fix: NCDD parser

lib/pumi/data_source/ncdd.rb

@@ -27,6 +27,14 @@ def administrative_unit
  "ភូមិ" => AdministrativeUnit.new(en: "Village", km: "ភូមិ", latin: "Phum", code_length: 8, group: "villages")
  }.freeze
 
+ attr_accessor :existing_data
+
+ def initialize(data_files: default_data_files)
+ @existing_data = data_files.each_with_object({}) do |data_file, result|
+ result[data_file.type] = data_file.read
+ end
+ end
+
  def load_data!(source_dir: "tmp", output_dir: "data")
  source_files(source_dir).each do |file|
  parse_source_file(file)
@@ -37,21 +45,21 @@ def load_data!(source_dir: "tmp", output_dir: "data")
 
  private
 
+ def data
+ @data ||= {}
+ end
+
  def parse_source_file(file)
  CSV.read(file, headers: CSV_HEADERS).each do |csv_row|
  row = build_row(csv_row)
 
  next unless row.code
  next if row.administrative_unit.code_length != row.code.length
 
- write_location(row)
+ add_data(row)
  end
  end
 
- def data
- @data ||= {}
- end
-
  def build_row(row)
  Row.new(
  code: parse_location_code(row),
@@ -69,9 +77,10 @@ def parse_location_code(row)
  code
  end
 
- def write_location(row)
+ def add_data(row)
  data[row.administrative_unit.group] ||= {}
- data[row.administrative_unit.group][row.code] = {
+ data[row.administrative_unit.group][row.code] = existing_data.dig(row.administrative_unit.group, row.code) || {}
+ data[row.administrative_unit.group][row.code].merge!(
  "name" => {
  "km" => row.name_km,
  "latin" => row.name_latin
@@ -81,7 +90,7 @@ def write_location(row)
  "latin" => row.administrative_unit.latin,
  "en" => row.administrative_unit.en
  }
- }
+ )
  end
 
  def source_files(source_dir)
@@ -95,6 +104,14 @@ def write_data!(output_dir)
  DataFile.new(data_group).write(data.fetch(data_group), data_directory: output_dir)
  end
  end
+
+ def default_data_files
+ [
+ DataFile.new(:districts),
+ DataFile.new(:communes),
+ DataFile.new(:villages)
+ ]
+ end
  end
  end
 end