aliyun-sdk 0.1.1

data/lib/aliyun/oss/object.rb

@@ -0,0 +1,15 @@
+# -*- encoding: utf-8 -*-
+
+module Aliyun
+  module OSS
+
+    ##
+    # Object表示OSS存储的一个对象
+    #
+    class Object < Struct::Base
+
+      attrs :key, :type, :size, :etag, :metas, :last_modified
+
+    end # Object
+  end # OSS
+end # Aliyun

data/lib/aliyun/oss/protocol.rb

@@ -0,0 +1,1432 @@
+# -*- encoding: utf-8 -*-
+
+require 'rest-client'
+require 'nokogiri'
+require 'time'
+
+module Aliyun
+  module OSS
+
+    ##
+    # Protocol implement the OSS Open API which is low-level. User
+    # should refer to {OSS::Client} for normal use.
+    #
+    class Protocol
+
+      STREAM_CHUNK_SIZE = 16 * 1024
+
+      include Logging
+
+      def initialize(config)
+        @config = config
+        @http = HTTP.new(config)
+      end
+
+      # List all the buckets.
+      # @param opts [Hash] options
+      # @option opts [String] :prefix return only those buckets
+      #  prefixed with it if specified
+      # @option opts [String] :marker return buckets after where it
+      #  indicates (exclusively). All buckets are sorted by name
+      #  alphabetically
+      # @option opts [Integer] :limit return only the first N
+      #  buckets if specified
+      # @return [Array<Bucket>, Hash] the returned buckets and a
+      #  hash including the next tokens, which includes:
+      #  * :prefix [String] the prefix used
+      #  * :delimiter [String] the delimiter used
+      #  * :marker [String] the marker used
+      #  * :limit [Integer] the limit used
+      #  * :next_marker [String] marker to continue list buckets
+      #  * :truncated [Boolean] whether there are more buckets to
+      #    be returned
+      def list_buckets(opts = {})
+        logger.info("Begin list buckets, options: #{opts}")
+
+        params = {
+          'prefix' => opts[:prefix],
+          'marker' => opts[:marker],
+          'max-keys' => opts[:limit]
+        }.reject { |_, v| v.nil? }
+
+        _, body = @http.get( {}, {:query => params})
+        doc = parse_xml(body)
+
+        buckets = doc.css("Buckets Bucket").map do |node|
+          Bucket.new(
+            {
+              :name => get_node_text(node, "Name"),
+              :location => get_node_text(node, "Location"),
+              :creation_time =>
+                get_node_text(node, "CreationDate") { |t| Time.parse(t) }
+            }, self
+          )
+        end
+
+        more = {
+          :prefix => 'Prefix',
+          :limit => 'MaxKeys',
+          :marker => 'Marker',
+          :next_marker => 'NextMarker',
+          :truncated => 'IsTruncated'
+        }.reduce({}) { |h, (k, v)|
+          value = get_node_text(doc.root, v)
+          value.nil?? h : h.merge(k => value)
+        }
+
+        update_if_exists(
+          more, {
+            :limit => ->(x) { x.to_i },
+            :truncated => ->(x) { x.to_bool }
+          }
+        )
+
+        logger.info("Done list buckets, buckets: #{buckets}, more: #{more}")
+
+        [buckets, more]
+      end
+
+      # Create a bucket
+      # @param name [String] the bucket name
+      # @param opts [Hash] options
+      # @option opts [String] :location the region where the bucket
+      #  is located
+      # @example
+      #   oss-cn-hangzhou
+      def create_bucket(name, opts = {})
+        logger.info("Begin create bucket, name: #{name}, opts: #{opts}")
+
+        location = opts[:location]
+        body = nil
+        if location
+          builder = Nokogiri::XML::Builder.new do |xml|
+            xml.CreateBucketConfiguration {
+              xml.LocationConstraint location
+            }
+          end
+          body = builder.to_xml
+        end
+
+        @http.put({:bucket => name}, {:body => body})
+
+        logger.info("Done create bucket")
+      end
+
+      # Put bucket acl
+      # @param name [String] the bucket name
+      # @param acl [String] the bucket acl
+      # @see OSS::ACL
+      def put_bucket_acl(name, acl)
+        logger.info("Begin put bucket acl, name: #{name}, acl: #{acl}")
+
+        sub_res = {'acl' => nil}
+        headers = {'x-oss-acl' => acl}
+        @http.put(
+          {:bucket => name, :sub_res => sub_res},
+          {:headers => headers, :body => nil})
+
+        logger.info("Done put bucket acl")
+      end
+
+      # Get bucket acl
+      # @param name [String] the bucket name
+      # @return [String] the acl of this bucket
+      def get_bucket_acl(name)
+        logger.info("Begin get bucket acl, name: #{name}")
+
+        sub_res = {'acl' => nil}
+        _, body = @http.get({:bucket => name, :sub_res => sub_res})
+
+        doc = parse_xml(body)
+        acl = get_node_text(doc.at_css("AccessControlList"), 'Grant')
+        logger.info("Done get bucket acl")
+
+        acl
+      end
+
+      # Put bucket logging settings
+      # @param name [String] the bucket name
+      # @param logging [BucketLogging] logging options
+      def put_bucket_logging(name, logging)
+        logger.info("Begin put bucket logging, "\
+                    "name: #{name}, logging: #{logging}")
+
+        if logging.enabled? && !logging.target_bucket
+          fail ClientError,
+               "Must specify target bucket when enabling bucket logging."
+        end
+
+        sub_res = {'logging' => nil}
+        body = Nokogiri::XML::Builder.new do |xml|
+          xml.BucketLoggingStatus {
+            if logging.enabled?
