[PROJ] about projinfo and --spatial-test

[Even Rouault]

Ah, I read your message too fast. I see I added that in 
https://github.com/OSGeo/PROJ/commit/f50221ed6606a67ef0354cacd449f02aca3faba7. 
Unfortunately there is no corresponding test and my memories have 
obviously vanished. This is perhaps a situation where --spatial-test 
intersects would return a non-ballpark operation and without it a 
ballpark operation. Although I suspect this might be more subtle than that.

Le 20/01/2023 à 20:51, Javier Jimenez Shaw a écrit :
> Thanks Even
>
> Unfortunately I do not have any example. I was just reading the code 
> of projinfo to see how to get CoordinateOperations in C++.
>
> .___ ._ ..._ .. . ._. .___ .. __ . _. . __..  ... .... ._ .__
> Entre dos pensamientos racionales
> hay infinitos pensamientos irracionales.
>
>
>
> On Fri, 20 Jan 2023 at 20:15, Even Rouault 
> <even.rouault at spatialys.com> wrote:
>
>     Hi Javier,
>
>     which exact projinfo invokation did you use ?
>
>     (I'm not that surprised and remember having already seeing that.
>     There are various heuristics and filtering that could cause this)
>
>     Even
>
>     Le 20/01/2023 à 20:11, Javier Jimenez Shaw a écrit :
>>     Hi
>>
>>     Inspecting the code of projinfo, I see a situation that may show
>>     the message
>>     "Note: using '--spatial-test intersects' would bring more
>>     relevant results".
>>     Looking at the code, it happens only when using "--spatial-test"
>>     option with "intersects" or default ("contains") both return 1
>>     operation... but different.
>>     How is that possible? I would expect that "intersects" contains
>>     all in "contains".
>>
>>     The code involved is this
>>
>>             list =
>>     CoordinateOperationFactory::create()->createOperations(
>>                 nnSourceCRS, nnTargetCRS, ctxt);
>>             if (!spatialCriterionExplicitlySpecified &&
>>                 spatialCriterion ==
>>     CoordinateOperationContext::SpatialCriterion::
>>     STRICT_CONTAINMENT) {
>>                 try {
>>                     ctxt->setSpatialCriterion(
>>     CoordinateOperationContext::SpatialCriterion::
>>                             PARTIAL_INTERSECTION);
>>                     auto list2 =
>>     CoordinateOperationFactory::create()->createOperations(
>>                             nnSourceCRS, nnTargetCRS, ctxt);
>>     spatialCriterionPartialIntersectionResultCount = list2.size();
>>                     if
>>     (spatialCriterionPartialIntersectionResultCount == 1 &&
>>                         list.size() == 1 &&
>>                         list2[0]->nameStr() != list[0]->nameStr()) {
>>     spatialCriterionPartialIntersectionMoreRelevant = true;
>>                     }
>>                 } catch (const std::exception &) {
>>                 }
>>             }
>>     ....
>>             if (spatialCriterionPartialIntersectionMoreRelevant) {
>>                std::cout << "Note: using '--spatial-test intersects'
>>     would bring "
>>                             "more relevant results."
>>                          << std::endl;
>>            }
>>
>>     Thanks
>>     Javier
>>     .___ ._ ..._ .. . ._.  .___ .. __ . _. . __..  ... .... ._ .__
>>     Entre dos pensamientos racionales
>>     hay infinitos pensamientos irracionales.
>>
>>
>>     _______________________________________________
>>     PROJ mailing list
>>     PROJ at lists.osgeo.org
>>     https://lists.osgeo.org/mailman/listinfo/proj
>
>     -- 
>     http://www.spatialys.com
>     My software is free, but my time generally not.
>
>     _______________________________________________
>     PROJ mailing list
>     PROJ at lists.osgeo.org
>     https://lists.osgeo.org/mailman/listinfo/proj
>
-- 
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/proj/attachments/20230120/6e914bfe/attachment.htm>