hbase-jruby 0.1.1-java

data/Rakefile

@@ -0,0 +1,7 @@
+require "bundler/gem_tasks"
+require 'rake/testtask'
+Rake::TestTask.new(:test) do |test|
+  test.libs << 'lib' << 'test'
+  test.pattern = 'test/**/test_*.rb'
+  test.verbose = true
+end

data/hbase-jruby.gemspec

@@ -0,0 +1,23 @@
+# -*- encoding: utf-8 -*-
+lib = File.expand_path('../lib', __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+require 'hbase-jruby/version'
+
+Gem::Specification.new do |gem|
+  gem.name          = "hbase-jruby"
+  gem.version       = HBase::JRuby::VERSION
+  gem.authors       = ["Junegunn Choi"]
+  gem.email         = ["junegunn.c@gmail.com"]
+  gem.description   = %q{Ruby-esque interface for accessing HBase from JRuby}
+  gem.summary       = %q{Ruby-esque interface for accessing HBase from JRuby}
+  gem.homepage      = "https://github.com/junegunn/hbase-jruby"
+  gem.platform      = 'java'
+
+  gem.files         = `git ls-files`.split($/)
+  gem.executables   = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
+  gem.test_files    = gem.files.grep(%r{^(test|spec|features)/})
+  gem.require_paths = ["lib"]
+
+  gem.add_development_dependency 'test-unit'
+  gem.add_development_dependency 'simplecov'
+end

data/lib/hbase-jruby.rb

@@ -0,0 +1,16 @@
+unless RUBY_PLATFORM =~ /java/
+  raise LoadError, 'Only supports JRuby'
+end
+
+require "hbase-jruby/version"
+require "hbase-jruby/dependency"
+require "hbase-jruby/util"
+require "hbase-jruby/byte_array"
+require "hbase-jruby/column_key"
+require "hbase-jruby/cell"
+require "hbase-jruby/admin"
+require "hbase-jruby/scoped"
+require "hbase-jruby/table"
+require "hbase-jruby/result"
+require 'hbase-jruby/hbase'
+

data/lib/hbase-jruby/admin.rb

@@ -0,0 +1,29 @@
+require 'thread'
+
+class HBase
+module Admin
+private
+  def with_admin
+    (@admin_mutex ||= Mutex.new).synchronize do
+      begin
+        admin = HBaseAdmin.new(@config)
+        yield admin
+      ensure
+        admin.close if admin
+      end
+    end
+  end
+
+  def wait_async_admin admin
+    while true
+      pair  = admin.getAlterStatus(@name.to_java_bytes)
+      yet   = pair.getFirst
+      total = pair.getSecond
+
+      break if yet == 0
+      sleep 1
+    end
+  end
+end#Admin
+end#HBase
+

data/lib/hbase-jruby/byte_array.rb

@@ -0,0 +1,39 @@
+class HBase
+# @private
+class ByteArray
+  attr_reader :java
+
+  def initialize value
+    @java = Util.to_bytes value
+  end
+
+  def eql? other
+    Arrays.equals(@java, other.java)
+  end
+  alias == eql?
+
+  def <=> other
+    Bytes.compareTo(@java, other.java)
+  end
+
+  def stopkey_bytes_for_prefix
+    arr = @java.to_a
+    csr = arr.length - 1
+    arr[csr] += 1
+    while csr >= 0 && arr[csr] > 127
+      csr -= 1
+      arr[csr] += 1
+    end
+    if csr < 0
+      nil
+    else
+      arr[0..csr].to_java(Java::byte)
+    end
+  end
+
+  def hash
+    Arrays.java_send(:hashCode, [Util::JAVA_BYTE_ARRAY_CLASS], @java)
+  end
+end#ByteArray
+end#HBase
+

