Object
Builds a new CloudFiles::StorageObject in the current container. If force_exist is set, the object must exist or a CloudFiles::Exception::NoSuchObject Exception will be raised. If not, an "empty" CloudFiles::StorageObject will be returned, ready for data via CloudFiles::StorageObject.write
# File lib/cloudfiles/storage_object.rb, line 15 def initialize(container, objectname, force_exists = false, make_path = false) @container = container @containername = container.name @name = objectname @make_path = make_path @storagepath = "#{CloudFiles.escape @containername}/#{CloudFiles.escape @name, '/'}" if force_exists raise CloudFiles::Exception::NoSuchObject, "Object #{@name} does not exist" unless container.object_exists?(objectname) end end
Size of the object (in bytes)
# File lib/cloudfiles/storage_object.rb, line 58 def bytes self.object_metadata[:bytes] end
Content type of the object data
# File lib/cloudfiles/storage_object.rb, line 73 def content_type self.object_metadata[:content_type] end
# File lib/cloudfiles/storage_object.rb, line 77 def content_type=(type) self.copy(:headers => {'Content-Type' => type}) end
Copy this object to a new location (optionally in a new container)
You must supply either a name for the new object or a container name, or both. If a :name is supplied without a :container, the object is copied within the current container. If the :container is specified with no :name, then the object is copied to the new container with its current name.
object.copy(:name => "images/funny/lolcat.jpg", :container => "pictures")
You may also supply a hash of headers in the :headers option. From there, you can set things like Content-Type, or other headers as available in the API document.
object.copy(:name => 'newfile.tmp', :headers => {'Content-Type' => 'text/plain'})
Returns the new CloudFiles::StorageObject for the copied item.
# File lib/cloudfiles/storage_object.rb, line 352 def copy(options = {}) raise CloudFiles::Exception::Syntax, "You must provide the :container, :name, or :headers for this operation" unless (options[:container] || options[:name] || options[:headers]) new_container = options[:container] || self.container.name new_name = options[:name] || self.name new_headers = options[:headers] || {} raise CloudFiles::Exception::Syntax, "The :headers option must be a hash" unless new_headers.is_a?(Hash) new_name.sub!(/^\//,'') headers = {'X-Copy-From' => "#{self.container.name}/#{self.name}", 'Content-Type' => self.content_type.sub(/;.+/, '')}.merge(new_headers) # , 'Content-Type' => self.content_type new_path = "#{CloudFiles.escape new_container}/#{CloudFiles.escape new_name, '/'}" response = self.container.connection.storage_request("PUT", new_path, headers) code = response.code raise CloudFiles::Exception::InvalidResponse, "Invalid response code #{response.code}" unless (response.code =~ /^20/) return CloudFiles::Container.new(self.container.connection, new_container).object(new_name) end
Retrieves the data from an object and stores the data in memory. The data is returned as a string. Throws a NoSuchObjectException if the object doesn't exist.
If the optional size and range arguments are provided, the call will return the number of bytes provided by size, starting from the offset provided in offset.
object.data => "This is the text stored in the file"
# File lib/cloudfiles/storage_object.rb, line 89 def data(size = -1, offset = 0, headers = {}) if size.to_i > 0 range = sprintf("bytes=%d-%d", offset.to_i, (offset.to_i + size.to_i) - 1) headers['Range'] = range end response = self.container.connection.storage_request("GET", @storagepath, headers) raise CloudFiles::Exception::NoSuchObject, "Object #{@name} does not exist" unless (response.code =~ /^20/) response.body end
Retrieves the data from an object and returns a stream that must be passed to a block. Throws a NoSuchObjectException if the object doesn't exist.
If the optional size and range arguments are provided, the call will return the number of bytes provided by size, starting from the offset provided in offset.
data = "" object.data_stream do |chunk| data += chunk end data => "This is the text stored in the file"
# File lib/cloudfiles/storage_object.rb, line 113 def data_stream(size = -1, offset = 0, headers = {}, &block) if size.to_i > 0 range = sprintf("bytes=%d-%d", offset.to_i, (offset.to_i + size.to_i) - 1) headers['Range'] = range end self.container.connection.storage_request("GET", @storagepath, headers, nil) do |response| raise CloudFiles::Exception::NoSuchObject, "Object #{@name} does not exist. Response code #{response.code}" unless (response.code =~ /^20./) response.read_body(&block) end end
ETag of the object data
# File lib/cloudfiles/storage_object.rb, line 68 def etag self.object_metadata[:etag] end
Date of the object's last modification
# File lib/cloudfiles/storage_object.rb, line 63 def last_modified self.object_metadata[:last_modified] end
A convenience method to stream data into an object from a local file (or anything that can be loaded by Ruby's open method)
You can provide an optional hash of headers, in case you want to do something like set the Content-Type manually.
