[PROJ] Determine "cut line" for arbitrary projection

[William Temperley]

Yes, that was helpful, thanks. It has been reasonably easy to use
 "proj_get_area_of_use" in the C++ API
This is probably all that's needed for projected coordinate reference
systems like EPSG:27700 where parameters are fixed (AFAIK).

In the end, I think the rest is probably a user or application-space
problem.
Mercator can probably be dealt with by just throwing away any
coordinates outside of what "proj_get_area_of_use" returns.

Seems the zones are hard-coded in the interrupted projections.
For cartographic purposes it'll probably be sufficient to build a clip mask
from the projected nodes defined the source.
e.g. imoll.cpp:

/*

  Zones:


    -180            -40                       180

      +--------------+-------------------------+

      |1             |2                        |

      |              |                         |

      |              |                         |

      |              |                         |

      |              |                         |

    0 +-------+------+-+-----------+-----------+

      |3      |4       |5          |6          |

      |       |        |           |           |

      |       |        |           |           |

      |       |        |           |           |

      |       |        |           |           |

      +-------+--------+-----------+-----------+

    -180    -100      -20         80          180

*/


On Mon, 10 Mar 2025 at 20:47, Greg Troxel via PROJ <proj at lists.osgeo.org>
wrote:

> William Temperley via PROJ <proj at lists.osgeo.org> writes:
>
> > Couldn't the cutlines be reasonably easily generated in geographic
> > coordinates from either user-supplied or default projection
> > parameters, using one of the above methods?
> > An example for a derivable extent would be an orthographic projection
> where
> > the cutline is just 90 degrees in spherical distance from the centre
> point.
>
> A further question might be how the cutlines relate to the 'area of use'
> in the EPSG database, but I'm afraid the answer is that the area of use
> is a geodetic bounding box.  It does seem that a projection database
> ideally would have this information.
>
> But, as you point out with Mercator, that leads to Mercator-80 and
> Mercator-85 entries, and that doesn't seem good.
>
> Not sure this ended up being helpful, but wanted to mention the possible
> relationship.
>
> _______________________________________________
> PROJ mailing list
> PROJ at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/proj
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20250311/9c3910c4/attachment.htm>