Can osm2pgsql-replication init be used with an extract from osmium-tags-filter?

[wcedmisten]

Hello! I want to keep an updated database of all restaurants on the planet. My current approach is to use osmium-tags-filter to filter the planet file in order to get a filtered PBF containing just the restaurants. This is significantly faster than filtering the data in osm2pgsql.

I'm wondering if using this extract with osm2pgsql-replication and the regular planet replication URL will result in unexpected behavior? Does osm2pgsql-replication need to initialize with all existing nodes on the planet in order to process updates correctly?

I'm imagining a process like this, but I'd like to know if it will result in the correct data:

• Download the standard planet PBF
• Run osmium tags filter to get just the restaurants
• Use that filtered PBF with the flex output for osm2pgsql that also filters for just restaurants
• Run osm2pgsl-replication init
• ...time passes
• Whenever an update is needed, run osm2pgsql-replication update

Thanks!

[joto]

No. This does not work. For instance a new way might come in with a certain tag that you are interested in but the nodes in that way are not in the update and not in your original data.