scashin133-s3 0.3.8

data/lib/s3/connection.rb

@@ -0,0 +1,222 @@
+module S3
+
+  # Class responsible for handling connections to amazon hosts
+  class Connection
+    include Parser
+
+    attr_accessor :access_key_id, :secret_access_key, :use_ssl, :timeout, :debug, :proxy
+    alias :use_ssl? :use_ssl
+
+    # Creates new connection object.
+    #
+    # ==== Options
+    # * <tt>:access_key_id</tt> - Access key id (REQUIRED)
+    # * <tt>:secret_access_key</tt> - Secret access key (REQUIRED)
+    # * <tt>:use_ssl</tt> - Use https or http protocol (false by
+    #   default)
+    # * <tt>:debug</tt> - Display debug information on the STDOUT
+    #   (false by default)
+    # * <tt>:timeout</tt> - Timeout to use by the Net::HTTP object
+    #   (60 by default)
+    # * <tt>:proxy</tt> - Hash for Net::HTTP Proxy settings
+    #   { :host => "proxy.mydomain.com", :port => "80, :user => "user_a", :password => "secret" }
+    # * <tt>:proxy</tt> - Hash for Net::HTTP Proxy settings
+    # * <tt>:chunk_size</tt> - Size of a chunk when streaming
+    #   (1048576 (1 MiB) by default)
+    def initialize(options = {})
+      @access_key_id = options.fetch(:access_key_id)
+      @secret_access_key = options.fetch(:secret_access_key)
+      @use_ssl = options.fetch(:use_ssl, false)
+      @debug = options.fetch(:debug, false)
+      @timeout = options.fetch(:timeout, 60)
+      @proxy = options.fetch(:proxy, nil)
+      @chunk_size = options.fetch(:chunk_size, 1048576)
+    end
+
+    # Makes request with given HTTP method, sets missing parameters,
+    # adds signature to request header and returns response object
+    # (Net::HTTPResponse)
+    #
+    # ==== Parameters
+    # * <tt>method</tt> - HTTP Method symbol, can be <tt>:get</tt>,
+    #   <tt>:put</tt>, <tt>:delete</tt>
+    #
+    # ==== Options:
+    # * <tt>:host</tt> - Hostname to connecto to, defaults
+    #   to <tt>s3.amazonaws.com</tt>
+    # * <tt>:path</tt> - path to send request to (REQUIRED)
+    # * <tt>:body</tt> - Request body, only meaningful for
+    #   <tt>:put</tt> request
+    # * <tt>:params</tt> - Parameters to add to query string for
+    #   request, can be String or Hash
+    # * <tt>:headers</tt> - Hash of headers fields to add to request
+    #   header
+    #
+    # ==== Returns
+    # Net::HTTPResponse object -- response from the server
+    def request(method, options)
+      host = options.fetch(:host, HOST)
+      path = options.fetch(:path)
+      body = options.fetch(:body, nil)
+      params = options.fetch(:params, {})
+      headers = options.fetch(:headers, {})
+
+      if params
+        params = params.is_a?(String) ? params : self.class.parse_params(params)
+        path << "?#{params}"
+      end
+
+      path = URI.escape(path)
+      request = Request.new(@chunk_size, method.to_s.upcase, !!body, method.to_s.upcase != "HEAD", path)
+
+      headers = self.class.parse_headers(headers)
+      headers.each do |key, value|
+        request[key] = value
+      end
+
+      if body
+        if body.respond_to?(:read)
+          request.body_stream = body
+        else
+          request.body = body
+        end
+        request.content_length = body.respond_to?(:lstat) ? body.stat.size : body.size
+      end
+
+      send_request(host, request)
+    end
+
+    # Helper function to parser parameters and create single string of
+    # params added to questy string
+    #
+    # ==== Parameters
+    # * <tt>params</tt> - Hash of parameters
+    #
+    # ==== Returns
+    # String -- containing all parameters joined in one params string,
+    # i.e. <tt>param1=val&param2&param3=0</tt>
+    def self.parse_params(params)
+      interesting_keys = [:max_keys, :prefix, :marker, :delimiter, :location]
+
+      result = []
+      params.each do |key, value|
+        if interesting_keys.include?(key)
+          parsed_key = key.to_s.gsub("_", "-")
+          case value
+          when nil
+            result << parsed_key
+          else
+            result << "#{parsed_key}=#{value}"
+          end
+        end
+      end
+      result.join("&")
+    end
+
+    # Helper function to change headers from symbols, to in correct
+    # form (i.e. with '-' instead of '_')
