require 'azure-signature' require 'active_support/core_ext/hash/conversions' module Azure module Armrest class StorageAccount < BaseModel attr_from_hash :name => :name, :blob_endpoint => [:properties, :primaryEndpoints, :blob] # Classes used to wrap container and blob information. class Container < BaseModel attr_from_hash :name => :Name end class ContainerProperty < BaseModel; end class Blob < BaseModel attr_from_hash :name => :Name, :lease_state => [:Properties, :LeaseState] end class BlobProperty < BaseModel; end class PrivateImage < BlobProperty; end class BlobServiceProperty < BaseModel; end class BlobServiceStat < BaseModel; end class BlobMetadata < BaseModel; end class BlobSnapshot < Blob; end # Classes used to wrap table information class Table < BaseModel; end class TableData < BaseModel; end # Classes used to wrap file shares class ShareDirectory < BaseModel; end class ShareFile < BaseModel; end # The version string used in headers sent as part any internal http # request. The default is 2016-05-31. attr_accessor :storage_api_version # The default access key used when creating a signature for internal http requests. attr_accessor :access_key # The parent configuration object attr_accessor :configuration def initialize(json, skip_accessors_definition = false) super @storage_api_version = '2016-05-31' end # Returns a list of tables for the given storage account +key+. Note # that full metadata is returned. # def tables(key = access_key) raise ArgumentError, "No access key specified" unless key response = table_response(key, nil, "Tables") JSON.parse(response.body)['value'].map { |t| Table.new(t) } end # Return information about a single table for the given storage # account +key+. If you are looking for the entities within the # table, use the table_data method instead. # def table_info(table, key = access_key) raise ArgumentError, "No access key specified" unless key response = table_response(key, nil, "Tables('#{table}')") Table.new(response.body) end # Returns a list of TableData objects for the given table +name+ using # account +key+. The exact nature of the TableData object depends on the # type of table that it is. # # You may specify :filter, :select or :top as options to restrict your # result set. # # By default you will receive a maximum of 1000 records. If you wish to # receive more records, you will need to use the continuation token. You # may also set the :all option to true if you want all records, though we # recommend using a filter as well if you use that option as there can # be thousands of results. # # You may also specify a :NextRowKey, :NextPartitionKey or :NextTableset # explicitly for paging. Normally you would just pass the # collection's continuation_token, however. See below for an example. # # When using continuation tokens, you should retain your original # filtering as well, or you may get unexpected results. # # Examples: # # # Get the first 10 rows of data from the last 3 days # date = (Time.now - (86400 * 3)).iso8601 # my_filter = "timestamp ge datetime'#{date}'" # options = {:top => 10, :filter => my_filter} # # results = storage_account.table_data(table, key, options) # # # Now get the next 10 records # if results.continuation_token # options[:continuation_token] = results.continuation_token # more_results = storage_account.table_data(table, key, options) # end # def table_data(name, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = build_query(options) response = table_response(key, query, name) klass = Azure::Armrest::StorageAccount::TableData data = Azure::Armrest::ArmrestCollection.create_from_response(response, klass) # Continuation tokens are parsed differently for storage data.continuation_token = parse_continuation_tokens(response) if options[:all] && data.continuation_token options[:continuation_token] = data.continuation_token data.push(*table_data(name, key, options)) data.continuation_token = nil # Clear when finished end data end ### Files and Directories # Create a new directory under the specified +share+ or parent directory. # # The only supported option at this time is a "timeout" option. # def create_directory(share, directory, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = {:restype => 'directory'}.merge(options).to_query response = file_response(key, query, 'put', '', File.join(share, directory)) Azure::Armrest::ResponseHeaders.new(response.headers).tap do |rh| rh.response_code = response.code end end # Delete the specified +share+ or parent directory. # # The only supported option at this time is a "timeout" option. # def delete_directory(share, directory, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = {:restype => 'directory'}.merge(options).to_query response = file_response(key, query, 'delete', '', File.join(share, directory)) Azure::Armrest::ResponseHeaders.new(response.headers).tap do |rh| rh.response_code = response.code end end # Get