gistfile1.rb

# Use Jruby to read hadoop sequence files

def load_libs(libs)
  Dir.glob(File.join(libs,"*.jar")).each { |f|
    require f
  }
end

load_libs ENV["HADOOP_HOME"]
load_libs File.join(ENV["HADOOP_HOME"], 'lib')

module H; end
module J; end

H::Configuration = Java::OrgApacheHadoopConf::Configuration
H::FileSystem = Java::OrgApacheHadoopFs::FileSystem
H::Path = Java::OrgApacheHadoopFs::Path
H::SequenceFile = Java::OrgApacheHadoopIo::SequenceFile
H::Writable = Java::OrgApacheHadoopIo::Writable
H::ReflectionUtils = Java::OrgApacheHadoopUtil::ReflectionUtils

J::URI = Java::JavaNet::URI

uri = '/mnt/hadoop/workspace/ddonnell/web_raw_requests.seq'
conf = H::Configuration.new
fs = H::FileSystem.get(J::URI.create(uri), conf)
path = H::Path.new(uri)

reader = H::SequenceFile::Reader.new(fs, path, conf)
key = H::ReflectionUtils.newInstance(reader.getKeyClass(), conf)
value = H::ReflectionUtils.newInstance(reader.getValueClass(), conf)

reader.next(key, value)
puts key.to_s
puts value.to_s