On Feb 23, 2009, at 4:42 PM, Neil Woodhouse wrote:
Howard,
Some additional points. I am using the C API, so I don't think
that the repair utility is available to me.
lasinfo's repair operations work by opening the LASWriterH in append
mode. lasinfo.c uses the C API. See the code in lasinfo.c for more
information
Here some pseudo code to
indicate my quandary.
Create the header, point and writer
For all blocks of data
Process the data block
Add the point to the LAS file
Accumulate the bounds of the points
Write the header
Close the file
Can you try the following?:
Create the header, point, and writer
For all blocks of data
process
add the point to the LASWriterH with _WritePoint
accumulate the bounds of the points
* Close the file with LASWriter_Destroy when you're done writing your
points.
* Open the file with LASReader_Create and fetch a copy of the
LASHeaderH then LASReader_Destroy
* Set the extents and other items you wish to tweak on this copy.
Don't mess with the point count, VLRs, or any other things that might
cause things to blow up
* Open the file with LASWriter_Create in Append mode using your
header copy
* Destroy the LASWriterH again with LASWriter_Destroy
We are not using any VLR records, but do use the set user data
function (LASPoint_SetUserData) for the point. I understand that
this is
an unsigned char.
UserData for 1.1 is a single unsigned char, right? I will update my
test case to try this.
BTW do you know of any free viewers that can view
this?
View the user data? I would assume that by definition, only the
application that created it would know how to interpret it.
The reason for this workflow is that we do not know the number of
points upfront and don't know the extents of them.
The LASWriter does not know the number of points it will write and
will do so with UpdateHeader for you as you destroy the writer with
LASWriter_Destroy or ~LASWriter when it goes out of scope. As for the
extents, I think the only way to currently update them now is using
the pseudocode I described (it should be the same thing as what's
currently in lasinfo)
Hope this helps,
Howard
_______________________________________________
Liblas-devel mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/liblas-devel
