3200

IDL: read ascii header of binary file

I'm having an enforced introduction to idl trying to debug some old code.

I have a binary image file that has an ascii header (It's a THEMIS IR BTR image of Mars, if that is of interest). The code opens the file as unit 1 using OPENR, then reads the first 256 bytes of it using ASSOC(1,BYTARR(256)). The return from that is 256 ascii character dex values, but they are mostly high or low numbers that do not correspond to alpha-numeric characters, and are not related to the header that I know is on the file.

One thing that may help with diagnostics: the original file is a g-zipped version of the file. If I try to open it directly (using less, for example) it allows me to read the header. But if I unzip it first (gzip -c filename.IMG.gz > filename.IMG) and then try to read it again I get binary gobbledegook. (less gives me a warning before opening: "filename.IMG may be a binary file. See it anyway?").

Any suggestions?

Here's the IDL code:

CLOSE,1
OPEN,1,FILENAME
A = ASSOC(1,BYTARR(256))
B = A[0]
print,'B  - ',B
H = STRING(B)
print,'H  - ',H

And this is what it gives me:

B  - 31 139 8 8 7 17 238 79 0 3 ... (and on for 256 characters) 
H  - [Some weird symbol]

I've tried it on a purely ascii test file and it works as expected.

Upvotes: 1

Views: 1182

Answers (2)

mgalloy
mgalloy

Reputation: 2386

Try the COMPRESS keyword to OPEN:

openr, 1, filename, /compress

The COMPRESS keyword refers to a compressed file, so it is both for reading and writing compressed files.

Upvotes: 1

LSerni
LSerni

Reputation: 57388

31 139 8 is the beginning of a GZIP header for a "deflated" file.

http://www.gzip.org/zlib/rfc-gzip.html#file-format

So yes, the file looks like it needs to be decompressed first.

Try decompressing the file with gunzip, and check the header again. If it is 31 139 08... again, it looks like it has been compressed twice.

Otherwise, whatever it is, it is likely that it's been finally decompressed. It remains to be seen why the uncompressed file isn't being decoded.

Upvotes: 1

Related Questions