[gdal-dev] Questions regarding phrasing in the documentation of `OCTTransformBounds`

Florian Katerndahl florian at katerndahl.com
Tue Apr 1 14:12:50 PDT 2025


Dear GDAL-devs,

while working on a small project, I (more or less inevitably) came 
across the problem of what to do when a bounding box geometry, that is 
to be translated to EPSG:4326, crosses the antimeridian in the target 
SRS (i.e. EPSG:4326). The docs for the C function `OCTTransformBounds` 
note that the user has to check if a) the destination/target SRS is 
geographic (which it is in my case), b) if the first SRS axis is 
latitude or longitude and c) depending on that check whether the maximum 
longitudinal bounding coordinate is smaller then than the minimum 
bounding longitudinal coordinate, accessing the correct axis.

Unfortunately, the above mentioned check whether the bounding box 
"crossed the antimeridian" confuses me. Applying this check on the 
output correctly detects crossing of the antimeridian in EPSG:4326.

While I may very well misunderstand the documentation and/or the 
underlying processing, I would be thankful for an answer to the 
following two questions:

 1. Is the documentation correct that input data must be checked and not
    the output data, i.e. `out_xmax` etc.?
 2. If so, why should this check be done on the input coordinates even
    though they may not be in a geographic coordinate reference system
    in the first place or have a SRS assigned that handles 180°
    crossings (e.g. Mollweide centered at 180°)?

Kind regards and thanks in advance,

Florian Katerndahl
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20250401/fc0ea324/attachment.htm>


More information about the gdal-dev mailing list