+              xml.LoggingEnabled {
+                xml.TargetBucket logging.target_bucket
+                xml.TargetPrefix logging.target_prefix if logging.target_prefix
+              }
+            end
+          }
+        end.to_xml
+
+        @http.put(
+          {:bucket => name, :sub_res => sub_res},
+          {:body => body})
+
+        logger.info("Done put bucket logging")
+      end
+
+      # Get bucket logging settings
+      # @param name [String] the bucket name
+      # @return [BucketLogging] logging options of this bucket
+      def get_bucket_logging(name)
+        logger.info("Begin get bucket logging, name: #{name}")
+
+        sub_res = {'logging' => nil}
+        _, body = @http.get({:bucket => name, :sub_res => sub_res})
+
+        doc = parse_xml(body)
+        opts = {:enable => false}
+
+        logging_node = doc.at_css("LoggingEnabled")
+        opts.update(
+          :target_bucket => get_node_text(logging_node, 'TargetBucket'),
+          :target_prefix => get_node_text(logging_node, 'TargetPrefix')
+        )
+        opts[:enable] = true if opts[:target_bucket]
+
+        logger.info("Done get bucket logging")
+
+        BucketLogging.new(opts)
+      end
+
+      # Delete bucket logging settings, a.k.a. disable bucket logging
+      # @param name [String] the bucket name
+      def delete_bucket_logging(name)
+        logger.info("Begin delete bucket logging, name: #{name}")
+
+        sub_res = {'logging' => nil}
+        @http.delete({:bucket => name, :sub_res => sub_res})
+
+        logger.info("Done delete bucket logging")
+      end
+
+      # Put bucket website settings
+      # @param name [String] the bucket name
+      # @param website [BucketWebsite] the bucket website options
+      def put_bucket_website(name, website)
+        logger.info("Begin put bucket website, "\
+                    "name: #{name}, website: #{website}")
+
+        unless website.index
+          fail ClientError, "Must specify index to put bucket website."
+        end
+
+        sub_res = {'website' => nil}
+        body = Nokogiri::XML::Builder.new do |xml|
+          xml.WebsiteConfiguration {
+            xml.IndexDocument {
+              xml.Suffix website.index
+            }
+            if website.error
+              xml.ErrorDocument {
+                xml.Key website.error
+              }
+            end
+          }
+        end.to_xml
+
+        @http.put(
+          {:bucket => name, :sub_res => sub_res},
+          {:body => body})
+
+        logger.info("Done put bucket website")
+      end
+
+      # Get bucket website settings
+      # @param name [String] the bucket name
+      # @return [BucketWebsite] the bucket website options
+      def get_bucket_website(name)
+        logger.info("Begin get bucket website, name: #{name}")
+
+        sub_res = {'website' => nil}
+        _, body = @http.get({:bucket => name, :sub_res => sub_res})
+
+        opts = {:enable => true}
+        doc = parse_xml(body)
+        opts.update(
+          :index => get_node_text(doc.at_css('IndexDocument'), 'Suffix'),
+          :error => get_node_text(doc.at_css('ErrorDocument'), 'Key')
+        )
+
+        logger.info("Done get bucket website")
+
+        BucketWebsite.new(opts)
+      end
+
+      # Delete bucket website settings
+      # @param name [String] the bucket name
+      def delete_bucket_website(name)
+        logger.info("Begin delete bucket website, name: #{name}")
+
+        sub_res = {'website' => nil}
+        @http.delete({:bucket => name, :sub_res => sub_res})
+
+        logger.info("Done delete bucket website")
+      end
+
+      # Put bucket referer
+      # @param name [String] the bucket name
+      # @param referer [BucketReferer] the bucket referer options
+      def put_bucket_referer(name, referer)
+        logger.info("Begin put bucket referer, "\
+                    "name: #{name}, referer: #{referer}")
+
+        sub_res = {'referer' => nil}
+        body = Nokogiri::XML::Builder.new do |xml|
+          xml.RefererConfiguration {
+            xml.AllowEmptyReferer referer.allow_empty?
+            xml.RefererList {
+              (referer.whitelist or []).each do |r|
+                xml.Referer r
+              end
+            }
+          }
+        end.to_xml
+
+        @http.put(
+          {:bucket => name, :sub_res => sub_res},
+          {:body => body})
+
+        logger.info("Done put bucket referer")
+      end
+
+      # Get bucket referer
+      # @param name [String] the bucket name
+      # @return [BucketReferer] the bucket referer options
+      def get_bucket_referer(name)
+        logger.info("Begin get bucket referer, name: #{name}")
+
+        sub_res = {'referer' => nil}
+        _, body = @http.get({:bucket => name, :sub_res => sub_res})
+
+        doc = parse_xml(body)
+        opts = {
+          :allow_empty =>
+            get_node_text(doc.root, 'AllowEmptyReferer', &:to_bool),
+          :whitelist => doc.css("RefererList Referer").map(&:text)
+        }
+
+        logger.info("Done get bucket referer")
+
+        BucketReferer.new(opts)
+      end
+
+      # Put bucket lifecycle settings
+      # @param name [String] the bucket name
+      # @param rules [Array<OSS::LifeCycleRule>] the
+      #  lifecycle rules
+      # @see OSS::LifeCycleRule
+      def put_bucket_lifecycle(name, rules)
+        logger.info("Begin put bucket lifecycle, name: #{name}, rules: "\
+                     "#{rules.map { |r| r.to_s }}")
+
+        sub_res = {'lifecycle' => nil}
+        body = Nokogiri::XML::Builder.new do |xml|
+          xml.LifecycleConfiguration {
+            rules.each do |r|
+              xml.Rule {
+                xml.ID r.id if r.id
+                xml.Status r.enabled? ? 'Enabled' : 'Disabled'
+
+                xml.Prefix r.prefix
+                xml.Expiration {
+                  if r.expiry.is_a?(Date)
+                    xml.Date Time.utc(
+                               r.expiry.year, r.expiry.month, r.expiry.day)
+                              .iso8601.sub('Z', '.000Z')
+                  elsif r.expiry.is_a?(Fixnum)
+                    xml.Days r.expiry
+                  else
+                    fail ClientError, "Expiry must be a Date or Fixnum."
+                  end
+                }
+              }
+            end
+          }
+        end.to_xml
+
+        @http.put(
+          {:bucket => name, :sub_res => sub_res},
+          {:body => body})
+
+        logger.info("Done put bucket lifecycle")
+      end
+
+      # Get bucket lifecycle settings
+      # @param name [String] the bucket name
+      # @return [Array<OSS::LifeCycleRule>] the
+      #  lifecycle rules. See {OSS::LifeCycleRule}
+      def get_bucket_lifecycle(name)
+        logger.info("Begin get bucket lifecycle, name: #{name}")
+
+        sub_res = {'lifecycle' => nil}
+        _, body = @http.get({:bucket => name, :sub_res => sub_res})
+
+        doc = parse_xml(body)
+        rules = doc.css("Rule").map do |n|
+          days = n.at_css("Expiration Days")
+          date = n.at_css("Expiration Date")
+
+          if (days && date) || (!days && !date)
+            fail ClientError, "We can only have one of Date and Days for expiry."
