View source for Preparing FKB data to MBTiles

This guide shows how to process FKB and related data into vector MBTiles datasets using FME 2021 or newer. The result are split into counties. Each county will consist of 3 MBTiles datasets: 
* FKB
* Road names (elveg)
* adresses (vegadresse)
These datasets will be used to update the product series '''NorwayTopoDetail'''.

== FKB ==
Source data must be downloaded from Geonorge with a valid '''BAAT''' user. The large amount of data makes it neccessary to use the tool '''MassivNedlasting''' from Kartverket. The following map themes should be downloaded into a single folder. Each theme should consist of one zipped ESRI file geodatabase for each county except for '''N5 Presentasjonsdata''', which as of 2022 is only available split into municipalities.
* FKB-AR5 FGDB-format
* FKB-Arealbruk FGDB-format
* FKB-Bane FGDB-format
* FKB-BygnAnlegg FGDB-format
* FKB-Bygning FGDB-format
* FKB-Ledning FGDB-format
* FKB-LedningVA FGDB-format
* FKB-Lufthavn FGDB-format
* FKB-Naturinfo FGDB-format
* FKB-Servitutt FGDB-format
* FKB-Tiltak FGDB-format
* FKB-TraktorvegSti FGDB-format
* FKB-Vann FGDB-format
* FKB-Veg FGDB-format
* Matrikkelen - Eiendomskart Teig FGDB-format
* N5 Høydekurve FGDB-format
* N5 Presentasjonsdata FGDB-format

When all files have been downloaded, use the FME workspace '''workspacerunner.fmw'''. Fill in the user parameters <code>input</code> and <code>output</code>. It is also possible to select which counties to process using the parameters <code>Fylker</code>. Runtime will vary, but all of Norway will be at least 12 hours. 

When all counties are finished, the Maria GDK products NorwayTopoDetail can be updated using the new MBTiles datasets. All products should first be set to "Draft" in Maria Map Maker. Then quit the application, and replace the existing MBTiles file in the "FKB" folder in the product. Restart Maria Map Maker, and inspect the overlays.

=== Troubleshooting ===
The workspace described here depends on the naming convention of file names, layer names and attributes in the source data. Unfortunately, these names are often changed by Kartverket from one year to the next, which usually results in errors or missing layers in the resulting datasets. Troubleshooting this is the most time-consuming part of the process. 
* When a layer is visibly missing in the result, a good starting point is to closely compare the source data with the result. The file '''FeatureGrouping.xml''' in the products contains the attribute filters for each style layer.
* Check the FME log files, which are written next to the workspace (fmw) file.
* Use '''extractmapinfo.exe''' to create an xml report on all the layers and attributes in the MBTiles file.

== Elveg ==

'''Elveg''' is a dataset representing the road network in Norway. In '''NorwayTopoDetail''' we use the road center lines to label road names and numbers. The Vbase line features themselves are not visible in the map.

We create Elveg as a separate MBTiles dataset from the base data. However, it is split by fylke in the same way.

Our starting point is one zipped Elveg GML file for each fylke. We use the FME workspace '''gml2mbtiles_Elveg2.0.fmw'''.

After running the workspace, you should have one MBTiles file for each fylke, which replaces the MBTiles files in the existing '''NorwayTopoDetail''' products.

== Matrikkelen - Adresse (Adress labels) ==

Points representing street adresses. It is placed in each '''NorwayTopoDetail''' as a separate dataset in the same way as '''Vbase'''

The starting point should be one zipped GML file for each fylke. Use the FME workspace '''gml2mbtiles_matrikkel-vegadresse.fmw'''. After running the workspace, you should have one MBTiles file for each fylke, which replaces the MBTiles files in the existing '''NorwayTopoDetail''' products.

[[Category:Preparing data]]