Hi Egon and Peter,
I just did a 'head -4' on the our sample gaussian (output) files and
compared the results. I used -4, because from interpreting the
Resolver.java code, I deduce (hopefully correctly so) that the Resolver
only looks at the first 4 lines of the file (as long as they are
shorter than 16K or so?).
Each of them start with the ' Entering Gaussian System'...., except for
4-cyanophenylnitrene-Benzazirine-TS.g94.out, which has that on the 2nd
line. So they all resolve properly. These files seem to have been
created by Gaussian 94, 98, and 03 and on Linux/Unix type systems.
If it is indeed the case that Windows (his g03 output) does not put '
Entering Gaussian System' but ' Entering Link 1' in the first line in
stead, the problem will be solved (at least for that file, and possibly
all G03W files?).
The problem remains with the other file, which has 5 lines (the last of
them empty) before the real output because I don't believe the Resolver
ever reads far enough ahead to see what it should be seeing.... So
another solution needs to be found for this kind of circumstances...
May I suggest as a possibility for the application (since it won't work
for the applet) to have an 'interpret as' popup that allows one to
define what the resolver should be (the default selection would be
'automatic' to have the current method of resolving do it). This would
allow somebody to determine whether or not the automatic resolving is
causing a problem, which could make them take a look at the contents of
the file to investigate the cause.
So in short:
1) adding the record 'Entering Link 1' would help
2) there will still be cases where the output file may not be resolved
properly
Cheers,
René
On Jul 9, 2005, at 6:43 AM, Dr P. Murray-Rust wrote:
On Jul 9 2005, [EMAIL PROTECTED] wrote:
Citeren René Kanters <[EMAIL PROTECTED]>:
> I took a quick look at the files you attached. In neither case is
the > resolver able to determine the type of file, i.e., it isn't
able to > figure out that it is a Gaussian file. The resolver
expects to see the > string "Entering Gaussian System" early on. In
neither of your files > that was there, which is peculiar because I
understand that Gaussian > really should do that.
I spent a considerable time writing parsers for Gaussian output.
Unfortunately it varies very considerably between versions - more than
other programs. So you cannot rely on messages like this occurring in
every case.
> > So to make these files work I did:
> In the acetone.g03.out:
> replace the first line:
> Entering Link 1 = c:\G03W\l1.exe PID= 2100.
> with
> Entering Gaussian System
Are the strings "Entering Link" common in g03 files? If so, we might
add them to the resolver...
Again this is system-dependent. Gaussian consists of components called
links and some, but not all, implementations emit these messages.
Without these messages it is almost impossible to write a full Gausian
parser - you cannot tell where one link stops and the next starts. It
is possible to cherry-pick some fields but nothing can be relied upon.
Other systems like GAMESS are starting to use CML/XML as the markup
and this will revolutionise this area.
P.
Egon
------------------------------------------------------- This SF.Net
email is sponsored by the 'Do More With Dual!' webinar happening July
14 at 8am PDT/11am EDT. We invite you to explore the latest in dual
core and dual graphics technology at this free one hour event hosted
by HP, AMD, and NVIDIA. To register visit
http://www.hp.com/go/dualwebinar
_______________________________________________ Jmol-developers
mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-developers
--
Peter Murray-Rust
Unilever Centre for Molecular Informatics
Chemistry Department, Cambridge University
Lensfield Road, CAMBRIDGE, CB2 1EW, UK
Tel: +44-1223-763069 Fax: +44 1223 763076
-------------------------------------------------------
This SF.Net email is sponsored by the 'Do More With Dual!' webinar
happening
July 14 at 8am PDT/11am EDT. We invite you to explore the latest in
dual
core and dual graphics technology at this free one hour event hosted
by HP,
AMD, and NVIDIA. To register visit http://www.hp.com/go/dualwebinar
_______________________________________________
Jmol-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-developers
-------------------------------------------------------
This SF.Net email is sponsored by the 'Do More With Dual!' webinar happening
July 14 at 8am PDT/11am EDT. We invite you to explore the latest in dual
core and dual graphics technology at this free one hour event hosted by HP,
AMD, and NVIDIA. To register visit http://www.hp.com/go/dualwebinar
_______________________________________________
Jmol-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-developers
