How do you get DOS/UNIX GSAS raw data into VMS format?

Brian Toby - E-mail: Brian.Toby@NIST.gov

Date: Mon, 09 Nov 1998 12:15:19 -0500
From: "Brian H. Toby" 
Organization: NIST Center for Neutron Research
To: RIETVELD_L Distribution List 
Subject: ASCII GSAS file formats

Here are the differences between the ASCII GSAS files on different
platforms (the ASCII files are the experiment, raw data and instrument
parameter files)

In all cases these files are 80 character, fixed-record length files but
Windows/DOS, UNIX and VMS have different ideas of what that means.

  VMS uses internal information in the file header to decide the contents 
  of a file; it is usually easy to move the file to another system since 
  most methods of transfer (ftp for example) convert it to a standard 
  file type on the receiving computer, but when copying a file to VMS, 
  you need to use a conversion utility. I typically use the 
  CONVERT/FDL=80.FDL/PAD command where 80.FDL has the appropriate 
  commands to make an 80 character fixed record length file (try EDIT/FDL 
  80.FDL to create). Note that older versions of GSAS on VMS used ISAM 
  access to .EXP files but the most recent uses 80 character, fixed-record 
  length direct access files, but the older version could do its own 
  conversions internally.

  UNIX uses a file with 80 characters per record and no line termination 
  so that the file appears to most other software as one very long line. 
  These files tend to be garbled by e-mail. The convstod and convdtos 
  programs convert standard ascii to direct-access (-stod) or the other 
  way round (-dtos). However, the versions in the GSAS package will 
  trash files if used on the wrong file type 
  [this was my sloppy programming -- see below for a fix].

  Windows/DOS uses a file with 80 characters per record followed by a 
  carriage-return & line-feed (so really 82 characters per record). 
  This does e-mail well, but some software may add or remove some of 
  the blank characters from each line, which GSAS cannot tolerate. 
  The CNVFILE program in GSAS will read standard ASCII files 
  (but not UNIX GSAS files) and pad them with the spaces needed to 
  make them the correct length.

I now have a better version of the convstod and convdtos program
(ftp://ftp.ncnr.nist.gov/pub/cryst/) 
[CCP14 Mirror] which does not trash files
when used incorrectly and where the sequential files are Windows/DOS
compatible. I have used it on UNIX for some time, but I suspect it works
fine when run in Win/DOS to read UNIX direct access files and convert
them to Win/DOS. 


See ftp://ftp.ncnr.nist.gov/pub/cryst/ 
[CCP14 Mirror] for comments on compilation and use.

N.B. pretty much all of the rest of the files used in GSAS are binary
and should never be moved from one platform to another, but that is not
a problem, since a powder experiment can be moved from one computer to
another by copying the .EXP file and the data files and then reformating
them appropriately.


Xiang Ouyang wrote:
> 
> Hi Dr. Cline,
> I would like to have a copy of your paper about the  GSAS refinement.
> BTW, what is the different for the raw data format for VMS, PC and UNIX
> when you use GSAS? Our data are in VMS format, have to use the convert
> before it run on the PC, and always get BOMBED when I tried to run the
> data in SGI.
> Thank you very much.
> 
> Xiang Ouyang
> Department of Chemistry
> Texas A & M University
> College Station,TX 77843

********************************************************************
Brian H. Toby, Ph.D.                    Leader, Crystallography Team
Brian.Toby@NIST.gov      NIST Center for Neutron Research, Stop 8562
voice: 301-975-4297     National Institute of Standards & Technology
FAX: 301-921-9847                        Gaithersburg, MD 20899-8562
********************************************************************