Hi Aitor,

If you are happy with a fix could you post a whole modified file as this is
the most reliable way for me to QA and merge changes.

Thanks,
Robert.

On 7 January 2015 at 15:50, Aitor Moreno <[email protected]> wrote:

> Hi again,
>
> Using *lasinfo *(included in libLAS), it is clear that the number of
> points declared in the header, and the actual number of points can differ.
>
> (excerpt of the output):
>
> ---------------------------------------------------------
>   Point Inspection Summary
> ---------------------------------------------------------
>   Header Point Count: *3837973*
>   Actual Point Count: *3813697*
>
> So, the solution should be as easy as removing the exception raise in the
> OSG loader. These lines:
>
>             //if (reader.GetHeader().GetPointRecordsCount() != i)
>             //{
>             //    return ReadResult::ERROR_IN_READING_FILE;
>             //}
>
>
> Bye,
>
>
> On Wed, Jan 7, 2015 at 4:00 PM, Aitor Moreno <[email protected]>
> wrote:
>
>> Hi Robert,
>>
>> Sorry I missed the last posts.
>>
>>
>> I tried the sample you tested and it is pretty weird. It seems that the
>> number of points declared in the header (
>> reader.GetHeader().GetPointRecordsCount() ) is wrong.
>>
>> The header returns *3837973 *as the number of points, but the reader
>> crashes after *3813697*. As the numbers do not match, the
>> *ReadResult::ERROR_IN_READING_FILE* exception is raised.
>>
>> I hacked the code to stop when the point number *3813697 *is reached,
>> and the file loads fine. Of course, I can not see what happens to the ghost
>> points (from *3813697* till *3837973*).
>>
>> I will try to check with other loaders if the file is badly formatted or
>> not, and additionally, adding some "validity control" to the OSG loader
>> *.*
>>
>>
>> Here the console output (verbose) using the following hacked code (a mere
>> break when the crash is expected)..
>>
>>
>>
>>             // now we do a second pass substracting the mid point to each
>> point
>>             reader.Reset();
>>             uint32_t i = 0;
>>             while (reader.ReadNextPoint())
>>             {
>> *                if ( i == 3813697 ) break; // hacked for
>> "LAS12_Sample_withRGB_Quick_Terrain_Modeler.las"*
>>                 liblas::Point const& p = reader.GetPoint();
>>
>>
>>
>>
>> File name: LAS12_Sample_withRGB_Quick_Terrain_Mo
>> Signature: LASF
>> Format   : 2
>> Project  : 00000000-0000-0000-0000-000000000000
>> Points count: *3837973*
>> VLRecords count: 3
>> Points by return: 3837973 0 0 0 0 0 0
>> Min Max calculation. Elapsed Time: 1669.29
>>
>> X: (728262.803178, 728998.135178)
>> Y: (4676439.019454, 4677014.685454)
>> Z: (292.647965, 327.077965)
>>
>> Read points: *3813697 *Elapsed Time: 1366.333353
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> On Mon, Dec 22, 2014 at 12:39 PM, Robert Osfield <
>> [email protected]> wrote:
>>
>>> Hi Rafa and Aitor,
>>>
>>> On 19 December 2014 at 00:04, Rafa Gaitan <[email protected]> wrote:
>>>
>>>> Hi Robert,
>>>>
>>>> Here you can find some samples for LAS files:
>>>> http://www.liblas.org/samples/
>>>>
>>>> If you have liblas built with laszip support, then .LAZ files can be
>>>> also opened.
>>>>
>>>
>>> Thanks for the link.  For .laz files we'd need to add in an alias for
>>> that extension to the plugin and Registry.cpp.
>>>
>>> I have just tried a couple of .las files, one loaded but another caused
>>> a crash:
>>>
>>> $ osgviewer LAS12_Sample_withRGB_Quick_Terrain_Modeler.las
>>> terminate called after throwing an instance of 'std::invalid_argument'
>>>   what():  invalid file signature
>>> Aborted (core dumped)
>>>
>>> I downloaded the above LAS file from the http://www.liblas.org/samples/
>>> directory.
>>>
>>> I haven't done any further investigation yet.
>>>
>>> Robert.
>>>
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> osg-submissions mailing list
>>> [email protected]
>>>
>>> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
>>>
>>>
>>
>>
>> --
>> Aitor Moreno
>>    aitormoreno [@] gmail.com
>>
>
>
>
> --
> Aitor Moreno
>    aitormoreno [@] gmail.com
>
> _______________________________________________
> osg-submissions mailing list
> [email protected]
>
> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
>
>
_______________________________________________
osg-submissions mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org

Reply via email to