/** * Splits osm data file into nodes, ways, relations files * * Relations and ways in osm doesn't have geometry and coordinates * they only have a references to nodes. * (Relations have members which are other relations, ways and nodes) * * So after we have filtered out which relations and ways we need by tags * we should load nodes underneath. * * Data stored in linear way (as a list of xml nodes). * So now it's faster to split data into three * files neither read whole file for many times. * * Actual job is performed in * me.osm.gazetter.split.Split * * Next slice phase: @see me.osm.gazetter.striper * */ package me.osm.gazetter.split;