From Jason Beverage, "I ran though the sample data that Gert provided earlier today and tracked down the problem. The xoffset in readHeightField and readImage were being applied even in projected coordinate systems. This patch should fix the issue."
This commit is contained in:
@@ -517,7 +517,7 @@ void DataSet::SourceData::readImage(DestinationData& destination)
|
||||
|
||||
// note, we have to handle the possibility of goegraphic datasets wrapping over on themselves when they pass over the dateline
|
||||
// to do this we have to test geographic datasets via two passes, each with a 360 degree shift of the source cata.
|
||||
double xoffset = d_bb.xMin() < s_bb.xMin() ? -360.0 : 0.0;
|
||||
double xoffset = ((d_bb.xMin() < s_bb.xMin()) && (d_bb._isGeographic)) ? -360.0 : 0.0;
|
||||
unsigned int numXChecks = d_bb._isGeographic ? 2 : 1;
|
||||
for(unsigned int ic = 0; ic < numXChecks; ++ic, xoffset += 360.0)
|
||||
{
|
||||
@@ -875,7 +875,7 @@ void DataSet::SourceData::readHeightField(DestinationData& destination)
|
||||
|
||||
// note, we have to handle the possibility of goegraphic datasets wrapping over on themselves when they pass over the dateline
|
||||
// to do this we have to test geographic datasets via two passes, each with a 360 degree shift of the source cata.
|
||||
double xoffset = d_bb.xMin() < s_bb.xMin() ? -360.0 : 0.0;
|
||||
double xoffset = ((d_bb.xMin() < s_bb.xMin()) && (d_bb._isGeographic)) ? -360.0 : 0.0;
|
||||
unsigned int numXChecks = d_bb._isGeographic ? 2 : 1;
|
||||
for(unsigned int ic = 0; ic < numXChecks; ++ic, xoffset += 360.0)
|
||||
{
|
||||
|
||||
Reference in New Issue
Block a user