In setMap, the layer gets the map's projection if it doesn't have its own. And since a layer can have a different SRS code than the map (but a compatible one, i.e. with OpenLayers.Projection.transforms[mapProj][layerProj] being OpenLayer.Projection.nullTransform), the axis order can be different.
This allows us to simplify the map and layer configuration, because now the projection also defines defaults for maxExtent, maxResolution and units.
This change also adds transforms for SRS aliases for EPSG:4326 and centralizes axis order information in OpenLayers.Projection.defaults.
Thanks dwins for spotting this and the quick patch, which, btw, was
perfect (i.e. with unit tests that show the problem). r=me (closes#2025)
git-svn-id: http://svn.openlayers.org/trunk/openlayers@9161 dc9f47b5-9b13-0410-9fdd-eb0c1a62fdaf
following modifications:
* WMSGetFeatureInfo class was contained twice in WMSGetFeatureInfo.js.
Removed one.
* Fixed @requires
* Added vendorParams option (with test) to allow for e.g. adding custom
params like RADIUS
* Changed the clickPosition key in the getfeatureinfo event parameter
to xy to comply with other event parameters in OpenLayers
* Modified concatenation of layerNames and styleNames as proposed in my
previous comment
* Made a separate handleResponse function from the previously closured
callback function and added xy to the scope
* Fixed some ND comments, especially removed links (<..>) from object
types that cannot be linked (e.g. {String})
* Inserted line breaks where lines exceeded 80 chars
* Fixed test for format option, because this is now formats and has a
different type
* Fixed tests in the example (this is no US census data)
* added ProxyHost and absolute WMS url to the example
* removed custom format from the click handler in the example so users
can also see the simpler instantiation with the default formats
r=me (closes#2007)
git-svn-id: http://svn.openlayers.org/trunk/openlayers@9159 dc9f47b5-9b13-0410-9fdd-eb0c1a62fdaf