+    #
+    # ==== Parameters
+    # * <tt>headers</tt> - Hash of pairs <tt>headername => value</tt>,
+    #   where value can be Range (for Range header) or any other value
+    #   which can be translated to string
+    #
+    # ==== Returns
+    # Hash of headers translated from symbol to string, containing
+    # only interesting headers
+    def self.parse_headers(headers)
+      interesting_keys = [:content_type, :cache_control, :x_amz_acl, :x_amz_storage_class, :range,
+                          :if_modified_since, :if_unmodified_since,
+                          :if_match, :if_none_match,
+                          :content_disposition, :content_encoding,
+                          :x_amz_copy_source, :x_amz_metadata_directive,
+                          :x_amz_copy_source_if_match,
+                          :x_amz_copy_source_if_none_match,
+                          :x_amz_copy_source_if_unmodified_since,
+                          :x_amz_copy_source_if_modified_since]
+
+      parsed_headers = {}
+      if headers
+        headers.each do |key, value|
+          if interesting_keys.include?(key)
+            parsed_key = key.to_s.gsub("_", "-")
+            parsed_value = value
+            case value
+            when Range
+              parsed_value = "bytes=#{value.first}-#{value.last}"
+            end
+            parsed_headers[parsed_key] = parsed_value
+          end
+        end
+      end
+      parsed_headers
+    end
+
+    private
+
+    def port
+      use_ssl ? 443 : 80
+    end
+
+    def proxy_settings
+      @proxy.values_at(:host, :port, :user, :password) unless @proxy.nil? || @proxy.empty?
+    end
+
+    def http(host)
+      http = Net::HTTP.new(host, port, *proxy_settings)
+      http.set_debug_output(STDOUT) if @debug
+      http.use_ssl = @use_ssl
+      http.verify_mode = OpenSSL::SSL::VERIFY_NONE if @use_ssl
+      http.read_timeout = @timeout if @timeout
+      http
+    end
+
+    def send_request(host, request, skip_authorization = false)
+      response = http(host).start do |http|
+        host = http.address
+
+        request["Date"] ||= Time.now.httpdate
+
+        if request.body
+          request["Content-Type"] ||= "application/octet-stream"
+          request["Content-MD5"] = Base64.encode64(Digest::MD5.digest(request.body)).chomp unless request.body.empty?
+        end
+
+        unless skip_authorization
+          request["Authorization"] = Signature.generate(:host => host,
+                                                        :request => request,
+                                                        :access_key_id => access_key_id,
+                                                        :secret_access_key => secret_access_key)
+        end
+
+        http.request(request)
+      end
+
+      if response.code.to_i == 307
+        if response.body
+          doc = Document.new response.body
+          send_request(doc.elements["Error"].elements["Endpoint"].text, request, true)
+        end
+      else
+        handle_response(response)
+      end
+    end
+
+    def handle_response(response)
+      case response.code.to_i
+      when 200...300
+        response
+      when 300...600
+        if response.body.nil? || response.body.empty?
+          raise Error::ResponseError.new(nil, response)
+        else
+          code, message = parse_error(response.body)
+          raise Error::ResponseError.exception(code).new(message, response)
+        end
+      else
+        raise(ConnectionError.new(response, "Unknown response code: #{response.code}"))
+      end
+      response
+    end
+  end
+end

data/lib/s3/exceptions.rb

@@ -0,0 +1,110 @@
+module S3
+  module Error
+
+    # All responses with a code between 300 and 599 that contain an
+    # <Error></Error> body are wrapped in an ErrorResponse which
+    # contains an Error object. This Error class generates a custom
+    # exception with the name of the xml Error and its message. All
+    # such runtime generated exception classes descend from
+    # ResponseError and contain the ErrorResponse object so that all
+    # code that makes a request can rescue ResponseError and get
+    # access to the ErrorResponse.
