[gdal-dev] Unknown ASCII raster format

[Even Rouault]

Javier,

yes I've came across similarly looking grid files occasionally. I'm not 
quite sure which ancestor format is behind. This is a bit similar to the 
IGN France ASCII grid .mnt format documented at 
https://geodesie.ign.fr/contenu/fichiers/documentation/grilles/notices/Grilles-MNT-TXT_Formats.pdf, 
except it has extra fields in the header to document the number of 
values per node, etc. Or actually, the closest would be the .gra format 
that was used at a time for older IGNF grids (like 
https://geodesie.ign.fr/contenu/fichiers/documentation/grilles/outremer/RAR07_bl.gra), 
which is identical to your sample, except that the values of the header 
in .gra are split on 3 lines... and the order of longitudes and 
latitudes are switched... Both were handled by the IGNFHeightASCIIGrid 
driver that was retired in 
https://github.com/OSGeo/gdal/commit/87721936b700b7e6a1ce52fba33d24c995c12472 
and moved to OSGeo/gdal-extra-drivers in 
https://github.com/OSGeo/gdal-extra-drivers/commit/3db74a53b6ee6a0eb6645a0e4ecab31af7215600 
. Reason being that IGNF releases one grid every few years, so I'm fine 
spending 1 minute to build an out of tree driver when needed.  Given the 
lack of any metadata key, it is quite difficult to write a reliable 
automatic driver because of all the potential variants on the header 
field order.

Even


Le 22/07/2024 à 18:17, Javier Jimenez Shaw via gdal-dev a écrit :
> Hi
>
> I have seen some raster files with this format
>
> -4.983333 14.983333 -79.983333 -66.016667 0.03333333333333333 
> 0.03333333333333333
> -5.258 -5.301 -5.345 -5.393 -5.444 -5.499 -5.555 -5.615 -5.682 -5.752 
> -5.823 -5.886 -5.938 -5.986 -6.036 -6.093 -6.157 -6.226 -6.296 -6.358 
> -6.413 -6.462 -6.516 -6.574 -6.636 -6.700 -6.761 -6.819 -6.868 -6.901 
> -6.933 -6.962 -7.004 -7.069 -7.156 -7.263 -7.374 -7.479 -7.581 -7.669 
> -7.748 -7.814 -7.873 -7.923 -7.973 -8.027 -8.084 -8.143 -8.208 -8.278 
> -8.355 -8.447 -8.552 -8.670 -8.790 -8.911
> ...
> The first line (six numbers) is the bounding box (4) and the grid 
> steps (2), all in degrees.
> Then all the pixel values (as in AAIGrid format).
>
> If I try to run gdalinfo, I get
>
> ERROR 1: Maximum number of characters allowed reached.
> ERROR 1: Couldn't determine X spacing
> gdalinfo failed - unable to open 'file.txt'.
>
> However, if I change the first line with this command to convert it 
> into a https://gdal.org/drivers/raster/aaigrid.html , is works
>
> awk 'NR==1 {print "NCOLS "($4-$3)/$6+1 "\nNROWS "($2-$1)/$5+1 
> "\nXLLCENTER "$3 "\nYLLCENTER "$1  "\nCELLSIZE "$6} NR>1 {print $0}' 
> <source_file>
>
> Producing something like this:
>
> NCOLS 420
> NROWS 600
> XLLCENTER -79.983333
> YLLCENTER -4.983333
> CELLSIZE 0.03333333333333333
> -5.258 -5.301 -5.345 -5.393 -5.444 -5.499 -5.555 -5.615 -5.682 -5.752 
> -5.823 -5.886 -5.938 -5.986 -6.036 -6.093 -6.157 -6.226 -6.296 -6.358 
> -6.413 -6.462 -6.516 -6.574 -6.636 -6.700 -6.761 -6.819 -6.868 -6.901 
> -6.933 -6.962 -7.004 -7.069 -7.156 -7.263 -7.374 -7.479 -7.581 -7.669 
> -7.748 -7.814 -7.873 -7.923 -7.973 -8.027 -8.084 -8.143 -8.208 -8.278 
> -8.355 -8.447 -8.552 -8.670 -8.790 -8.91
> ...
>
> See that I am assuming that the coordinates in the first line are at 
> the pixel center.
>
> Do you know what is that initial format? Is it a "variation" of 
> "AAIGrid -- Arc/Info ASCII Grid"? Something else? Is it worth it to 
> add it to GDAL? (maybe the awk script is enough)
>
> Thanks
> .___ ._ ..._ .. . ._.  .___ .. __ . _. . __..  ... .... ._ .__
>
> _______________________________________________
> gdal-dev mailing list
> gdal-dev at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/gdal-dev

-- 
http://www.spatialys.com
My software is free, but my time generally not.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20240722/774f7225/attachment.htm>