+          end
+
+          LifeCycleRule.new(
+            :id => get_node_text(n, 'ID'),
+            :prefix => get_node_text(n, 'Prefix'),
+            :enable => get_node_text(n, 'Status') { |x| x == 'Enabled' },
+            :expiry => days ? days.text.to_i : Date.parse(date.text)
+          )
+        end
+        logger.info("Done get bucket lifecycle")
+
+        rules
+      end
+
+      # Delete *all* lifecycle rules on the bucket
+      # @note this will delete all lifecycle rules
+      # @param name [String] the bucket name
+      def delete_bucket_lifecycle(name)
+        logger.info("Begin delete bucket lifecycle, name: #{name}")
+
+        sub_res = {'lifecycle' => nil}
+        @http.delete({:bucket => name, :sub_res => sub_res})
+
+        logger.info("Done delete bucket lifecycle")
+      end
+
+      # Set bucket CORS(Cross-Origin Resource Sharing) rules
+      # @param name [String] the bucket name
+      # @param rules [Array<OSS::CORSRule] the CORS
+      #  rules
+      # @see OSS::CORSRule
+      def set_bucket_cors(name, rules)
+        logger.info("Begin set bucket cors, bucket: #{name}, rules: "\
+                     "#{rules.map { |r| r.to_s }.join(';')}")
+
+        sub_res = {'cors' => nil}
+        body = Nokogiri::XML::Builder.new do |xml|
+          xml.CORSConfiguration {
+            rules.each do |r|
+              xml.CORSRule {
+                r.allowed_origins.each { |x| xml.AllowedOrigin x }
+                r.allowed_methods.each { |x| xml.AllowedMethod x }
+                r.allowed_headers.each { |x| xml.AllowedHeader x }
+                r.expose_headers.each { |x| xml.ExposeHeader x }
+                xml.MaxAgeSeconds r.max_age_seconds if r.max_age_seconds
+              }
+            end
+          }
+        end.to_xml
+
+        @http.put(
+          {:bucket => name, :sub_res => sub_res},
+          {:body => body})
+
+        logger.info("Done delete bucket lifecycle")
+      end
+
+      # Get bucket CORS rules
+      # @param name [String] the bucket name
+      # @return [Array<OSS::CORSRule] the CORS rules
+      def get_bucket_cors(name)
+        logger.info("Begin get bucket cors, bucket: #{name}")
+
+        sub_res = {'cors' => nil}
+        _, body = @http.get({:bucket => name, :sub_res => sub_res})
+
+        doc = parse_xml(body)
+        rules = []
+
+        doc.css("CORSRule").map do |n|
+          allowed_origins = n.css("AllowedOrigin").map(&:text)
+          allowed_methods = n.css("AllowedMethod").map(&:text)
+          allowed_headers = n.css("AllowedHeader").map(&:text)
+          expose_headers = n.css("ExposeHeader").map(&:text)
+          max_age_seconds = get_node_text(n, 'MaxAgeSeconds', &:to_i)
+
+          rules << CORSRule.new(
+            :allowed_origins => allowed_origins,
+            :allowed_methods => allowed_methods,
+            :allowed_headers => allowed_headers,
+            :expose_headers => expose_headers,
+            :max_age_seconds => max_age_seconds)
+        end
+
+        logger.info("Done get bucket cors")
+
+        rules
+      end
+
+      # Delete all bucket CORS rules
+      # @note this will delete all CORS rules of this bucket
+      # @param name [String] the bucket name
+      def delete_bucket_cors(name)
+        logger.info("Begin delete bucket cors, bucket: #{name}")
+
+        sub_res = {'cors' => nil}
+
+        @http.delete({:bucket => name, :sub_res => sub_res})
+
+        logger.info("Done delete bucket cors")
+      end
+
+      # Delete a bucket
+      # @param name [String] the bucket name
+      # @note it will fails if the bucket is not empty (it contains
+      #  objects)
+      def delete_bucket(name)
+        logger.info("Begin delete bucket: #{name}")
+
+        @http.delete({:bucket => name})
+
+        logger.info("Done delete bucket")
+      end
+
+      # Put an object to the specified bucket, a block is required
+      # to provide the object data.
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param opts [Hash] Options
+      # @option opts [String] :content_type the HTTP Content-Type
+      #  for the file, if not specified client will try to determine
+      #  the type itself and fall back to HTTP::DEFAULT_CONTENT_TYPE
+      #  if it fails to do so
+      # @option opts [Hash<Symbol, String>] :metas key-value pairs
+      #  that serve as the object meta which will be stored together
+      #  with the object
+      # @yield [HTTP::StreamWriter] a stream writer is
+      #  yielded to the caller to which it can write chunks of data
+      #  streamingly
+      # @example
+      #   chunk = get_chunk
+      #   put_object('bucket', 'object') { |sw| sw.write(chunk) }
+      def put_object(bucket_name, object_name, opts = {}, &block)
+        logger.debug("Begin put object, bucket: #{bucket_name}, object: "\
+                     "#{object_name}, options: #{opts}")
+
+        headers = {'Content-Type' => opts[:content_type]}
+        (opts[:metas] || {})
+          .each { |k, v| headers["x-oss-meta-#{k.to_s}"] = v.to_s }
+
+        @http.put(
+          {:bucket => bucket_name, :object => object_name},
+          {:headers => headers, :body => HTTP::StreamPayload.new(&block)})
+
+        logger.debug('Done put object')
+      end
+
+      # Append to an object of a bucket. Create an "Appendable
+      # Object" if the object does not exist. A block is required to
+      # provide the appending data.
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param position [Integer] the position to append
+      # @param opts [Hash] Options
+      # @option opts [String] :content_type the HTTP Content-Type
+      #  for the file, if not specified client will try to determine
+      #  the type itself and fall back to HTTP::DEFAULT_CONTENT_TYPE
+      #  if it fails to do so
+      # @option opts [Hash<Symbol, String>] :metas key-value pairs
+      #  that serve as the object meta which will be stored together
+      #  with the object
+      # @return [Integer] next position to append
+      # @yield [HTTP::StreamWriter] a stream writer is
+      #  yielded to the caller to which it can write chunks of data
+      #  streamingly
+      # @note
+      #   1. Can not append to a "Normal Object"
+      #   2. The position must equal to the object's size before append
+      #   3. The :content_type is only used when the object is created
+      def append_object(bucket_name, object_name, position, opts = {}, &block)
+        logger.debug("Begin append object, bucket: #{bucket_name}, object: "\
+                      "#{object_name}, position: #{position}, options: #{opts}")
+
+        sub_res = {'append' => nil, 'position' => position}
+        headers = {'Content-Type' => opts[:content_type]}
+        (opts[:metas] || {})
+          .each { |k, v| headers["x-oss-meta-#{k.to_s}"] = v.to_s }
+
+        h, _ = @http.post(
+             {:bucket => bucket_name, :object => object_name, :sub_res => sub_res},
+             {:headers => headers, :body => HTTP::StreamPayload.new(&block)})
+
+        logger.debug('Done append object')
+
+        wrap(h[:x_oss_next_append_position], &:to_i) || -1
+      end
+
+      # List objects in a bucket.