+    class ResponseError < StandardError
+      attr_reader :response
+
+      # Creates new S3::ResponseError.
+      #
+      # ==== Parameters
+      # * <tt>message</tt> - what went wrong
+      # * <tt>response</tt> - Net::HTTPResponse object or nil
+      def initialize(message, response)
+        @response = response
+        super(message)
+      end
+
+      # Factory for all other Exception classes in module, each for
+      # every error response available from AmazonAWS
+      #
+      # ==== Parameters
+      # * <tt>code</tt> - Code name of exception
+      #
+      # ==== Returns
+      # Descendant of ResponseError suitable for that exception code
+      # or ResponseError class if no class found
+      def self.exception(code)
+        S3::Error.const_get(code)
+      rescue NameError
+        ResponseError
+      end
+    end
+
+    #:stopdoc:
+
+    class AccessDenied < ResponseError; end
+    class AccountProblem < ResponseError; end
+    class AmbiguousGrantByEmailAddress < ResponseError; end
+    class BadDigest < ResponseError; end
+    class BucketAlreadyExists < ResponseError; end
+    class BucketAlreadyOwnedByYou < ResponseError; end
+    class BucketNotEmpty < ResponseError; end
+    class CredentialsNotSupported < ResponseError; end
+    class CrossLocationLoggingProhibited < ResponseError; end
+    class EntityTooSmall < ResponseError; end
+    class EntityTooLarge < ResponseError; end
+    class ExpiredToken < ResponseError; end
+    class IncompleteBody < ResponseError; end
+    class IncorrectNumberOfFilesInPostRequestPOST < ResponseError; end
+    class InlineDataTooLarge < ResponseError; end
+    class InternalError < ResponseError; end
+    class InvalidAccessKeyId < ResponseError; end
+    class InvalidAddressingHeader < ResponseError; end
+    class InvalidArgument < ResponseError; end
+    class InvalidBucketName < ResponseError; end
+    class InvalidDigest < ResponseError; end
+    class InvalidLocationConstraint < ResponseError; end
+    class InvalidPayer < ResponseError; end
+    class InvalidPolicyDocument < ResponseError; end
+    class InvalidRange < ResponseError; end
+    class InvalidSecurity < ResponseError; end
+    class InvalidSOAPRequest < ResponseError; end
+    class InvalidStorageClass < ResponseError; end
+    class InvalidTargetBucketForLogging < ResponseError; end
+    class InvalidToken < ResponseError; end
+    class InvalidURI < ResponseError; end
+    class KeyTooLong < ResponseError; end
+    class MalformedACLError < ResponseError; end
+    class MalformedACLError < ResponseError; end
+    class MalformedPOSTRequest < ResponseError; end
+    class MalformedXML < ResponseError; end
+    class MaxMessageLengthExceeded < ResponseError; end
+    class MaxPostPreDataLengthExceededErrorYour < ResponseError; end
+    class MetadataTooLarge < ResponseError; end
+    class MethodNotAllowed < ResponseError; end
+    class MissingAttachment < ResponseError; end
+    class MissingContentLength < ResponseError; end
+    class MissingRequestBodyError < ResponseError; end
+    class MissingSecurityElement < ResponseError; end
+    class MissingSecurityHeader < ResponseError; end
+    class NoLoggingStatusForKey < ResponseError; end
+    class NoSuchBucket < ResponseError; end
+    class NoSuchKey < ResponseError; end
+    class NotImplemented < ResponseError; end
+    class NotSignedUp < ResponseError; end
+    class OperationAborted < ResponseError; end
+    class PermanentRedirect < ResponseError; end
+    class PreconditionFailed < ResponseError; end
+    class Redirect < ResponseError; end
+    class RequestIsNotMultiPartContent < ResponseError; end
+    class RequestTimeout < ResponseError; end
+    class RequestTimeTooSkewed < ResponseError; end
+    class RequestTorrentOfBucketError < ResponseError; end
+    class SignatureDoesNotMatch < ResponseError; end
+    class SlowDown < ResponseError; end
+    class TemporaryRedirect < ResponseError; end
+    class TokenRefreshRequired < ResponseError; end
