Download the TXT version of the ONS postcode file from http://www.ons.gov.uk/ons/guide-method/geography/products/postcode-directories/-nspp-/index.html
Run the preprocessor with the txt file as input
preprocessor input/ONSPD_FEB_2013_UK_O.txt
This is a memory intensive operation, if you face out of memory issues split the file in two (or more files), run through the preprocessor, join the two files and the run the preprocessor again, e.g.:
split -l 1300000 input/ONSPD_FEB_2013_UK_O.txt input/FEB
preprocessor.rb input/FEBaa
preprocessor.rb input/FEBab
cat input/FEBaa_processed.txt input/FEBab_processed.txt > input/combined
preprocessor.rb input/combined
The main program takes the following as input:
input file: e.g. input/combined_processed
The Northing value for the south west corner of the bounding box to run in
The Easting value for the south west corner of the bounding box to run in
The Northing value for the north east corner of the bounding box to run in
The Easting value for the north east corner of the bounding box to run in
The northing increment - used to batch up requests to overpass - leave at 1000 if unsure
The easting increment - used to batch up requests to overpass - leave at 1000 if unsure
The overpass server to retrieve from
This example runs in an area in southwest london
osm.rb "input/combined_processed" 518659 0169663 529831 0179224 1000 1000 localhost