+      # @param bucket_name [String] the bucket name
+      # @param opts [Hash] options
+      # @option opts [String] :prefix return only those buckets
+      #  prefixed with it if specified
+      # @option opts [String] :marker return buckets after where it
+      #  indicates (exclusively). All buckets are sorted by name
+      #  alphabetically
+      # @option opts [Integer] :limit return only the first N
+      #  buckets if specified
+      # @option opts [String] :delimiter the delimiter to get common
+      #  prefixes of all objects
+      # @option opts [String] :encoding the encoding of object key
+      #  in the response body. Only {OSS::KeyEncoding::URL} is
+      #  supported now.
+      # @example
+      #  Assume we have the following objects:
+      #     /foo/bar/obj1
+      #     /foo/bar/obj2
+      #     ...
+      #     /foo/bar/obj9999999
+      #     /foo/xxx/
+      #  use 'foo/' as the prefix, '/' as the delimiter, the common
+      #  prefixes we get are: '/foo/bar/', '/foo/xxx/'. They are
+      #  coincidentally the sub-directories under '/foo/'. Using
+      #  delimiter we avoid list all the objects whose number may be
+      #  large.
+      # @return [Array<Objects>, Hash] the returned object and a
+      #  hash including the next tokens, which includes:
+      #  * :common_prefixes [String] the common prefixes returned
+      #  * :prefix [String] the prefix used
+      #  * :delimiter [String] the delimiter used
+      #  * :marker [String] the marker used
+      #  * :limit [Integer] the limit used
+      #  * :next_marker [String] marker to continue list objects
+      #  * :truncated [Boolean] whether there are more objects to
+      #    be returned
+      def list_objects(bucket_name, opts = {})
+        logger.debug("Begin list object, bucket: #{bucket_name}, options: #{opts}")
+
+        params = {
+          'prefix' => opts[:prefix],
+          'delimiter' => opts[:delimiter],
+          'marker' => opts[:marker],
+          'max-keys' => opts[:limit],
+          'encoding-type' => opts[:encoding]
+        }.reject { |_, v| v.nil? }
+
+        _, body = @http.get({:bucket => bucket_name}, {:query => params})
+
+        doc = parse_xml(body)
+
+        encoding = get_node_text(doc.root, 'EncodingType')
+
+        objects = doc.css("Contents").map do |node|
+          Object.new(
+            :key => get_node_text(node, "Key") { |x| decode_key(x, encoding) },
+            :type => get_node_text(node, "Type"),
+            :size => get_node_text(node, "Size", &:to_i),
+            :etag => get_node_text(node, "ETag"),
+            :last_modified =>
+              get_node_text(node, "LastModified") { |x| Time.parse(x) }
+          )
+        end || []
+
+        more = {
+          :prefix => 'Prefix',
+          :delimiter => 'Delimiter',
+          :limit => 'MaxKeys',
+          :marker => 'Marker',
+          :next_marker => 'NextMarker',
+          :truncated => 'IsTruncated',
+          :encoding => 'EncodingType'
+        }.reduce({}) { |h, (k, v)|
+          value = get_node_text(doc.root, v)
+          value.nil?? h : h.merge(k => value)
+        }
+
+        update_if_exists(
+          more, {
+            :limit => ->(x) { x.to_i },
+            :truncated => ->(x) { x.to_bool },
+            :delimiter => ->(x) { decode_key(x, encoding) },
+            :marker => ->(x) { decode_key(x, encoding) },
+            :next_marker => ->(x) { decode_key(x, encoding) }
+          }
+        )
+
+        common_prefixes = []
+        doc.css("CommonPrefixes Prefix").map do |node|
+          common_prefixes << decode_key(node.text, encoding)
+        end
+        more[:common_prefixes] = common_prefixes unless common_prefixes.empty?
+
+        logger.debug("Done list object. objects: #{objects}, more: #{more}")
+
+        [objects, more]
+      end
+
+      # Get an object from the bucket. A block is required to handle
+      # the object data chunks.
+      # @note User can get the whole object or only part of it by specify
+      #  the bytes range;
+      # @note User can specify conditions to get the object like:
+      #  if-modified-since, if-unmodified-since, if-match-etag,
+      #  if-unmatch-etag. If the object to get fails to meet the
+      #  conditions, it will not be returned;
+      # @note User can indicate the server to rewrite the response headers
+      #  such as content-type, content-encoding when get the object
+      #  by specify the :rewrite options. The specified headers will
+      #  be returned instead of the original property of the object.
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param opts [Hash] options
+      # @option opts [Array<Integer>] :range bytes range to get from
+      #  the object, in the format: xx-yy
+      # @option opts [Hash] :condition preconditions to get the object
+      #   * :if_modified_since (Time) get the object if its modified
+      #     time is later than specified
+      #   * :if_unmodified_since (Time) get the object if its
+      #     unmodified time if earlier than specified
+      #   * :if_match_etag (String) get the object if its etag match
+      #     specified
+      #   * :if_unmatch_etag (String) get the object if its etag
+      #     doesn't match specified
+      # @option opts [Hash] :rewrite response headers to rewrite
+      #   * :content_type (String) the Content-Type header
+      #   * :content_language (String) the Content-Language header
+      #   * :expires (Time) the Expires header
+      #   * :cache_control (String) the Cache-Control header
+      #   * :content_disposition (String) the Content-Disposition header
+      #   * :content_encoding (String) the Content-Encoding header
+      # @return [OSS::Object] The object meta
+      # @yield [String] it gives the data chunks of the object to
+      #  the block
+      def get_object(bucket_name, object_name, opts = {}, &block)
+        logger.debug("Begin get object, bucket: #{bucket_name}, "\
+                     "object: #{object_name}")
+
+        range = opts[:range]
+        conditions = opts[:condition]
+        rewrites = opts[:rewrite]
+
+        headers = {}
+        headers['Range'] = get_bytes_range(range) if range
+        headers.merge!(get_conditions(conditions)) if conditions
+
+        sub_res = {}
+        if rewrites
+          [ :content_type,
+            :content_language,
+            :cache_control,
+            :content_disposition,
+            :content_encoding
+          ].each do |k|
+            key = "response-#{k.to_s.sub('_', '-')}"
+            sub_res[key] = rewrites[k] if rewrites.key?(k)
+          end
+          sub_res["response-expires"] =
+            rewrites[:expires].httpdate if rewrites.key?(:expires)
+        end
+
+        h, _ = @http.get(
+             {:bucket => bucket_name, :object => object_name,
+              :sub_res => sub_res},
+             {:headers => headers}
+           ) { |chunk| yield chunk if block_given? }
+
+        metas = {}
+        meta_prefix = 'x_oss_meta_'
+        h.select { |k, _| k.to_s.start_with?(meta_prefix) }
+          .each { |k, v| metas[k.to_s.sub(meta_prefix, '')] = v.to_s }
+
+        obj = Object.new(
+          :key => object_name,
+          :type => h[:x_oss_object_type],
+          :size => wrap(h[:content_length], &:to_i),
+          :etag => h[:etag],
+          :metas => metas,
+          :last_modified => wrap(h[:last_modified]) { |x| Time.parse(x) })
+
+        logger.debug("Done get object")
+
+        obj
+      end
+
+      # Get the object meta rather than the whole object.