properties for the specified +share+ or parent directory. # # The only supported option at this time is a "timeout" option. # def directory_properties(share, directory, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = {:restype => 'directory'}.merge(options).to_query response = file_response(key, query, 'get', '', File.join(share, directory)) ShareDirectory.new(response.headers) end # Get metadata for the specified +share+ or parent directory. # # The only supported option at this time is a "timeout" option. # def directory_metadata(share, directory, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = {:restype => 'directory', :comp => 'metadata'}.merge(options).to_query response = file_response(key, query, 'head', '', File.join(share, directory)) ShareDirectory.new(response.headers) end # Returns a list of files for the specified file-share. You may also # optionally specify a +directory+ in "share/directory" format. # # You may specify multiple +options+ to limit the result set. The # possible options are: # # * prefix # * marker # * maxresults # * timeout # * all # # If the :all option is set to true, then this method will collect all # records. Otherwise, it is capped at 5000 records by Azure, or whatever # you set :maxresults to. If you set both the :all option and the :maxresults # option, then all records will be collected in :maxresults batches. # def files(share, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = "comp=list&restype=directory" skip_defs = options[:skip_accessors_definition] options.each do |okey, ovalue| unless okey == :skip_accessors_definition query += "&#{okey}=#{[ovalue].flatten.join(',')}" end end response = file_response(key, query, 'get', nil, share) hash = Hash.from_xml(response.body)['EnumerationResults']['Entries'] results = [] if hash && hash['Directory'] Array.wrap(hash['Directory']).each { |dir| results << ShareDirectory.new(dir, skip_defs) } end if hash && hash['File'] Array.wrap(hash['File']).each { |file| results << ShareFile.new(file, skip_defs) } end if options[:all] && hash['NextMarker'] options[:marker] = hash['NextMarker'] results.concat(files(share, key, options)) end results end # Returns the raw contents of the specified file. # # The only supported option at this time is a "timeout" option. # def file_content(share, file, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = options.to_query response = file_response(key, query, 'get', '', File.join(share, file)) response.body end # Returns the raw contents of the specified file. # # The only supported option at this time is a "timeout" option. # def file_properties(share, file, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = options.to_query response = file_response(key, query, 'head', '', File.join(share, file)) Azure::Armrest::ResponseHeaders.new(response.headers).tap do |rh| rh.response_code = response.code end end # Create the specified share file. You may specify any of the following # options: # # * cache_control # * content_disposition # * content_length (default: 0) # * content_encoding # * content_language # * content_md5 # * content_type (default: application/octet-stream) # * meta_name # * timeout # * version # # Note that this does not set the content of the file, it only creates # in the file share. # def create_file(share, file, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key timeout = options.delete(:timeout) # Part of request url = File.join(properties.primary_endpoints.file, share, file) url += "?timeout=#{timeout}" if timeout hash = options.transform_keys.each { |okey| 'x-ms-' + okey.to_s.tr('_', '-') } hash['verb'] = 'PUT' # Mandatory and/or sane defaults hash['x-ms-type'] = 'file' hash['x-ms-content-length'] ||= 0 hash['x-ms-content-type'] ||= 'application/octet-stream' headers = build_headers(url, key, :file, hash) response = ArmrestService.send( :rest_put, :url => url, :payload => '', :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) Azure::Armrest::ResponseHeaders.new(response.headers).tap do |rh| rh.response_code = response.code end end # Delete the specified share file. # # The only supported option at this time is a "timeout" option. # def delete_file(share, file, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = options.to_query response = file_response(key, query, 'delete', '', File.join(share, file)) Azure::Armrest::ResponseHeaders.new(response.headers).tap do |rh| rh.response_code = response.code end end # Copy a +src_file+ to a destination +dst_file+ within the same storage account. # def copy_file(src_container, src_file, dst_container = src_container, dst_file = nil, key = access_key) raise ArgumentError, "No access key specified" unless key dst_file ||= File.basename(src_blob) dst_url = File.join(properties.primary_endpoints.file, dst_container, dst_file) src_url = File.join(properties.primary_endpoints.file, src_container, src_file) options = {'x-ms-copy-source' => src_url, :verb => 'PUT'} headers = build_headers(dst_url, key, :file, options) response = ArmrestService.send( :rest_put, :url => dst_url, :payload => '', :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) Azure::Armrest::ResponseHeaders.new(response.headers).tap do |rh| rh.response_code = response.code end end # Add content to +file+ on +share+. The +options+ hash supports # three options, :content, :timeout and :write. # # The :content option is just a string, i.e. the content you want # to add to the file. Azure allows you to add a maximum of 4mb worth # of content per request. # # The :timeout option is nil by default. The :write option defaults to # 'update'. If you want to clear a file, set it to 'clear'. # def add_file_content(share, file, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key timeout = options.delete(:timeout) content = options.delete(:content) url = File.join(properties.primary_endpoints.file, share, file) + "?comp=range" url += "&timeout=#{timeout}" if timeout hash = options.transform_keys.each { |okey| 'x-ms-' + okey.to_s.tr('_', '-') } hash['verb'] = 'PUT' hash['x-ms-write'] ||= 'update' if hash['x-ms-write'] == 'clear' hash['content-length'] = 0 hash['x-ms-range'] = "bytes=0-" else range = 0..(content.size - 1) hash['content-length'] = content.size hash['x-ms-range'] = "bytes=#{range.min}-#{range.max}" end headers = build_headers(url, key, :file, hash) response = ArmrestService.send( :rest_put, :url => url, :payload => content, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) Azure::Armrest::ResponseHeaders.new(response.headers).tap do |rh| rh.response_code = response.code end end ### Containers # Return a list of container names for the given storage account +key+. # If no key is provided, it is assumed that the StorageAccount object # includes the access_key property. # # # The following options are supported: # # * prefix # * delimiter # * maxresults # * include # * timeout # * all # # By default Azure uses a value of 5000 for :maxresults. # # If the :include option is specified, it should contain an array of # one element: metadata. More options may be added by Microsoft # at a later date. # # If the :all option is set to true, then this method will collect all # records. Otherwise, it is capped at 5000 records by Azure, or whatever # you set :maxresults to. If you set both the :all option and the :maxresults # option, then all records will be collected in :maxresults batches. # # Example: # # sas = Azure::Armrest::StorageAccountService.new(conf) # key = sas.list_account_keys['key1'] # acct = sas.get('your_storage_account', 'your_resource_group') # # p acct.containers(key) # p acct.containers(key, :include => ['metadata']) # p acct.containers(key, :maxresults => 1) # def containers(key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = "comp=list" skip_defs = options[:skip_accessors_definition] options.each do |okey, ovalue| unless okey == :skip_accessors_definition query += "&#{okey}=#{[ovalue].flatten.join(',')}" end end response = blob_response(key, query) hash = Hash.from_xml(response.body)['EnumerationResults']['Containers'] results = [] if hash && hash['Container'] Array.wrap(hash['Container']).each { |c| results << Container.new(c, skip_defs) } end if options[:all] && hash['NextMarker'] options[:marker] = hash['NextMarker'] results.concat(containers(key, options)) end results end # Returns the properties for the given container +name+ using account +key+. # def container_properties(name, key = access_key) raise ArgumentError, "No access key specified" unless key response = blob_response(key, "restype=container", name) ContainerProperty.new(response.headers) end # Returns the properties for the given container +name+ using account +key+. # # If the returned object does not contain x_ms_blob_public_access then # the container is private to the account owner. You can also use the # :private? method to determine if the account is public or private. # def container_acl(name, key = access_key) raise ArgumentError, "No access key specified" unless key response = blob_response(key, "restype=container&comp=acl", name) response.headers[:private?] = response.headers.include?