data/lib/hbase-jruby/cell.rb

@@ -0,0 +1,122 @@
+class HBase
+# Boxed Ruby class for org.apache.hadoop.hbase.KeyValue
+# @!attribute [r] java
+#   @return [org.apache.hadoop.hbase.KeyValue]
+class Cell
+  attr_reader :java
+
+  # Creates a boxed object for a KeyValue object
+  # @param [org.apache.hadoop.hbase.KeyValue] key_value
+  def initialize key_value
+    @java = key_value
+    @ck   = nil
+  end
+
+  # Returns the rowkey of the cell decoded as the given type
+  # @param [Symbol] type The type of the rowkey.
+  #   Can be one of :string, :symbol, :fixnum, :float, :bignum, :bigdecimal, :boolean and :raw.
+  # @return [String, byte[]]
+  def rowkey type = :string
+    Util.from_bytes type, @java.getRow
+  end
+
+  # Returns the ColumnKey object for the cell
+  # @return [ColumnKey]
+  def column_key
+    @ck ||= ColumnKey.new @java.getFamily, @java.getQualifier
+  end
+
+  # Returns the name of the column family of the cell
+  # @return [String]
+  def family
+    String.from_java_bytes @java.getFamily
+  end
+  alias cf family
+
+  # Returns the column qualifier of the cell
+  # @param [Symbol] type The type of the qualifier.
+  #   Can be one of :string, :symbol, :fixnum, :float, :bignum, :bigdecimal, :boolean and :raw.
+  # @return [Object]
+  def qualifier type = :string
+    Util.from_bytes type, @java.getQualifier
+  end
+  alias cq qualifier
+
+  # Returns the timestamp of the cell
+  # @return [Fixnum]
+  def timestamp
+    @java.getTimestamp
+  end
+  alias ts timestamp
+
+  # Returns the value of the cell as a Java byte array 
+  # @return [byte[]]
+  def value
+    @java.getValue
+  end
+  alias raw value
+
+  # Returns the column value as a String
+  # @return [String]
+  def string
+    Util.from_bytes :string, value
+  end
+  alias str string
+
+  # Returns the column value as a Symbol
+  # @return [Symbol]
+  def symbol
+    Util.from_bytes :symbol, value
+  end
+  alias sym symbol
+
+  # Returns the column value as a Fixnum
+  # @return [Fixnum]
+  def fixnum
+    Util.from_bytes :fixnum, value
+  end
+  alias integer fixnum
+  alias int     fixnum
+
+  # Returns the column value as a Bignum
+  # @return [Bignum]
+  def bignum
+    Util.from_bytes :bignum, value
+  end
+  alias biginteger bignum
+  alias bigint     bignum
+
+  # Returns the column value as a BigDecimal
+  # @return [BigDecimal]
+  def bigdecimal
+    Util.from_bytes :bigdecimal, value
+  end
+
+  # Returns the column value as a Float
+  # @return [Float]
+  def float
+    Util.from_bytes :float, value
+  end
+  alias double float
+
+  # Returns the column value as a boolean value
+  # @return [true, false]
+  def boolean
+    Util.from_bytes :boolean, value
+  end
+  alias bool boolean
+
+  # Implements column key order
+  # @param [Cell] other
+  # @return [Fixnum] -1, 0, or 1
+  def <=> other
+    KeyValue.COMPARATOR.compare(@java, other.java)
+  end
+
+  # @return [String]
+  def inspect
+    %[#{cf}:#{cq} = "#{string}"@#{ts}]
+  end
+end#Cell
+end#HBase
+

data/lib/hbase-jruby/column_key.rb

@@ -0,0 +1,63 @@
+class HBase
+class << self
+  # Shortcut method to HBase::ColumnKey.new
+  # @param [Object] cf Column family
+  # @param [Object] cq Column qualifier
+  def ColumnKey cf, cq
+    ColumnKey.new cf, cq
+  end
+end
+# Boxed class for column keys
+class ColumnKey
+  attr_reader :cf
+  alias family cf
+
+  # Creates a ColumnKey object
+  # @param [Object] cf Column family
+  # @param [Object] cq Column qualifier
+  def initialize cf, cq
+    @cf = String.from_java_bytes Util.to_bytes(cf)
+    @cq = Util.to_bytes(cq)
+  end
+
+  # @param [Symbol] type
+  def cq type = :string
+    Util.from_bytes type, @cq
+  end
+  alias qualifier cq
+
+  # @param [Object] other
+  def eql? other
+    other = other_as_ck(other)
+    @cf == other.cf && Arrays.equals(@cq, other.cq(:raw))
+  end
+  alias == eql?
+
+  # @param [Object] other
+  def <=> other
+    other = other_as_ck(other)
+    d = @cf <=> other.cf
+    d != 0 ? d : Bytes.compareTo(@cq, other.cq(:raw))
+  end
+
+  def hash
+    [@cf, Arrays.java_send(:hashCode, [Util::JAVA_BYTE_ARRAY_CLASS], @cq)].hash
+  end
+
+  def to_s
+    [@cf, @cq.empty? ? nil : cq].compact.join(':')
+  end
+
+private
+  def other_as_ck other
+    case other
+    when ColumnKey
+      other
+    else
+      cf, cq = Util.parse_column_name(other)
+      ColumnKey.new(cf, cq)
+    end
+  end
+end#ColumnKey
+end#HBase
+

