Re: [Open Babel] Limits on number of input files

[Chris M. <c.m...@ga...>]

Michael Fisher wrote:
> I’ve started using the OpenBabel GUI.  I have two directories with .mol 
> files I want to get SMILES strings for.  One has about 3500 files, the 
> other >13000.  There appears to be an input file limit of about 700 
> files.  Obviously, I could do 700 at a time, but I then have to 
> concatenate all the files (at least twenty for the big directory).  Am I 
> right that there’s a limit?  Is there a way to get around it?  

It seems to work for me.

I used the recently released v2.2.1 GUI on Windows XP, although I 
don't think there was any change which should make it different from 
v2.2.0. Not having several thousand SMILES files to hand, I took a 
file with 3603 molecules in SMILES format and did a smi to smi 
conversion to output file A*.smi. This splitting operation produced 
3603 files, A1.smi ... A3603.smi

I then did another smi to smi conversion from A*.smi to recombined.smi 
and got file containing all 3603 molecules. I think this is the sort 
of agglomeration operation you wanted to do, and it worked ok.

I also tried doing another smi to smi conversion from A*.smi to 
B*.smi and got another 3603 files BA1.smi ...BA3603.smi. So a batch 
conversion also worked ok.

There seems to be a bug in displaying the input files when there is a 
wildcard character, but this occurs only when the mousewheel is 
operated, and isn't size related.

I don't see any reason why the Windows command line shouldn't do these 
operations ok, and putting the input file name in quotes, as Noel 
mentioned, should get over the embedded spaces.

I think your problem may be an erroneous SMILES in the 701st file. Try 
  checking "Continue with next object after error, if possible" or -e 
on the command line.

Chris

> 
> FYI, the filenames contain numerous spaces, periods and commas, so the 
> command line version of babel (which I’ve tried) crashes.  Also, I’m 
> presently stuck doing this in Windows XP.  If I switch to Fedora, will 
> the command line version ignore punctuations?  Will there still be a 
> limit on input files in the GUI?