Hello Isuranga You are right, the prototype implementation uses a "filter" keyword which is not yet conform to a standard. Writing a CQL parser still need to be done. For this task we can either take the Geotk parser as a starting point [1][2] (which uses code generated by ANTLR), or we can write a parser manually with only a few functionalities for starting. Indeed, it needs to be in the project scope.
The mapping from Landsat, Modis, GeoTIFF and netCDF metadata to ISO 19115 may not be 100% complete, but I think that their coverage are sufficient for this project. We may complete them if we find obvious holes, but that would be mostly opportunistic improvements. SOAP endpoints would be useful too. Thanks, Martin [1] https://github.com/Geomatys/geotoolkit/tree/master/modules/feature/geotk-feature/src/main/java/org/geotoolkit/cql [2] https://github.com/Geomatys/geotoolkit/blob/master/modules/feature/geotk-feature/src/main/antlr4/org/geotoolkit/cql/CQL.g4 Le 24/02/2018 à 16:23, Isuranga Perera a écrit : > Hi Martin, > > I went through OGC CSW specification and Hao's and Chin's works as > well. Basically Hao has implemented get endpoints. So does the > project scope includes POST and SOAP endpoints as well? > If I'm not mistaken in the prototype implementation filters(e.g. east, > west, north and south bounds) are not yet conform to a standard query > language like CQL so that this task should be included in the project > scope. In addition to that can I know if mapping from Landsat, GeoTIFF > and Modis metadata to ISO 19115 implementaion is complete ( is it > support all informations in metadata. do I have to complete that as > well?). > > Appreciate if you can clarify those problems. > > Best Regards > Isuranga Perera
