Windkit: Possible bug when infering number of sectors from RSF file

[Florian Mintgen]

Hi WAsP team,

First of all thank you for providing these tools.

I think the windkit conversion from RSF files contains a bug: When it tries to infer the number of sectors in the source file (in _infer_resource_file_nsec), the lines are split and counted:

    return (len(line.split()) - 9) // 3

However, this doesn't seem to account for occasions where some of the columns in the RSF file adhere to the column width specs, but don't have a space in between, like here:

Point   726452.0 5781855.8   130.5149.0 9.54 2.244        633.900 12  78 116  235  24  77  207  35  72  246  99  83  278  92  80  260  36  77  244  51  77  232  93  82  238 124  87  240 140 101  247 110 115  263 118 122  272
Point   726452.0 5781880.8   130.8149.0 9.54 2.244        633.700 12  78 116  235  24  77  207  35  72  246  99  83  278  92  80  260  36  77  244  51  77  232  93  82  238 124  87  240 140 101  247 110 115  263 118 122  272

It can also be problematic when the text string in the first ten columns is an empty string.

Is there a reason not to read the number of sectors from columns 70-72?

Cheers,
Florian
 

[Neil Davis]

Thanks for the bug report. I have forwarded it to our team, and hopefully we will have a fix out before too long.

[Florian Mintgen]

Thanks Neil. Another thing I came across is that when I change the second line in the RSF such that _infer_resource_file_nsec results in 12 sectors, the site object ends up having 13 sectors now, containing both 0.0 and 360.0 deg. Is that the way it's supposed to be, and handled later down the line, or could that be another bug?

 

xarray.Dataset

• Dimensions: (h: 1, wd: 13, ws: 23, x: 1071, y: 886)
• Coordinates:
  • Name (x, y, h) object ...
  • wd (wd) float64 0.0 30.0 60.0 ... 300.0 330.0 360.0

   

   

Edited July 27, 2022 by Florian Mintgen
Small typo