+      # @note User can specify conditions to get the object like:
+      #  if-modified-since, if-unmodified-since, if-match-etag,
+      #  if-unmatch-etag. If the object to get fails to meet the
+      #  conditions, it will not be returned.
+      #
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param opts [Hash] options
+      # @option opts [Hash] :condition preconditions to get the
+      #  object meta. The same as #get_object
+      # @return [OSS::Object] The object meta
+      def get_object_meta(bucket_name, object_name, opts = {})
+        logger.debug("Begin get object meta, bucket: #{bucket_name}, "\
+                     "object: #{object_name}, options: #{opts}")
+
+        headers = {}
+        headers.merge!(get_conditions(opts[:condition])) if opts[:condition]
+
+        h, _ = @http.head(
+             {:bucket => bucket_name, :object => object_name},
+             {:headers => headers})
+
+        metas = {}
+        meta_prefix = 'x_oss_meta_'
+        h.select { |k, _| k.to_s.start_with?(meta_prefix) }
+          .each { |k, v| metas[k.to_s.sub(meta_prefix, '')] = v.to_s }
+
+        obj = Object.new(
+          :key => object_name,
+          :type => h[:x_oss_object_type],
+          :size => wrap(h[:content_length], &:to_i),
+          :etag => h[:etag],
+          :metas => metas,
+          :last_modified => wrap(h[:last_modified]) { |x| Time.parse(x) })
+
+        logger.debug("Done get object meta")
+
+        obj
+      end
+
+      # Copy an object in the bucket. The source object and the dest
+      # object must be in the same bucket.
+      # @param bucket_name [String] the bucket name
+      # @param src_object_name [String] the source object name
+      # @param dst_object_name [String] the dest object name
+      # @param opts [Hash] options
+      # @option opts [String] :acl specify the dest object's
+      #  ACL. See {OSS::ACL}
+      # @option opts [String] :meta_directive specify what to do
+      #  with the object's meta: copy or replace. See
+      #  {OSS::MetaDirective}
+      # @option opts [String] :content_type the HTTP Content-Type
+      #  for the file, if not specified client will try to determine
+      #  the type itself and fall back to HTTP::DEFAULT_CONTENT_TYPE
+      #  if it fails to do so
+      # @option opts [Hash<Symbol, String>] :metas key-value pairs
+      #  that serve as the object meta which will be stored together
+      #  with the object
+      # @option opts [Hash] :condition preconditions to get the
+      #  object. See #get_object
+      # @return [Hash] the copy result
+      #  * :etag [String] the etag of the dest object
+      #  * :last_modified [Time] the last modification time of the
+      #    dest object
+      def copy_object(bucket_name, src_object_name, dst_object_name, opts = {})
+        logger.debug("Begin copy object, bucket: #{bucket_name}, "\
+                     "source object: #{src_object_name}, dest object: "\
+                     "#{dst_object_name}, options: #{opts}")
+
+        headers = {
+          'x-oss-copy-source' =>
+            @http.get_resource_path(bucket_name, src_object_name),
+          'Content-Type' => opts[:content_type]
+        }
+        (opts[:metas] || {})
+          .each { |k, v| headers["x-oss-meta-#{k.to_s}"] = v.to_s }
+
+        {
+          :acl => 'x-oss-object-acl',
+          :meta_directive => 'x-oss-metadata-directive'
+        }.each { |k, v| headers[v] = opts[k] if opts[k] }
+
+        headers.merge!(get_copy_conditions(opts[:condition])) if opts[:condition]
+
+        _, body = @http.put(
+          {:bucket => bucket_name, :object => dst_object_name},
+          {:headers => headers})
+
+        doc = parse_xml(body)
+        copy_result = {
+          :last_modified => get_node_text(
+            doc.root, 'LastModified') { |x| Time.parse(x) },
+          :etag => get_node_text(doc.root, 'ETag')
+        }.reject { |_, v| v.nil? }
+
+        logger.debug("Done copy object")
+
+        copy_result
+      end
+
+      # Delete an object from the bucket
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      def delete_object(bucket_name, object_name)
+        logger.debug("Begin delete object, bucket: #{bucket_name}, "\
+                     "object:  #{object_name}")
+
+        @http.delete({:bucket => bucket_name, :object => object_name})
+
+        logger.debug("Done delete object")
+      end
+
+      # Batch delete objects
+      # @param bucket_name [String] the bucket name
+      # @param object_names [Enumerator<String>] the object names
+      # @param opts [Hash] options
+      # @option opts [Boolean] :quiet indicates whether the server
+      #  should return the delete result of the objects
+      # @option opts [String] :encoding-type the encoding type for
+      #  object key in the response body, only
+      #  {OSS::KeyEncoding::URL} is supported now
+      # @return [Array<String>] object names that have been
+      #  successfully deleted or empty if :quiet is true
+      def batch_delete_objects(bucket_name, object_names, opts = {})
+        logger.debug("Begin batch delete object, bucket: #{bucket_name}, "\
+                     "objects: #{object_names}, options: #{opts}")
+
+        sub_res = {'delete' => nil}
+        body = Nokogiri::XML::Builder.new do |xml|
+          xml.Delete {
+            xml.Quiet opts[:quiet]? true : false
+            object_names.each do |o|
+              xml.Object {
+                xml.Key o
+              }
+            end
+          }
+        end.to_xml
+
+        query = {}
+        query['encoding-type'] = opts[:encoding] if opts[:encoding]
+
+        _, body = @http.post(
+             {:bucket => bucket_name, :sub_res => sub_res},
+             {:query => query, :body => body})
+
+        deleted = []
+        unless opts[:quiet]
+          doc = parse_xml(body)
+          encoding = get_node_text(doc.root, 'EncodingType')
+          doc.css("Deleted").map do |n|
+            deleted << get_node_text(n, 'Key') { |x| decode_key(x, encoding) }
+          end
+        end
+
+        logger.debug("Done delete object")
+
+        deleted
+      end
+
+      # Put object acl
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param acl [String] the object's ACL. See {OSS::ACL}
