-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/124522/#review83166
-----------------------------------------------------------


It would be neat if we could provide an .xsd file for validation of .kml files 
that use elements in the mx namespace. Google provides it here for gx: 
https://developers.google.com/kml/schema/kml22gx.xsd
We could upload that .xsd to marble.kde.org also such that tools can validate 
.kml files automatically.

With the .xsd in place you could also verify .osm => .kml conversion results 
locally, see http://stackoverflow.com/a/573443

Regarding verbosity of the osm placemark data in KML I'm fine with the current 
approach.


src/lib/marble/geodata/handlers/kml/KmlMemberTagHandler.cpp (line 55)
<https://git.reviewboard.kde.org/r/124522/#comment57430>

    As the index is read from a parser attribute I'd prefer a safety check here 
that memberIndex is within the bounds of innerBoundaries.



src/lib/marble/geodata/handlers/kml/KmlNdTagHandler.cpp (line 50)
<https://git.reviewboard.kde.org/r/124522/#comment57491>

    I don't think returning 0 is considered an error. There are a couple of 
handlers that do that.


Can you check the various documentation snippets whether the mx: namespace is 
consistently included?

- Dennis Nienhüser


On July 31, 2015, 5:37 p.m., Marius Stanciu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/124522/
> -----------------------------------------------------------
> 
> (Updated July 31, 2015, 5:37 p.m.)
> 
> 
> Review request for Marble.
> 
> 
> Repository: marble
> 
> 
> Description
> -------
> 
> - This completes the parsing/writing cycle: load ".osm" -> export as ".kml" 
> -> import ".kml" -> export ".osm" -> repeat
> 
> - I have made the custom XML schema very verbose, so as to be understandable 
> and have a KMLish XML style.
> If you think it's too verbose. I can eliminate the <member> and <nd> wrapper 
> tags and save their data
> as attributes
> 
> - One small change had to be done: ExtendedData is now written after the 
> Geometry within a <Placemark> tag ( @see change in KmlFeatureTagWriter.cpp )
> ( OsmPlacemarkData depends on the geometry, so that has to be initialized 
> before, when parsing ".kml" files )
> 
> 
> - There is one known issue: if you import polygons from a ".osm" file, the 
> polygon()->outerBoundary().nodeType() is GeoDataLineStringType, not 
> GeoDataLinearRingType as it should be
> ( this, when written in ".kml" will result such a thing:
>            <Polygon>
>                 <outerBoundaryIs>
>                     <LineString>  // this should be <LinearRing>
>                         <coordinates>-5.7129150009,-16.0033060822 
> -5.7131777661,-16.0038067061 -5.7124571297,-16.0041031487 
> -5.7119945396,-16.0034240531 -5.7124492636,-16.0031684483 
> -5.7129150009,-16.0033060822</coordinates>
>                     </LineString> )
>                     
> I know what causes the issue, and i'm currently currently trying to fix it ( 
> GeoDataLinearRing::GeoDataLinearRing( const GeoDataGeometry & other ) 
> constructor does not initialize
> it's own GeoDataLinearRingPrivate member )
> 
> 
> Diffs
> -----
> 
>   src/lib/marble/geodata/CMakeLists.txt cb89e7f 
>   src/lib/marble/geodata/handlers/kml/KmlMemberTagHandler.h PRE-CREATION 
>   src/lib/marble/geodata/handlers/kml/KmlMemberTagHandler.cpp PRE-CREATION 
>   src/lib/marble/geodata/handlers/kml/KmlNdTagHandler.h PRE-CREATION 
>   src/lib/marble/geodata/handlers/kml/KmlNdTagHandler.cpp PRE-CREATION 
>   src/lib/marble/geodata/handlers/kml/KmlOsmPlacemarkDataTagHandler.h 
> PRE-CREATION 
>   src/lib/marble/geodata/handlers/kml/KmlOsmPlacemarkDataTagHandler.cpp 
> PRE-CREATION 
>   src/lib/marble/geodata/handlers/kml/KmlTagTagHandler.h PRE-CREATION 
>   src/lib/marble/geodata/handlers/kml/KmlTagTagHandler.cpp PRE-CREATION 
>   src/lib/marble/geodata/writers/kml/KmlFeatureTagWriter.cpp f4a77ff 
>   src/lib/marble/geodata/writers/kml/KmlOsmPlacemarkDataTagWriter.h 
> PRE-CREATION 
>   src/lib/marble/geodata/writers/kml/KmlOsmPlacemarkDataTagWriter.cpp 
> PRE-CREATION 
> 
> Diff: https://git.reviewboard.kde.org/r/124522/diff/
> 
> 
> Testing
> -------
> 
> created stuff within editor then done the cycle: export as ".osm" ----> load 
> ".osm" ----> export as ".kml" ----> import ".kml" ----> export ".osm"
> compared the initial ".osm" file with the last one, they are the same. ( 
> except tag order inconsistencies cause by the QHash mechanism )
> Unit tests coming soon.
> 
> 
> Thanks,
> 
> Marius Stanciu
> 
>

_______________________________________________
Marble-devel mailing list
[email protected]
https://mail.kde.org/mailman/listinfo/marble-devel

Reply via email to