+    class TooManyBuckets < ResponseError; end
+    class UnexpectedContent < ResponseError; end
+    class UnresolvableGrantByEmailAddress < ResponseError; end
+    class UserKeyMustBeSpecified < ResponseError; end
+  end
+end

data/lib/s3/object.rb

@@ -0,0 +1,251 @@
+module S3
+
+  # Class responsible for handling objects stored in S3 buckets
+  class Object
+    include Parser
+    extend Forwardable
+
+    attr_accessor :content_type, :content_disposition, :content_encoding, :cache_control
+    attr_reader :last_modified, :etag, :size, :bucket, :key, :acl, :storage_class
+    attr_writer :content
+
+    def_instance_delegators :bucket, :name, :service, :bucket_request, :vhost?, :host, :path_prefix
+    def_instance_delegators :service, :protocol, :port, :secret_access_key
+    private_class_method :new
+
+    # Compares the object with other object. Returns true if the key
+    # of the objects are the same, and both have the same buckets (see
+    # Bucket equality)
+    def ==(other)
+      other.equal?(self) || (other.instance_of?(self.class) && self.key == other.key && self.bucket == other.bucket)
+    end
+
+    # Returns full key of the object: e.g. <tt>bucket-name/object/key.ext</tt>
+    def full_key
+      [name, key].join("/")
+    end
+
+    # Assigns a new +key+ to the object, raises ArgumentError if given
+    # key is not valid key name
+    def key=(key)
+      raise ArgumentError.new("Invalid key name: #{key}") unless key_valid?(key)
+      @key ||= key
+    end
+
+    # Assigns a new ACL to the object. Please note that ACL is not
+    # retrieved from the server and set to "public-read" by default.
+    #
+    # ==== Example
+    #   object.acl = :public_read
+    def acl=(acl)
+      @acl = acl.to_s.gsub("_", "-") if acl
+    end
+
+    # Assigns a new storage class (RRS) to the object. Please note
+    # that the storage class is not retrieved from the server and set
+    # to "STANDARD" by default.
+    #
+    # ==== Example
+    #   object.storage_class = :reduced_redundancy
+    def storage_class=(storage_class)
+      @storage_class = storage_class.to_s.upcase if storage_class
+    end
+
+    # Retrieves the object from the server. Method is used to download
+    # object information only (content type, size and so on). It does
+    # NOT download the content of the object (use the #content method
+    # to do it).
+    def retrieve
+      object_headers
+      self
+    end
+
+    # Retrieves the object from the server, returns true if the object
+    # exists or false otherwise. Uses #retrieve method, but catches
+    # S3::Error::NoSuchKey exception and returns false when it happens
+    def exists?
+      retrieve
+      true
+    rescue Error::NoSuchKey
+      false
+    end
+
+    # Download the content of the object, and caches it. Pass true
+    # to clear the cache and download the object again.
+    def content(reload = false)
+      get_object if reload or @content.nil?
+      @content
+    end
+
+    # Saves the object, returns true if successfull.
+    def save
+      put_object
+      true
+    end
+
+    # Copies the file to another key and/or bucket.
+    #
+    # ==== Options
+    # * <tt>:key</tt> - New key to store object in
+    # * <tt>:bucket</tt> - New bucket to store object in (instance of
+    #   S3::Bucket)
+    # * <tt>:acl</tt> - ACL of the copied object (default:
+    #   "public-read")
+    # * <tt>:content_type</tt> - Content type of the copied object
+    #   (default: "application/octet-stream")
+    def copy(options = {})
+      copy_object(options)
+    end
+
+    # Destroys the file on the server
+    def destroy
+      delete_object
+      true
+    end
+
+    # Returns Object's URL using protocol specified in service,
+    # e.g. <tt>http://domain.com.s3.amazonaws.com/key/with/path.extension</tt>
+    def url
+      URI.escape("#{protocol}#{host}/#{path_prefix}#{key}")
+    end
+
+    # Returns a temporary url to the object that expires on the
+    # timestamp given. Defaults to one hour expire time.