+      def put_object_acl(bucket_name, object_name, acl)
+        logger.debug("Begin update object acl, bucket: #{bucket_name}, "\
+                     "object: #{object_name}, acl: #{acl}")
+
+        sub_res = {'acl' => nil}
+        headers = {'x-oss-object-acl' => acl}
+
+        @http.put(
+          {:bucket => bucket_name, :object => object_name, :sub_res => sub_res},
+          {:headers => headers})
+
+        logger.debug("Done update object acl")
+      end
+
+      # Get object acl
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # [return] the object's acl. See {OSS::ACL}
+      def get_object_acl(bucket_name, object_name)
+        logger.debug("Begin get object acl, bucket: #{bucket_name}, "\
+                     "object: #{object_name}")
+
+        sub_res = {'acl' => nil}
+        _, body = @http.get(
+             {bucket: bucket_name, object: object_name, sub_res: sub_res})
+
+        doc = parse_xml(body)
+        acl = get_node_text(doc.at_css("AccessControlList"), 'Grant')
+
+        logger.debug("Done get object acl")
+
+        acl
+      end
+
+      # Get object CORS rule
+      # @note this is usually used by browser to make a "preflight"
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param origin [String] the Origin of the reqeust
+      # @param method [String] the method to request access:
+      #  Access-Control-Request-Method
+      # @param headers [Array<String>] the headers to request access:
+      #  Access-Control-Request-Headers
+      # @return [CORSRule] the CORS rule of the object
+      def get_object_cors(bucket_name, object_name, origin, method, headers = [])
+        logger.debug("Begin get object cors, bucket: #{bucket_name}, object: "\
+                     "#{object_name}, origin: #{origin}, method: #{method}, "\
+                     "headers: #{headers.join(',')}")
+
+        h = {
+          'Origin' => origin,
+          'Access-Control-Request-Method' => method,
+          'Access-Control-Request-Headers' => headers.join(',')
+        }
+
+        return_headers, _ = @http.options(
+                          {:bucket => bucket_name, :object => object_name},
+                          {:headers => h})
+
+        logger.debug("Done get object cors")
+
+        CORSRule.new(
+          :allowed_origins => return_headers[:access_control_allow_origin],
+          :allowed_methods => return_headers[:access_control_allow_methods],
+          :allowed_headers => return_headers[:access_control_allow_headers],
+          :expose_headers => return_headers[:access_control_expose_headers],
+          :max_age_seconds => return_headers[:access_control_max_age]
+        )
+      end
+
+      ##
+      # Multipart uploading
+      #
+
+      # Initiate a a multipart uploading transaction
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param opts [Hash] options
+      # @option opts [String] :content_type the HTTP Content-Type
+      #  for the file, if not specified client will try to determine
+      #  the type itself and fall back to HTTP::DEFAULT_CONTENT_TYPE
+      #  if it fails to do so
+      # @option opts [Hash<Symbol, String>] :metas key-value pairs
+      #  that serve as the object meta which will be stored together
+      #  with the object
+      # @return [String] the upload id
+      def initiate_multipart_upload(bucket_name, object_name, opts = {})
+        logger.info("Begin initiate multipart upload, bucket: "\
+                    "#{bucket_name}, object: #{object_name}, options: #{opts}")
+
+        sub_res = {'uploads' => nil}
+        headers = {'Content-Type' => opts[:content_type]}
+        (opts[:metas] || {})
+          .each { |k, v| headers["x-oss-meta-#{k.to_s}"] = v.to_s }
+
+        _, body = @http.post(
+             {:bucket => bucket_name, :object => object_name,
+              :sub_res => sub_res},
+             {:headers => headers})
+
+        doc = parse_xml(body)
+        txn_id = get_node_text(doc.root, 'UploadId')
+
+        logger.info("Done initiate multipart upload: #{txn_id}.")
+
+        txn_id
+      end
+
+      # Upload a part in a multipart uploading transaction.
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param txn_id [String] the upload id
+      # @param part_no [Integer] the part number
+      # @yield [HTTP::StreamWriter] a stream writer is
+      #  yielded to the caller to which it can write chunks of data
+      #  streamingly
+      def upload_part(bucket_name, object_name, txn_id, part_no, &block)
+        logger.debug("Begin upload part, bucket: #{bucket_name}, object: "\
+                     "#{object_name}, txn id: #{txn_id}, part No: #{part_no}")
+
+        sub_res = {'partNumber' => part_no, 'uploadId' => txn_id}
+        headers, _ = @http.put(
+          {:bucket => bucket_name, :object => object_name, :sub_res => sub_res},
+          {:body => HTTP::StreamPayload.new(&block)})
+
+        logger.debug("Done upload part")
+
+        Multipart::Part.new(:number => part_no, :etag => headers[:etag])
+      end
+
+      # Upload a part in a multipart uploading transaction by copying
+      # from an existent object as the part's content. It may copy
+      # only part of the object by specifying the bytes range to read.
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param txn_id [String] the upload id
+      # @param part_no [Integer] the part number
+      # @param source_object [String] the source object name to copy from
+      # @param opts [Hash] options
+      # @option opts [Array<Integer>] :range the bytes range to
+      #  copy, int the format: [begin(inclusive), end(exclusive]
+      # @option opts [Hash] :condition preconditions to copy the
+      #  object. See #get_object
+      def upload_part_by_copy(
+            bucket_name, object_name, txn_id, part_no, source_object, opts = {})
+        logger.debug("Begin upload part by copy, bucket: #{bucket_name}, "\
+                     "object: #{object_name}, source object: #{source_object}"\
+                     "txn id: #{txn_id}, part No: #{part_no}, options: #{opts}")
+
+        range = opts[:range]
+        conditions = opts[:condition]
+
+        if range && (!range.is_a?(Array) || range.size != 2)
+          fail ClientError, "Range must be an array containing 2 Integers."