Throws an Errno::ENOENT if the file cannot be read.
object.data => "This is my data" object.load_from_filename("/tmp/file.txt") => true object.load_from_filename("/home/rackspace/myfile.tmp", 'Content-Type' => 'text/plain') object.data => "This data was in the file /tmp/file.txt" object.load_from_filename("/tmp/nonexistent.txt") => Errno::ENOENT: No such file or directory - /tmp/nonexistent.txt
# File lib/cloudfiles/storage_object.rb, line 268 def load_from_filename(filename, headers = {}, check_md5 = false) f = open(filename) if check_md5 require 'digest/md5' md5_hash = Digest::MD5.file(filename) headers["Etag"] = md5_hash.to_s() end self.write(f, headers) f.close true end
Returns the object's manifest.
object.manifest => "container/prefix"
# File lib/cloudfiles/storage_object.rb, line 155 def manifest self.object_metadata[:manifest] end
Returns the object's metadata as a nicely formatted hash, stripping off the X-Meta-Object- prefix that the system prepends to the key name.
object.metadata => {"ruby"=>"cool", "foo"=>"bar"}
# File lib/cloudfiles/storage_object.rb, line 129 def metadata metahash = {} self.object_metadata[:metadata].each{ |key, value| metahash[key.gsub(/x-object-meta-/, '').gsub(/\+\-/, ' ')] = URI.decode(value).gsub(/\+\-/, ' ') } metahash end
Takes the same options as the copy method, only it does a copy followed by a delete on the original object.
Returns the new CloudFiles::StorageObject for the moved item. You should not attempt to use the old object after doing a move.
# File lib/cloudfiles/storage_object.rb, line 372 def move(options = {}) new_object = self.copy(options) self.container.delete_object(self.name) self.freeze return new_object end
Retrieves Metadata for the object
# File lib/cloudfiles/storage_object.rb, line 35 def object_metadata @object_metadata ||= ( response = self.container.connection.storage_request("HEAD", @storagepath) raise CloudFiles::Exception::NoSuchObject, "Object #{@name} does not exist" unless (response.code =~ /^20/) resphash = {} metas = response.to_hash.select { |k,v| k.match(/^x-object-meta/) } metas.each do |x,y| resphash[x] = (y.respond_to?(:join) ? y.join('') : y.to_s) end { :manifest => response["x-object-manifest"], :bytes => response["content-length"], :last_modified => Time.parse(response["last-modified"]), :etag => response["etag"], :content_type => response["content-type"], :metadata => resphash } ) end
If the parent container is public (CDN-enabled), returns the SSL CDN URL to this object. Otherwise, return nil
public_object.public_ssl_url => "https://c61.ssl.cf0.rackcdn.com/myfile.jpg" private_object.public_ssl_url => nil
# File lib/cloudfiles/storage_object.rb, line 323 def public_ssl_url self.container.public? ? self.container.cdn_ssl_url + "/#{CloudFiles.escape @name, '/'}" : nil end
If the parent container is public (CDN-enabled), returns the SSL CDN URL to this object. Otherwise, return nil
public_object.public_streaming_url => "https://c61.stream.rackcdn.com/myfile.jpg" private_object.public_streaming_url => nil
# File lib/cloudfiles/storage_object.rb, line 334 def public_streaming_url self.container.public? ? self.container.cdn_streaming_url + "/#{CloudFiles.escape @name, '/'}" : nil end
If the parent container is public (CDN-enabled), returns the CDN URL to this object. Otherwise, return nil
public_object.public_url => "http://c0001234.cdn.cloudfiles.rackspacecloud.com/myfile.jpg" private_object.public_url => nil
# File lib/cloudfiles/storage_object.rb, line 312 def public_url self.container.public? ? self.container.cdn_url + "/#{CloudFiles.escape @name, '/'}" : nil end
Purges CDN Edge Cache for all objects inside of this container
:email, An valid email address or comma seperated list of emails to be notified once purge is complete .