+    def temporary_url(expires_at = Time.now + 3600)
+      signature = Signature.generate_temporary_url_signature(:bucket => name,
+                                                             :resource => key,
+                                                             :expires_at => expires_at,
+                                                             :secret_access_key => secret_access_key)
+
+      "#{url}?AWSAccessKeyId=#{self.bucket.service.access_key_id}&Expires=#{expires_at.to_i.to_s}&Signature=#{signature}"
+    end
+
+    # Returns Object's CNAME URL (without <tt>s3.amazonaws.com</tt>
+    # suffix) using protocol specified in Service,
+    # e.g. <tt>http://domain.com/key/with/path.extension</tt>. (you
+    # have to set the CNAME in your DNS before using the CNAME URL
+    # schema).
+    def cname_url
+      URI.escape("#{protocol}#{name}/#{key}") if bucket.vhost?
+    end
+
+    def inspect #:nodoc:
+      "#<#{self.class}:/#{name}/#{key}>"
+    end
+
+    private
+
+    attr_writer :last_modified, :etag, :size, :original_key, :bucket
+
+    def copy_object(options = {})
+      key = options[:key] or raise ArgumentError, "No key given"
+      raise ArgumentError.new("Invalid key name: #{key}") unless key_valid?(key)
+      bucket = options[:bucket] || self.bucket
+
+      headers = {}
+
+      headers[:x_amz_acl] = options[:acl] || acl || "public-read"
+      headers[:content_type] = options[:content_type] || content_type || "application/octet-stream"
+      headers[:content_encoding] = options[:content_encoding] if options[:content_encoding]
+      headers[:content_disposition] = options[:content_disposition] if options[:content_disposition]
+      headers[:cache_control] = options[:cache_control] if options[:cache_control]
+      headers[:x_amz_copy_source] = full_key
+      headers[:x_amz_metadata_directive] = "REPLACE"
+      headers[:x_amz_copy_source_if_match] = options[:if_match] if options[:if_match]
+      headers[:x_amz_copy_source_if_none_match] = options[:if_none_match] if options[:if_none_match]
+      headers[:x_amz_copy_source_if_unmodified_since] = options[:if_modified_since] if options[:if_modified_since]
+      headers[:x_amz_copy_source_if_modified_since] = options[:if_unmodified_since] if options[:if_unmodified_since]
+
+      response = bucket.send(:bucket_request, :put, :path => key, :headers => headers)
+      object_attributes = parse_copy_object_result(response.body)
+
+      object = Object.send(:new, bucket, object_attributes.merge(:key => key, :size => size))
+      object.acl = response["x-amz-acl"]
+      object.content_type = response["content-type"]
+      object.content_encoding = response["content-encoding"]
+      object.content_disposition = response["content-disposition"]
+      object.cache_control = response["cache-control"]
+      object
+    end
+
+    def get_object(options = {})
+      response = object_request(:get, options)
+      parse_headers(response)
+    end
+
+    def object_headers(options = {})
+      response = object_request(:head, options)
+      parse_headers(response)
+    rescue Error::ResponseError => e
+      if e.response.code.to_i == 404
+        raise Error::ResponseError.exception("NoSuchKey").new("The specified key does not exist.", nil)
+      else
+        raise e
+      end
+    end
+
+    def put_object
+      response = object_request(:put, :body => content, :headers => dump_headers)
+      parse_headers(response)
+    end
+
+    def delete_object(options = {})
+      object_request(:delete)
+    end
+
+    def initialize(bucket, options = {})
+      self.bucket = bucket
+      self.key = options[:key]
+      self.last_modified = options[:last_modified]
+      self.etag = options[:etag]
+      self.size = options[:size]
+      self.cache_control = options[:cache_control]
+    end
+
+    def object_request(method, options = {})
+      bucket_request(method, options.merge(:path => key))
+    end
+
+    def last_modified=(last_modified)
+      @last_modified = Time.parse(last_modified) if last_modified
+    end
+
+    def etag=(etag)
+      @etag = etag[1..-2] if etag
+    end
+
+    def key_valid?(key)
+      if (key.nil? or key.empty? or key =~ %r#//#)
+        false
+      else