(:x_ms_blob_public_access) ? false : true ContainerProperty.new(response.headers) end # Return the blob properties for the given +blob+ found in +container+. You may # optionally provide a date to get information for a snapshot. # def blob_properties(container, blob, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key url = File.join(blob_endpoint_from_hash, container, blob) url += "?snapshot=" + options[:date] if options[:date] headers = build_headers(url, key, :blob, :verb => 'HEAD') response = ArmrestService.send( :rest_head, :url => url, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) BlobProperty.new(response.headers.merge(:container => container, :name => blob), options[:skip_accessors_definition]) end # Update the given +blob+ in +container+ with the provided options. The # possible options are: # # cache_control # content_disposition # content_encoding # content_language # content_length # content_md5 # content_type # lease_id # version # # The content_length option is only value for page blobs, and is used # to resize the blob. # def update_blob_properties(container, blob, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key url = File.join(properties.primary_endpoints.blob, container, blob) + "?comp=properties" hash = options.transform_keys { |okey| "x-ms-blob-" + okey.to_s.tr('_', '-') } hash['verb'] = 'PUT' headers = build_headers(url, key, :blob, hash) response = ArmrestService.send( :rest_put, :url => url, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) BlobProperty.new(response.headers.merge(:container => container, :name => blob)) end # Return a list of blobs for the given +container+ using the given +key+ # or the access_key property of the StorageAccount object. # # The following options are supported: # # * prefix # * delimiter # * maxresults # * include # * timeout # # By default Azure uses a value of 5000 for :maxresults. # # If the :include option is specified, it should contain an array of # one or more of the following values: snapshots, metadata, copy or # uncommittedblobs. # # If the :all option is set to true, then this method will collect all # records. Otherwise, it is capped at 5000 records by Azure, or whatever # you set :maxresults to. If you set both the :all option and the :maxresults # option, then all records will be collected in :maxresults batches. # # Example: # # sas = Azure::Armrest::StorageAccountService.new(conf) # key = sas.list_account_keys['key1'] # acct = sas.get('your_storage_account', 'your_resource_group') # # p acct.blobs('vhds', key) # p acct.blobs('vhds', key, :timeout => 30) # p acct.blobs('vhds', key, :include => ['snapshots', 'metadata']) # def blobs(container, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = "restype=container&comp=list" skip_defs = options[:skip_accessors_definition] options.each do |okey, ovalue| unless okey == :skip_accessors_definition query += "&#{okey}=#{[ovalue].flatten.join(',')}" end end response = blob_response(key, query, container) hash = Hash.from_xml(response.body)['EnumerationResults']['Blobs'] results = [] if hash && hash['Blob'] Array.wrap(hash['Blob']).each do |h| h[:container] = container object = h.key?('Snapshot') ? BlobSnapshot.new(h, skip_defs) : Blob.new(h, skip_defs) results << object end end if options[:all] && hash['NextMarker'] options[:marker] = hash['NextMarker'] results.concat(blobs(container, key, options)) end results end # Returns an array of all blobs for all containers. The +options+ hash # may contain the same arguments that a call to StorageAccount#blobs # would accept. # def all_blobs(key = access_key, max_threads = 10, options = {}) raise ArgumentError, "No access key specified" unless key array = [] mutex = Mutex.new opts = { :skip_accessors_definition => options[:skip_accessors_definition] } Parallel.each(containers(key, opts), :in_threads => max_threads) do |container| begin mutex.synchronize { array.concat(blobs(container.name_from_hash, key, options)) } rescue Errno::ECONNREFUSED, Azure::Armrest::TimeoutException => err msg = "Unable to gather blob information for #{container.name}: #{err}" Azure::Armrest::Configuration.log.try(:log, Logger::WARN, msg) next end end array end # Returns the blob service properties for the current storage account. # def blob_service_properties(key = access_key) raise ArgumentError, "No access key specified" unless key response = blob_response(key, "restype=service&comp=properties") hash = Hash.from_xml(response.body)['StorageServiceProperties'] BlobServiceProperty.new(hash) end # Return metadata for the given +blob+ within +container+. You may # specify a +date+ to retrieve metadata for a specific snapshot. # def blob_metadata(container, blob, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key query = "comp=metadata" query << "&snapshot=" + options[:date] if options[:date] response = blob_response(key, query, container, blob) BlobMetadata.new(response.headers) end # Retrieves statistics related to replication for the Blob service. Only # available on the secondary location endpoint when read-access # geo-redundant replication is enabled for the storage account. # def blob_service_stats(key = access_key) raise ArgumentError, "No access key specified" unless key response = blob_response(key, "restype=service&comp=stats") hash = Hash.from_xml(response.body)['StorageServiceStats'] BlobServiceStat.new(hash) end # Copy the blob from the source container/blob to the destination container/blob. # If no destination blob name is provided, it will use the same name as the source. # # Example: # # source = "Microsoft.Compute/Images/your_container/your-img-osDisk.123xyz.vhd" # storage_acct.copy_blob('system', source, 'vhds', nil, your_key) # def copy_blob(src_container, src_blob, dst_container, dst_blob = nil, key = access_key) raise ArgumentError, "No access key specified" unless key dst_blob ||= File.basename(src_blob) dst_url = File.join(properties.primary_endpoints.blob, dst_container, dst_blob) src_url = File.join(properties.primary_endpoints.blob, src_container, src_blob) options = {'x-ms-copy-source' => src_url, 'if-none-match' => '*', :verb => 'PUT'} headers = build_headers(dst_url, key, :blob, options) response = ArmrestService.send( :rest_put, :url => dst_url, :payload => '', :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) blob = blob_properties(dst_container, dst_blob, key) blob.response_headers = Azure::Armrest::ResponseHeaders.new(response.headers) blob.response_code = response.code blob end # Delete the given +blob+ found in +container+. Pass a :date option # if you wish to delete a snapshot. # def delete_blob(container, blob, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key url = File.join(properties.primary_endpoints.blob, container, blob) url += "?snapshot=" + options[:date] if options[:date] headers = build_headers(url, key, :blob, :verb => 'DELETE') response = ArmrestService.send( :rest_delete, :url => url, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) headers = Azure::Armrest::ResponseHeaders.new(response.headers) headers.response_code = response.code headers end # Create new blob for a container. # # The options parameter is a hash that contains information used # when creating the blob: # # * type - "BlockBlob", "PageBlob" or "AppendBlob". Mandatory. # # * content_disposition # * content_encoding # * content_language # * content_md5 # * content_type # * cache_control # * lease_id # * payload (block blobs only) # * sequence_number (page blobs only) # * timeout (part of the request) # # Returns a ResponseHeaders object since this method is asynchronous. # def create_blob(container, blob, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key timeout = options.delete(:timeout) payload = options.delete(:payload) || '' url = File.join(properties.primary_endpoints.blob, container, blob) url += "&timeout=#{timeout}" if timeout hash = options.transform_keys do |okey| if okey.to_s =~ /^if/i okey.to_s.tr('_', '-') elsif %w[date meta_name lease_id version].include?(okey.to_s) 'x-ms-' + okey.to_s.tr('_', '-') else 'x-ms-blob-' + okey.to_s.tr('_', '-') end end unless hash['x-ms-blob-type'] raise ArgumentError, "The :type option must be specified" end hash['x-ms-date'] ||= Time.now.httpdate hash['x-ms-version'] ||= storage_api_version hash['verb'] = 'PUT' # Content length must be 0 (blank) for Page or Append blobs if %w[pageblob appendblob].include?(hash['x-ms-blob-type'].downcase) hash['content-length'] = '' else hash['content-length'] ||= hash['x-ms-blob-content-length'] end # Override the default empty string hash['content-type'] ||= hash['x-ms-blob-content-type'] || 'application/octet-stream' headers = build_headers(url, key, :blob, hash) response = ArmrestService.send( :rest_put, :url => url, :payload => payload, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) resp_headers = Azure::Armrest::ResponseHeaders.new(response.headers) resp_headers.response_code = response.code resp_headers end # Create a read-only snapshot of a blob. # # Possible options are: # # * meta_name # * lease_id # * client_request_id # * if_modified_since # * if_unmodified_since # * if_match # * if_none_match # * timeout # # Returns a ResponseHeaders object since this is an asynchronous method. # def create_blob_snapshot(container, blob, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key timeout = options.delete(:timeout) # Part of request url = File.join(properties.primary_endpoints.blob, container, blob) + "?comp=snapshot" url += "&timeout=#{timeout}" if timeout hash = options.transform_keys do |okey| if okey.to_s =~ /^if/i okey.to_s.tr('_', '-') else 'x-ms-blob-' + okey.to_s.tr('_', '-') end end hash['verb'] = 'PUT' headers = build_headers(url, key, :blob, hash) response = ArmrestService.send( :rest_put, :url => url, :payload => '', :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) headers = Azure::Armrest::ResponseHeaders.new(response.headers) headers.response_code = response.code headers end # Get the contents of the given +blob+ found in +container+ using the # given +options+. This is a low level method to read a range of bytes # from the blob directly. The possible options are: # # * range - A range of bytes to collect. # * start_byte - The starting byte for collection. # * end_byte - The end byte for collection. Use this or :length with :start_byte. # * length - The number of bytes to collect starting at +start_byte+. # * entire_image - Read all bytes for the blob. # * md5 - If true, the response headers will include MD5 checksum information. # * date - Get the blob snapshot for the given date. # # If you do not specify a :range or :start_byte, then an error will be # raised unless you explicitly set the :entire_image option to true. # However, that is not recommended because the blobs can be huge. # # Unlike other methods, this method returns a raw response object rather # than a wrapper model. Get the information you need using: # # * response.body - blob data. # * response.headers - blob metadata. # # Example: # # ret = @storage_acct.get_blob(@container, @blob, key, :start_byte => start_byte, :length => length) # content_md5 = ret.headers[:content_md5].unpack("m0").first.unpack("H*").first # returned_md5 = Digest::MD5.hexdigest(ret.body) # raise "Checksum error: #{range_str}, blob: #{@container}/#{@blob}" unless content_md5 == returned_md5 # return ret.body # def get_blob_raw(container, blob, key = access_key, options = {}) raise ArgumentError, "No access key specified" unless key url = File.join(properties.primary_endpoints.blob, container, blob) url += "?snapshot=" + options[:date] if options[:date] additional_headers = { 'verb' => 'GET' } range_str = nil if options[:range] range_str = "bytes=#{options[:range].min}-#{options[:range].max}" elsif options[:start_byte] range_str = "bytes=#{options[:start_byte]}-" if options[:end_byte] range_str << options[:end_byte].to_s elsif options[:length] range_str << (options[:start_byte] + options[:length] - 1).to_s end end if range_str additional_headers['x-ms-range'] = range_str additional_headers['x-ms-range-get-content-md5'] = true if options[:md5] else raise ArgumentError, "must specify byte range or entire_image flag" unless options[:entire_image] end headers = build_headers(url, key, :blob, additional_headers) ArmrestService.send( :rest_get, :url => url, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) end private # Build a query string from a hash of options. # def build_query(options) array = [] options.each do |key, value| next if key == :all if [:filter, :select, :top].include?(key) array << "$#{key}=#{value}" if value elsif key == :continuation_token value.each { |k, token| array << "#{k}=#{token}" if token } else array << "#{key}=#{value}" if value end end array.join('&') end # Get the continuation tokens from the response headers for paging results. # def parse_continuation_tokens(response) headers = response.headers token = { :NextPartitionKey => headers[:x_ms_continuation_nextpartitionkey], :NextRowKey => headers[:x_ms_continuation_nextrowkey], :NextTableName => headers[:x_ms_continuation_nexttablename] } # If there are no continuation values at all, then return nil token = nil if token.all? { |_key, value| value.nil? } token end # Using the blob primary endpoint as a base, join any arguments to the # the url and submit an http request. # def blob_response(key, query, *args) url = File.join(blob_endpoint_from_hash, *args) + "?#{query}" headers = build_headers(url, key, 'blob') ArmrestService.send( :rest_get, :url => url, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) end # Using the file primary endpoint as a base, join any arguments to the # the url and submit an http request. # def file_response(key, query, request_type = 'get', payload = '', *args) url = File.join(properties.primary_endpoints.file, *args) url += "?#{query}" if query && !query.empty? request_method = "rest_#{request_type}".to_sym headers = build_headers(url, key, :file, :verb => request_type.to_s.upcase) params = { :url => url, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify } if %w[put post].include?(request_type.to_s.downcase) params[:payload] = payload end ArmrestService.send(request_method, params) end # Using the blob primary endpoint as a base, join any arguments to the # the url and submit an http request. def table_response(key, query = nil, *args) url = File.join(properties.primary_endpoints.table, *args) headers = build_headers(url, key, 'table') headers['Accept'] = 'application/json;odata=fullmetadata' # Must happen after headers are built unless query.nil? || query.empty? url << "?#{query}" end ArmrestService.send( :rest_get, :url => url, :headers => headers, :proxy => configuration.proxy, :ssl_version => configuration.ssl_version, :ssl_verify => configuration.ssl_verify ) end # Set the headers needed, including the Authorization header. # def build_headers(url, key, sig_type = nil, additional_headers = {}) sig = Signature.new(url, key) sig_type ||= 'blob' # RestClient will set the Content-Type to application/x-www-form-urlencoded. # We must override this setting or the request will fail in some cases. content_type = additional_headers['content-type'] || '' headers = { 'content-type' => content_type, 'x-ms-date' => Time.now.httpdate, 'x-ms-version' => storage_api_version, 'auth_string' => true } headers.merge!(additional_headers) headers['authorization'] = sig.signature(sig_type, headers) headers end end end end