Adrian Custer a écrit :
> 1) Why does the ProjectedCRS not have any scope or valid area set when a
> Mercator_1SP is made from WKT? It seems like the library could impose a
> maximum extent for any data attempting to be Mercator. In the debugger,
> none of these appear to be set.
The scope and valid area are fetch from the EPSG database when using any
gt2-epsg-* factory other
than epsg-wkt. There is no way to specify this information in WKT format. We
could hard-code a
maximum extent in every operation method, but it still a maximum extent, not a
valid area (which is
usually much smaller than the maximum extent, and can not be guessed - we need
to look in the database).
Latest ISO 19111 revision (not yet reflected into GeoAPI interfaces) added a
field in
CoordinateSystemAxis, which allow to specify the axis extent. So you get the
same fonctionnality
than a "maximum extent", but latest ISO 19111 put this particular information
at the axis level. It
does make sense, since the extent of the longitude axis (for example) is
independent of the
particular CRS in use, while "valid area" and "scope" are CRS-specific. This
"coordinate system axis
range" feature is not get implemented in Geotools, but this is planned.
> 2) Is Meractor_1SP the name of the PROJCS or of the PROJECTION element
> in the WKT?
The PROJECTION element. The name of "PROJCS" is arbitrary and can be anything.
> 3) To make a Mercator programatically, I've found
> FactoryGroup.createProjectedCRS(...)
> and
> DefaultProjectedCRS(...)
> but are there any other ways to create a Mercator ProjectedCRS? What's
> the easiest way to create a Mercator programatically?
The preferred way is to use factories. Unfortunatly, the current way to create
a ProjectedCRS from a
factory is broken (over-complicated). This is the subject of:
http://jira.codehaus.org/browse/GEO-61
Which I hope to fix in the next GeoAPI version. The suggested way (you should
be able to do that
using directly Geotools constructors, not yet with factory, but fixing factory
that way is the goal
of GEO-61) is to create first a DefiningConversion object, and then give this
DefiningConversion
object to the ProjectedCRS constructor.
A workaround is defined there:
http://javadoc.geotools.fr/snapshot/org/geotools/referencing/factory/FactoryGroup.html#createProjectedCRS(java.util.Map,%20org.opengis.referencing.crs.GeographicCRS,%20org.opengis.referencing.operation.Conversion,%20org.opengis.referencing.cs.CartesianCS)
But this method should actually be part of GeoAPI CRSFactory, in replacement of
current one. GEO-61
is exactly about that. I suggest that you use the above-cited method and make
clear that this is a
temporary workaround - a method with the same name and signature should be
added in CRSFactory in
GeoAPI 2.1.
Note that GeoAPI factory method other than ProjectedCRS and DerivedCRS should
be okay.
> 4) The above methods require various combination of things, one of which
> is:
> String => Is this name arbitrary (whatever I want)?
Yes, except for MathTransform which must be one of the registered name.
> OperationMethod => What's this?
Metadata about the MathTransform implementation (not the parameter, just the
method). For example
the formulas for Mercator projection. You can avoid creating an OperationMethod
if you first create
a DefiningConversion using the first convenience constructor, then the
above-cited "workaround"
create method. The important thing is to get the right ParameterValueGroup:
pg = mathTransformFactory.getDefaultParameters("Mercator_1SP");
pg.parameter("falseEasting").setValue(...);
// etc.
Martin.
Martin.
-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Geotools-gt2-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-gt2-users
