#!/usr/bin/ruby
-require 'rubygems'
-require 'pg'
-require 'yaml'
-require 'time'
-require 'fileutils'
-require 'xml/libxml'
-require 'zlib'
+require "rubygems"
+require "pg"
+require "yaml"
+require "time"
+require "fileutils"
+require "xml/libxml"
+require "zlib"
# after this many changes, a changeset will be closed
CHANGES_LIMIT = 50000
attr_reader :id, :created_at, :closed_at, :num_changes
def initialize(row)
- @id = row['id'].to_i
- @created_at = Time.parse(row['created_at'])
- @closed_at = Time.parse(row['closed_at'])
- @num_changes = row['num_changes'].to_i
+ @id = row["id"].to_i
+ @created_at = Time.parse(row["created_at"])
+ @closed_at = Time.parse(row["closed_at"])
+ @num_changes = row["num_changes"].to_i
end
def closed?(t)
class Replicator
def initialize(config)
@config = YAML.load(File.read(config))
- @state = YAML.load(File.read(@config['state_file']))
- @conn = PGconn.connect(@config['db'])
+ @state = YAML.load(File.read(@config["state_file"]))
+ @conn = PGconn.connect(@config["db"])
@now = Time.now.getutc
end
def open_changesets
- last_run = @state['last_run']
+ last_run = @state["last_run"]
last_run = (@now - 60) if last_run.nil?
- @state['last_run'] = @now
+ @state["last_run"] = @now
# pretty much all operations on a changeset will modify its closed_at
# time (see rails_port's changeset model). so it is probably enough
# for us to look at anything that was closed recently, and filter from
def changeset_dump(changesets)
doc = XML::Document.new
doc.root = XML::Node.new("osm")
- { 'version' => '0.6',
- 'generator' => 'replicate_changesets.rb',
- 'copyright' => "OpenStreetMap and contributors",
- 'attribution' => "http://www.openstreetmap.org/copyright",
- 'license' => "http://opendatacommons.org/licenses/odbl/1-0/" }
+ { "version" => "0.6",
+ "generator" => "replicate_changesets.rb",
+ "copyright" => "OpenStreetMap and contributors",
+ "attribution" => "http://www.openstreetmap.org/copyright",
+ "license" => "http://opendatacommons.org/licenses/odbl/1-0/" }
.each { |k, v| doc.root[k] = v }
changesets.each do |cs|
xml = XML::Node.new("changeset")
- xml['id'] = cs.id.to_s
- xml['created_at'] = cs.created_at.getutc.xmlschema
- xml['closed_at'] = cs.closed_at.getutc.xmlschema if cs.closed?(@now)
- xml['open'] = cs.open?(@now).to_s
- xml['num_changes'] = cs.num_changes.to_s
+ xml["id"] = cs.id.to_s
+ xml["created_at"] = cs.created_at.getutc.xmlschema
+ xml["closed_at"] = cs.closed_at.getutc.xmlschema if cs.closed?(@now)
+ xml["open"] = cs.open?(@now).to_s
+ xml["num_changes"] = cs.num_changes.to_s
res = @conn.exec("select u.id, u.display_name, c.min_lat, c.max_lat, c.min_lon, c.max_lon from users u join changesets c on u.id=c.user_id where c.id=#{cs.id}")
- xml['user'] = res[0]['display_name']
- xml['uid'] = res[0]['id']
-
- unless res[0]['min_lat'].nil? ||
- res[0]['max_lat'].nil? ||
- res[0]['min_lon'].nil? ||
- res[0]['max_lon'].nil?
- xml['min_lat'] = (res[0]['min_lat'].to_f / GEO_SCALE).to_s
- xml['max_lat'] = (res[0]['max_lat'].to_f / GEO_SCALE).to_s
- xml['min_lon'] = (res[0]['min_lon'].to_f / GEO_SCALE).to_s
- xml['max_lon'] = (res[0]['max_lon'].to_f / GEO_SCALE).to_s
+ xml["user"] = res[0]["display_name"]
+ xml["uid"] = res[0]["id"]
+
+ unless res[0]["min_lat"].nil? ||
+ res[0]["max_lat"].nil? ||
+ res[0]["min_lon"].nil? ||
+ res[0]["max_lon"].nil?
+ xml["min_lat"] = (res[0]["min_lat"].to_f / GEO_SCALE).to_s
+ xml["max_lat"] = (res[0]["max_lat"].to_f / GEO_SCALE).to_s
+ xml["min_lon"] = (res[0]["min_lon"].to_f / GEO_SCALE).to_s
+ xml["max_lon"] = (res[0]["max_lon"].to_f / GEO_SCALE).to_s
end
res = @conn.exec("select k, v from changeset_tags where changeset_id=#{cs.id}")
res.each do |row|
tag = XML::Node.new("tag")
- tag['k'] = row['k']
- tag['v'] = row['v']
+ tag["k"] = row["k"]
+ tag["v"] = row["v"]
xml << tag
end
# saves new state (including the changeset dump xml)
def save!
- File.open(@config['state_file'], "r") do |fl|
+ File.open(@config["state_file"], "r") do |fl|
fl.flock(File::LOCK_EX)
- sequence = (@state.key?('sequence') ? @state['sequence'] + 1 : 0)
- data_file = @config['data_dir'] + format("/%03d/%03d/%03d.osm.gz", sequence / 1000000, (sequence / 1000) % 1000, (sequence % 1000))
- tmp_state = @config['state_file'] + ".tmp"
+ sequence = (@state.key?("sequence") ? @state["sequence"] + 1 : 0)
+ data_file = @config["data_dir"] + format("/%03d/%03d/%03d.osm.gz", sequence / 1000000, (sequence / 1000) % 1000, (sequence % 1000))
+ tmp_state = @config["state_file"] + ".tmp"
tmp_data = "/tmp/changeset_data.osm.tmp"
# try and write the files to tmp locations and then
# move them into place later, to avoid in-progress
Zlib::GzipWriter.open(tmp_data) do |fh|
fh.write(changeset_dump(open_changesets))
end
- @state['sequence'] = sequence
+ @state["sequence"] = sequence
File.open(tmp_state, "w") do |fh|
fh.write(YAML.dump(@state))
end
FileUtils.mv(tmp_data, data_file)
- FileUtils.mv(tmp_state, @config['state_file'])
+ FileUtils.mv(tmp_state, @config["state_file"])
fl.flock(File::LOCK_UN)
rescue
projects / chef.git / blobdiff