obj.purge_from_cdn => true or obj.purge_from_cdn("User@domain.com") => true or obj.purge_from_cdn("User@domain.com, User2@domain.com") => true
# File lib/cloudfiles/storage_object.rb, line 240 def purge_from_cdn(email=nil) raise Exception::CDNNotAvailable unless cdn_available? headers = {} headers = {"X-Purge-Email" => email} if email response = self.container.connection.cdn_request("DELETE", @storagepath, headers) raise CloudFiles::Exception::Connection, "Error Unable to Purge Object: #{@name}" unless (response.code.to_s =~ /^20.$/) true end
Refreshes the object metadata
# File lib/cloudfiles/storage_object.rb, line 28 def refresh @object_metadata = nil true end
A convenience method to stream data from an object into a local file
Throws an Errno::ENOENT if the file cannot be opened for writing due to a path error, and Errno::EACCES if the file cannot be opened for writing due to permissions.
object.data => "This is my data" object.save_to_filename("/tmp/file.txt") => true $ cat /tmp/file.txt "This is my data" object.save_to_filename("/tmp/owned_by_root.txt") => Errno::EACCES: Permission denied - /tmp/owned_by_root.txt
# File lib/cloudfiles/storage_object.rb, line 296 def save_to_filename(filename) File.open(filename, 'wb+') do |f| self.data_stream do |chunk| f.write chunk end end true end
Sets the manifest for an object. By passing a string as an argument, you can set the manifest for an object. However, setting manifest will overwrite any existing manifest for the object.
Throws NoSuchObjectException if the object doesn't exist. Throws InvalidResponseException if the request fails.
# File lib/cloudfiles/storage_object.rb, line 165 def set_manifest(manifest) headers = {'X-Object-Manifest' => manifest} response = self.container.connection.storage_request("PUT", @storagepath, headers) raise CloudFiles::Exception::NoSuchObject, "Object #{@name} does not exist" if (response.code == "404") raise CloudFiles::Exception::InvalidResponse, "Invalid response code #{response.code}" unless (response.code =~ /^20/) true end
Sets the metadata for an object. By passing a hash as an argument, you can set the metadata for an object. However, setting metadata will overwrite any existing metadata for the object.
Throws NoSuchObjectException if the object doesn't exist. Throws InvalidResponseException if the request fails.
# File lib/cloudfiles/storage_object.rb, line 140 def set_metadata(metadatahash) headers = {} metadatahash.each{ |key, value| headers['X-Object-Meta-' + key.to_s.capitalize] = value.to_s } response = self.container.connection.storage_request("POST", @storagepath, headers) raise CloudFiles::Exception::NoSuchObject, "Object #{@name} does not exist" if (response.code == "404") raise CloudFiles::Exception::InvalidResponse, "Invalid response code #{response.code}" unless (response.code =~ /^20/) true end
Takes supplied data and writes it to the object, saving it. You can supply an optional hash of headers, including Content-Type and ETag, that will be applied to the object.
If you would rather stream the data in chunks, instead of reading it all into memory at once, you can pass an IO object for the data, such as: object.write(open('/path/to/file.mp3'))
You can compute your own MD5 sum and send it in the "ETag" header. If you provide yours, it will be compared to the MD5 sum on the server side. If they do not match, the server will return a 422 status code and a CloudFiles::Exception::MisMatchedChecksum Exception will be raised. If you do not provide an MD5 sum as the ETag, one will be computed on the server side.
Updates the container cache and returns true on success, raises exceptions if stuff breaks.
object = container.create_object("newfile.txt") object.write("This is new data") => true object.data => "This is new data"
If you are passing your data in via STDIN, just do an
object.write
with no data (or, if you need to pass headers)
object.write(nil,{'header' => 'value})
# File lib/cloudfiles/storage_object.rb, line 202 def write(data = nil, headers = {}) raise CloudFiles::Exception::Syntax, "No data or header updates supplied" if ((data.nil? && $stdin.tty?) and headers.empty?) if headers['Content-Type'].nil? type = MIME::Types.type_for(self.name).first.to_s if type.empty? headers['Content-Type'] = "application/octet-stream" else headers['Content-Type'] = type end end # If we're taking data from standard input, send that IO object to cfreq data = $stdin if (data.nil? && $stdin.tty? == false) response = self.container.connection.storage_request("PUT", @storagepath, headers, data) code = response.code raise CloudFiles::Exception::InvalidResponse, "Invalid content-length header sent" if (code == "412") raise CloudFiles::Exception::MisMatchedChecksum, "Mismatched etag" if (code == "422") raise CloudFiles::Exception::InvalidResponse, "Invalid response code #{code}" unless (code =~ /^20./) make_path(File.dirname(self.name)) if @make_path == true self.refresh true end
Generated with the Darkfish Rdoc Generator 2.