+        end
+
+        headers = {
+          'x-oss-copy-source' =>
+            @http.get_resource_path(bucket_name, source_object)
+        }
+        headers['Range'] = get_bytes_range(range) if range
+        headers.merge!(get_copy_conditions(conditions)) if conditions
+
+        sub_res = {'partNumber' => part_no, 'uploadId' => txn_id}
+
+        headers, _ = @http.put(
+          {:bucket => bucket_name, :object => object_name, :sub_res => sub_res},
+          {:headers => headers})
+
+        logger.debug("Done upload part by copy: #{source_object}.")
+
+        Multipart::Part.new(:number => part_no, :etag => headers[:etag])
+      end
+
+      # Complete a multipart uploading transaction
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param txn_id [String] the upload id
+      # @param parts [Array<Multipart::Part>] all the
+      #  parts in this transaction
+      def complete_multipart_upload(bucket_name, object_name, txn_id, parts)
+        logger.debug("Begin complete multipart upload, "\
+                     "txn id: #{txn_id}, parts: #{parts.map(&:to_s)}")
+
+        sub_res = {'uploadId' => txn_id}
+
+        body = Nokogiri::XML::Builder.new do |xml|
+          xml.CompleteMultipartUpload {
+            parts.each do |p|
+              xml.Part {
+                xml.PartNumber p.number
+                xml.ETag p.etag
+              }
+            end
+          }
+        end.to_xml
+
+        @http.post(
+          {:bucket => bucket_name, :object => object_name, :sub_res => sub_res},
+          {:body => body})
+
+        logger.debug("Done complete multipart upload: #{txn_id}.")
+      end
+
+      # Abort a multipart uploading transaction
+      # @note All the parts are discarded after abort. For some parts
+      #  being uploaded while the abort happens, they may not be
+      #  discarded. Call abort_multipart_upload several times for this
+      #  situation.
+      # @param bucket_name [String] the bucket name
+      # @param object_name [String] the object name
+      # @param txn_id [String] the upload id
+      def abort_multipart_upload(bucket_name, object_name, txn_id)
+        logger.debug("Begin abort multipart upload, txn id: #{txn_id}")
+
+        sub_res = {'uploadId' => txn_id}
+
+        @http.delete(
+          {:bucket => bucket_name, :object => object_name, :sub_res => sub_res})
+
+        logger.debug("Done abort multipart: #{txn_id}.")
+      end
+
+      # Get a list of all the on-going multipart uploading
+      # transactions. That is: thoses started and not aborted.
+      # @param bucket_name [String] the bucket name
+      # @param opts [Hash] options:
+      # @option opts [String] :id_marker return only thoese transactions with
+      #  txn id after :id_marker
+      # @option opts [String] :key_marker the object key marker for
+      #  a multipart upload transaction.
+      #  1. if +:id_marker+ is not set, return only those
+      #     transactions with object key *after* +:key_marker+;
+      #  2. if +:id_marker+ is set, return only thoese transactions
+      #     with object key *equals* +:key_marker+ and txn id after
+      #     +:id_marker+
+      # @option opts [String] :prefix the prefix of the object key
+      #  for a multipart upload transaction. if set only return
+      #  those transactions with the object key prefixed with it
+      # @option opts [String] :delimiter the delimiter for the
+      #  object key for a multipart upload transaction.
+      # @option opts [String] :encoding the encoding of object key
+      #  in the response body. Only {OSS::KeyEncoding::URL} is
+      #  supported now.
+      # @return [Array<Multipart::Transaction>, Hash]
+      #  the returned transactions and a hash including next tokens,
+      #  which includes:
+      #  * :prefix [String] the prefix used
+      #  * :delimiter [String] the delimiter used
+      #  * :limit [Integer] the limit used
+      #  * :id_marker [String] the upload id marker used
+      #  * :next_id_marker [String] upload id marker to continue list
+      #    multipart transactions
+      #  * :key_marker [String] the object key marker used
+      #  * :next_key_marker [String] object key marker to continue
+      #    list multipart transactions
+      #  * :truncated [Boolean] whether there are more transactions
+      #    to be returned
+      #  * :encoding [String] the object key encoding used
+      def list_multipart_uploads(bucket_name, opts = {})
+        logger.debug("Begin list multipart uploads, "\
+                     "bucket: #{bucket_name}, opts: #{opts}")
+
+        sub_res = {'uploads' => nil}
+        params = {
+          'prefix' => opts[:prefix],
+          'delimiter' => opts[:delimiter],
+          'upload-id-marker' => opts[:id_marker],
+          'key-marker' => opts[:key_marker],
+          'max-uploads' => opts[:limit],
+          'encoding-type' => opts[:encoding]
+        }.reject { |_, v| v.nil? }
+
+        _, body = @http.get(
+          {:bucket => bucket_name, :sub_res => sub_res},
+          {:query => params})
+
+        doc = parse_xml(body)
+
+        encoding = get_node_text(doc.root, 'EncodingType')
+
+        txns = doc.css("Upload").map do |node|
+          Multipart::Transaction.new(
+            :id => get_node_text(node, "UploadId"),
+            :object => get_node_text(node, "Key") { |x| decode_key(x, encoding) },
+            :bucket => bucket_name,
+            :creation_time =>
+              get_node_text(node, "Initiated") { |t| Time.parse(t) }
+          )
+        end || []
+
+        more = {
+          :prefix => 'Prefix',
+          :delimiter => 'Delimiter',
+          :limit => 'MaxUploads',
+          :id_marker => 'UploadIdMarker',
+          :next_id_marker => 'NextUploadIdMarker',
+          :key_marker => 'KeyMarker',
+          :next_key_marker => 'NextKeyMarker',
+          :truncated => 'IsTruncated',
+          :encoding => 'EncodingType'
+        }.reduce({}) { |h, (k, v)|
+          value = get_node_text(doc.root, v)
+          value.nil?? h : h.merge(k => value)
+        }
+
+        update_if_exists(
+          more, {
+            :limit => ->(x) { x.to_i },
+            :truncated => ->(x) { x.to_bool },
+            :delimiter => ->(x) { decode_key(x, encoding) },
+            :key_marker => ->(x) { decode_key(x, encoding) },
+            :next_key_marker => ->(x) { decode_key(x, encoding) }
+          }
+        )
+
+        logger.debug("Done list multipart transactions")
+
+        [txns, more]
+      end
+
+      # Get a list of parts that are successfully uploaded in a
+      # transaction.
