[Octopus-users] abnormality in the work of the parser

[Xavier Andrade]

Hi Igor,

Thanks for your report, this is certainly a bug, I will have a look at it, 
but probably the solution will be to forbid having a variable defined 
twice.

Cheers,

Xavier


On Sat, 28 Jun 2008, ??????? ?.?. wrote:

>     Dear developers,
>     There seem to exist an abnormality in the work of the parser (in Octopus 3.0.1).
> Namely, let the input file have the form
> -----\
> Units = eV_Angstrom
>
> rHH=0.7
> r_H=4.0
>
> %Species
> 'H' | 1.0 | spec_ps_psf | 1 | 1 | 1
> %
> %Coordinates
> 'H' | -rHH/2 | 0.0 | 0.0 | no
> 'H' |  rHH/2 | 0.0 | 0.0 | no
> %
>
> BoxShape = cylinder
> Radius = r_H
> Xlength = rHH/2+r_H
> Spacing = 0.1
>
> CalculationMode = gs
> -----/
> , and we run it as follows:
> -----\
> export OCT_PARSE_ENV=1
> export OCT_rHH=1.6
> ../src/main/octopus 2>>gs.err 1|tee gs.lst
> -----/
> ; then in exec/out.oct we obtain
> -----\
> Opened block 'Coordinates'
>  (0, 1) = -0.8
>  (1, 1) = 0.8
> xlength = 4.35
> -----/
> - xlength = 4.0+0.7/2 = 4.35 instead of 4.0+1.6/2 = 4.8;
> i.e., when calculating coordinates, it uses the exported value
> (as one could expect), but when calculating Xlength, it uses the old value
> (which is not what the user intented).
>     If we comment out the assignment to rHH in the input file,
> or hold the original input file under different name and modify the input file
> in a cylcle over rHH, e.g.,
> cat h2.inp |sed 's_rHH=0.7_rHH='$rHH'_' >inp
> , we obtain the desirable values: coordinates - +-0.8, and xlength = 4.8 .
> -----
> .
>     With best regards     Igor.
> _______________________________________________
> Octopus-users mailing list
> Octopus-users at tddft.org
> http://www.tddft.org/mailman/listinfo/octopus-users
>