OSM phone number normalisation

derive_changes.sh

#!/bin/bash


osmosis \
--fast-read-xml file="out.xml" \
--buffer \
--sort \
--read-pbf-fast file="lyon_france.osm.pbf" \
--sort \
--buffer \
--derive-change \
--log-progress-change \
--write-xml-change file="change.osc"

transform.sh

#!/bin/bash

osmosis \
--read-pbf-fast file="lyon_france.osm.pbf" \
--buffer \
--tag-transform file="transform.xml" \
--log-progress \
--buffer \
--write-xml file="out.xml"

transform.xml

<?xml version="1.0"?>
<translations>
  <translation>
    <name>French phone numbers normalisation</name>
    <description>French phone numbers normalisation</description>
    <match>
      <tag k="phone|contact:phone" match_id="phone1" v=".*(\d).*?(\d).*?(\d).*?(\d).*?(\d).*?(\d).*?(\d).*?(\d).*?(\d)"/>
    </match>
    <output>
      <copy-all/>
      <tag from_match="phone1" v="+33 {1} {2}{3} {4}{5} {6}{7} {8}{9}"/>
    </output>
  </translation>
</translations>

view_out

# Allow to quickly see defects post-transformation, before deriving the osc file

sed -ne 's/.*k=\"\(contact:\)*phone\" v=\"\(.*\)\".*/\2/p' out.xml | sort | uniq | less