+      # @param txn_id [String] the upload id
+      # @param opts [Hash] options:
+      # @option opts [Integer] :marker the part number marker after
+      #  which to return parts
+      # @option opts [Integer] :limit max number parts to return
+      # @return [Array<Multipart::Part>, Hash] the returned parts and
+      #  a hash including next tokens, which includes:
+      #  * :marker [Integer] the marker used
+      #  * :limit [Integer] the limit used
+      #  * :next_marker [Integer] marker to continue list parts
+      #  * :truncated [Boolean] whether there are more parts to be
+      #    returned
+      def list_parts(bucket_name, object_name, txn_id, opts = {})
+        logger.debug("Begin list parts, bucket: #{bucket_name}, object: "\
+                     "#{object_name}, txn id: #{txn_id}, options: #{opts}")
+
+        sub_res = {'uploadId' => txn_id}
+        params = {
+          'part-number-marker' => opts[:marker],
+          'max-parts' => opts[:limit],
+          'encoding-type' => opts[:encoding]
+        }.reject { |_, v| v.nil? }
+
+        _, body = @http.get(
+          {:bucket => bucket_name, :object => object_name, :sub_res => sub_res},
+          {:query => params})
+
+        doc = parse_xml(body)
+        parts = doc.css("Part").map do |node|
+          Multipart::Part.new(
+            :number => get_node_text(node, 'PartNumber', &:to_i),
+            :etag => get_node_text(node, 'ETag'),
+            :size => get_node_text(node, 'Size', &:to_i),
+            :last_modified =>
+              get_node_text(node, 'LastModified') { |x| Time.parse(x) })
+        end || []
+
+        more = {
+          :limit => 'MaxParts',
+          :marker => 'PartNumberMarker',
+          :next_marker => 'NextPartNumberMarker',
+          :truncated => 'IsTruncated',
+          :encoding => 'EncodingType'
+        }.reduce({}) { |h, (k, v)|
+          value = get_node_text(doc.root, v)
+          value.nil?? h : h.merge(k => value)
+        }
+
+        update_if_exists(
+          more, {
+            :limit => ->(x) { x.to_i },
+            :truncated => ->(x) { x.to_bool }
+          }
+        )
+
+        logger.debug("Done list parts, parts: #{parts}, more: #{more}")
+
+        [parts, more]
+      end
+
+      # Get bucket/object url
+      # @param [String] bucket the bucket name
+      # @param [String] object the bucket name
+      # @return [String] url for the bucket/object
+      def get_request_url(bucket, object = nil)
+        @http.get_request_url(bucket, object)
+      end
+
+      # Get user's access key id
+      # @return [String] the access key id
+      def get_access_key_id
+        @config.access_key_id
+      end
+
+      # Sign a string using the stored access key secret
+      # @param [String] string_to_sign the string to sign
+      # @return [String] the signature
+      def sign(string_to_sign)
+        Util.sign(@config.access_key_secret, string_to_sign)
+      end
+
+      private
+
+      # Parse body content to xml document
+      # @param content [String] the xml content
+      # @return [Nokogiri::XML::Document] the parsed document
+      def parse_xml(content)
+        doc = Nokogiri::XML(content) do |config|
+          config.options |= Nokogiri::XML::ParseOptions::NOBLANKS
+        end
+
+        doc
+      end
+
+      # Get the text of a xml node
+      # @param node [Nokogiri::XML::Node] the xml node
+      # @param tag [String] the node tag
+      # @yield [String] the node text is given to the block
+      def get_node_text(node, tag, &block)
+        n = node.at_css(tag) if node
+        value = n.text if n
+        block && value ? yield(value) : value
+      end
+
+      # Decode object key using encoding. If encoding is nil it
+      # returns the key directly.
+      # @param key [String] the object key
+      # @param encoding [String] the encoding used
+      # @return [String] the decoded key
+      def decode_key(key, encoding)
+        return key unless encoding
+
+        unless KeyEncoding.include?(encoding)
+          fail ClientError, "Unsupported key encoding: #{encoding}"
+        end
+
+        if encoding == KeyEncoding::URL
+          return CGI.unescape(key)
+        end
+      end
+
+      # Transform x if x is not nil
+      # @param x [Object] the object to transform
+      # @yield [Object] the object if given to the block
+      # @return [Object] the transformed object
+      def wrap(x, &block)
+        yield x if x
+      end
+
+      # Get conditions for HTTP headers
+      # @param conditions [Hash] the conditions
+      # @return [Hash] conditions for HTTP headers
+      def get_conditions(conditions)
+        {
+          :if_modified_since => 'If-Modified-Since',
+          :if_unmodified_since => 'If-Unmodified-Since',
+        }.reduce({}) { |h, (k, v)|
+          conditions.key?(k)? h.merge(v => conditions[k].httpdate) : h
+        }.merge(
+          {
+            :if_match_etag => 'If-Match',
+            :if_unmatch_etag => 'If-None-Match'
+          }.reduce({}) { |h, (k, v)|
+            conditions.key?(k)? h.merge(v => conditions[k]) : h
+          }
+        )
+      end
+
+      # Get copy conditions for HTTP headers
+      # @param conditions [Hash] the conditions
+      # @return [Hash] copy conditions for HTTP headers
+      def get_copy_conditions(conditions)
+        {
+          :if_modified_since => 'x-oss-copy-source-if-modified-since',
+          :if_unmodified_since => 'x-oss-copy-source-if-unmodified-since',
+        }.reduce({}) { |h, (k, v)|
+          conditions.key?(k)? h.merge(v => conditions[k].httpdate) : h
+        }.merge(
+          {
+            :if_match_etag => 'x-oss-copy-source-if-match',
+            :if_unmatch_etag => 'x-oss-copy-source-if-none-match'
+          }.reduce({}) { |h, (k, v)|
+            conditions.key?(k)? h.merge(v => conditions[k]) : h
+          }
+        )
+      end
+
+      # Get bytes range
+      # @param range [Array<Integer>] range
+      # @return [String] bytes range for HTTP headers
+      def get_bytes_range(range)
+        if range &&
+           (!range.is_a?(Array) || range.size != 2 ||
+            !range.at(0).is_a?(Fixnum) || !range.at(1).is_a?(Fixnum))
+          fail ClientError, "Range must be an array containing 2 Integers."
+        end
+
+        "bytes=#{range.at(0)}-#{range.at(1) - 1}"
+      end
+
+      # Update values for keys that exist in hash
+      # @param hash [Hash] the hash to be updated
+      # @param kv [Hash] keys & blocks to updated
+      def update_if_exists(hash, kv)
+        kv.each { |k, v| hash[k] = v.call(hash[k]) if hash.key?(k) }
+      end
+
+    end # Protocol
+  end # OSS
+end # Aliyun