data/lib/hbase-jruby/dependency.rb

@@ -0,0 +1,69 @@
+require 'java'
+require 'open-uri'
+require 'tempfile'
+
+# HBase connection
+class HBase
+  class << self
+    # Resolve Hadoop and HBase dependency with Maven or hbase command (Experimental)
+    # @param [String] dist Distribution version or path to pom.xml file
+    # @param [true, false] verbose Verbose output
+    # @return [Array<String>] Loaded JAR files
+    def resolve_dependency! dist, verbose = false
+      silencer = verbose ? '' : '> /dev/null'
+      tempfiles = []
+      jars =
+        if dist == :hbase
+          # Check for hbase executable
+          hbase = `which hbase`
+          raise RuntimeError, "Cannot find executable `hbase`" if hbase.empty?
+          `hbase classpath`.split(':')
+        else
+          # Check for Maven executable
+          mvn = `which mvn`
+          raise RuntimeError, "Cannot find executable `mvn`" if mvn.empty?
+
+          distname = dist.downcase.sub(/\.xml$/, '')
+          path = [
+            File.expand_path("../pom/#{distname}.xml", __FILE__),
+            dist.to_s,
+          ].select { |f| File.exists? f }.first
+
+          # Try github head
+          unless path
+            begin
+              xml = open("https://raw.github.com/junegunn/hbase-jruby/master/lib/hbase-jruby/pom/#{distname}.xml").read
+              tempfiles << tf = Tempfile.new("#{distname}.xml")
+              tf.close(false)
+              path = tf.path
+              File.open(path, 'w') do |f|
+                f << xml
+              end
+            rescue OpenURI::HTTPError => e
+              # No such distribution anywhere
+            end
+          end
+
+          raise ArgumentError, "Invalid distribution: #{dist}" unless path
+
+          # Download dependent JAR files and build classpath string
+          tempfiles << tf = Tempfile.new('hbase-jruby-classpath')
+          tf.close(false)
+          system "mvn org.apache.maven.plugins:maven-dependency-plugin:2.5.1:resolve org.apache.maven.plugins:maven-dependency-plugin:2.5.1:build-classpath -Dsilent=true -Dmdep.outputFile=#{tf.path} -f #{path} #{silencer}"
+
+          raise RuntimeError.new("Error occurred. Set verbose parameter to see the log.") unless $?.exitstatus == 0
+
+          File.read(tf.path).split(':')
+        end
+
+      # Load jars
+      jars.select { |jar| File.exists?(jar) && File.extname(jar) == '.jar' }.select do |jar|
+        require jar
+      end
+
+      Util.import_java_classes!
+    ensure
+      tempfiles.each { |tempfile| tempfile.unlink rescue nil }
+    end
+  end
+end

data/lib/hbase-jruby/hbase.rb

@@ -0,0 +1,77 @@
+require 'java'
+
+# @author Junegunn Choi <junegunn.c@gmail.com>
+# @!attribute [r] config
+#   @return [org.apache.hadoop.conf.Configuration]
+class HBase
+  attr_reader :config
+
+  include Admin
+
+  # Connects to HBase
+  # @param [Hash] config A key-value pairs to build HBaseConfiguration from
+  def initialize config = {}
+    Util.import_java_classes!
+
+    @config =
+      case config
+      when org.apache.hadoop.conf.Configuration
+        config
+      else
+        HBaseConfiguration.create.tap do |hbcfg|
+          config.each do |k, v|
+            hbcfg.set k.to_s, v.to_s
+          end
+        end
+      end
+    @htable_pool = HTablePool.new @config, java.lang.Integer::MAX_VALUE
+  end
+
+  # Returns an HBaseAdmin object for administration
+  # @yield [org.apache.hadoop.hbase.client.HBaseAdmin]
+  # @return [org.apache.hadoop.hbase.client.HBaseAdmin]
+  def admin
+    if block_given?
+      with_admin { |admin| yield admin }
+    else
+      HBaseAdmin.new @config
+    end
+  end
+
+  # Closes HTablePool and connection
+  # @return [nil]
+  def close
+    @htable_pool.close
+    HConnectionManager.deleteConnection(@config, true)
+  end
+
+  # Returns the list of HBase::Table instances
+  # @return [Array<HBase::Table>]
+  def tables
+    table_names.map { |tn| table(tn) }
+  end
+
+  # Returns the list of table names
+  # @return [Array<String>]
+  def table_names
+    with_admin { |admin| admin.list_tables.map(&:name_as_string) }
+  end
+
+  # Creates HBase::Table instance for the specified name
+  # @param [#to_s] table_name The name of the table
+  # @return [HBase::Table]
+  def table table_name
+    ht = HBase::Table.send :new, @config, @htable_pool, table_name
+
+    if block_given?
+      begin
+        yield ht
+      ensure
+        ht.close
+      end
+    else
+      ht
+    end
+  end
+end
+