+        true
+      end
+    end
+
+    def dump_headers
+      headers = {}
+      headers[:x_amz_acl] = @acl || "public-read"
+      headers[:x_amz_storage_class] = @storage_class || "STANDARD"
+      headers[:content_type] = @content_type || "application/octet-stream"
+      headers[:content_encoding] = @content_encoding if @content_encoding
+      headers[:content_disposition] = @content_disposition if @content_disposition
+      headers[:cache_control] = @cache_control if @cache_control
+      headers
+    end
+
+    def parse_headers(response)
+      self.etag = response["etag"] if response.key?("etag")
+      self.content_type = response["content-type"] if response.key?("content-type")
+      self.content_disposition = response["content-disposition"] if response.key?("content-disposition")
+      self.cache_control = response["cache-control"] if response.key?("cache-control")
+      self.content_encoding = response["content-encoding"] if response.key?("content-encoding")
+      self.last_modified = response["last-modified"] if response.key?("last-modified")
+      if response.key?("content-range")
+        self.size = response["content-range"].sub(/[^\/]+\//, "").to_i
+      else
+        self.size = response["content-length"]
+        self.content = response.body
+      end
+    end
+  end
+end

data/lib/s3/objects_extension.rb

@@ -0,0 +1,37 @@
+module S3
+  module ObjectsExtension
+    # Builds the object in the bucket with given key
+    def build(key)
+      Object.send(:new, proxy_owner, :key => key)
+    end
+
+    # Finds first object with given name or raises the exception if
+    # not found
+    def find_first(name)
+      object = build(name)
+      object.retrieve
+    end
+    alias :find :find_first
+
+    # Finds the objects in the bucket.
+    #
+    # ==== Options
+    # * <tt>:prefix</tt> - Limits the response to keys which begin
+    #   with the indicated prefix
+    # * <tt>:marker</tt> - Indicates where in the bucket to begin
+    #   listing
+    # * <tt>:max_keys</tt> - The maximum number of keys you'd like
+    #   to see
+    # * <tt>:delimiter</tt> - Causes keys that contain the same
+    #   string between the prefix and the first occurrence of the
+    #   delimiter to be rolled up into a single result element
+    def find_all(options = {})
+      proxy_owner.send(:list_bucket, options)
+    end
+
+    # Destroys all keys in the bucket
+    def destroy_all
+      proxy_target.each { |object| object.destroy }
+    end
+  end
+end

data/lib/s3/parser.rb

@@ -0,0 +1,52 @@
+module S3
+  module Parser
+    include REXML
+
+    def rexml_document(xml)
+      xml.force_encoding(Encoding::UTF_8) if xml.respond_to? :force_encoding
+      Document.new(xml)
+    end
+
+    def parse_list_all_my_buckets_result(xml)
+      names = []
+      rexml_document(xml).elements.each("ListAllMyBucketsResult/Buckets/Bucket/Name") { |e| names << e.text }
+      names
+    end
+
+    def parse_location_constraint(xml)
+      rexml_document(xml).elements["LocationConstraint"].text
+    end
+
+    def parse_list_bucket_result(xml)
+      objects_attributes = []
+      rexml_document(xml).elements.each("ListBucketResult/Contents") do |e|
+        object_attributes = {}
+        object_attributes[:key] = e.elements["Key"].text
+        object_attributes[:etag] = e.elements["ETag"].text
+        object_attributes[:last_modified] = e.elements["LastModified"].text
+        object_attributes[:size] = e.elements["Size"].text
+        objects_attributes << object_attributes
+      end
+      objects_attributes
+    end
+
+    def parse_copy_object_result(xml)
+      object_attributes = {}
+      document = rexml_document(xml)
+      object_attributes[:etag] = document.elements["CopyObjectResult/ETag"].text
+      object_attributes[:last_modified] = document.elements["CopyObjectResult/LastModified"].text
+      object_attributes
+    end
+
+    def parse_error(xml)
+      document = rexml_document(xml)
+      code = document.elements["Error/Code"].text
+      message = document.elements["Error/Message"].text
+      [code, message]
+    end
+    
+    def parse_is_truncated xml
+      rexml_document(xml).elements["ListBucketResult/IsTruncated"].text =='true'
+    end
+  end
+end