let me have a look at it over the weekend. i am fairly certain that i can cut out some computing which is not necessary for sextante raster, which might already fix this issue.
..ede On 9/17/2020 15:34, Giuseppe Aruta wrote: > In any case I need compare the same errors with Roberto: > we are working around this "bug" since a couple of days and right now we > cannot perfectly duplicate the same errors: > I have errors only on one file, anyhowI can load and view all the file, > included the fault one) > Robberto verified the error in several files of the list and was not able > to load and view some of them > Peppe > > Il giorno gio 17 set 2020 alle ore 15:29 Giuseppe Aruta < > giuseppe.ar...@gmail.com> ha scritto: > >> Hi Ede, >> sorry, I sent the message just a microsecond I received yours >> >>> how about we modify the sources eg. TIFFUtil and others instead to fit >> your need instead? >> Yes >> >>> i don't understand. you mean retrieving georeferencing values, image >> dimensions, both? >> georeferencing values. In the way sextante load files there is a first >> step (in AddRasterImageWizard class to understand those values without >> loading the image (for TIFF files) - than all the process is activated >> through the other classes (excuse me for my very poor way to explain) >> >> >> Il giorno gio 17 set 2020 alle ore 15:22 <edgar.sol...@web.de> ha scritto: >> >>> hey Peppe, >>> >>> On 9/17/2020 14:58, Giuseppe Aruta wrote: >>>> Hi Ede, >>>> the class >>>> com.vividsolutions.jump.workbench.imagery.geoimg.GeoReferencedRaster >>> (and >>>> related) work very well on loading raster in sextante. >>>> In order to clean up from any message error belonging to classes non >>> used >>>> by Sextante (ReferencedImage at similar and the method .paint. I made a >>>> copy of it (and GeoRaster class), cleaned from all those unused method, >>>> renamed (as SextanteReferencedRaster etc) and use it instead of the >>>> original to load Raster as Sextante. >>> >>> while i understand your approach here, which is good for learning and >>> understanding things i really do not like duplicated code in production. >>> how about we modify the sources eg. TIFFUtil and others instead to fit your >>> need instead? >>> >>>> My idea was also to modify the class in order to have a method that >>>> calculate the location of a raster without computing the image/raster. >>> >>> i don't understand. you mean retrieving georeferencing values, image >>> dimensions, both? >>> >>>> I was aware to have back the same error as Roberto. But the error was >>>> limited to only one file (among the one Roberto sent: Adrena070.tif) >>> which >>>> shows some values of the cells as doubles with many decimals, possibly >>>> derived from irrational number? >>> >>> give me some time to check that out. i am currently not sure what's going >>> on, but suspect some JAI operations at fault here. >>> >>>> Now I see your new mail about commons imaging: could it be that they are >>>> used somehow by the class GeoRaster? >>> >>> somehow, yes. but with no regard to the use via TIFFUtil class if the >>> forced loader is used. >>> >>>> And this explains the similar warning? >>> >>> should be completely unrelated. ArrayIndexOutOfBoundsException are the >>> NPEs of array handling. pretty common if not actually every possibility was >>> thought of before implementing the code. >>> >>> ..ede >>> >>> >>> >>> >>> _______________________________________________ >>> Jump-pilot-devel mailing list >>> Jump-pilot-devel@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel >>> >> > > > > _______________________________________________ > Jump-pilot-devel mailing list > Jump-pilot-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel > _______________________________________________ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
