moserp-s3sync 1.2.6

data/lib/S3_s3sync_mod.rb

@@ -0,0 +1,143 @@
+#  This software code is made available "AS IS" without warranties of any
+#  kind.  You may copy, display, modify and redistribute the software
+#  code either by itself or as incorporated into your code; provided that
+#  you do not remove any proprietary notices.  Your use of this software
+#  code is at your own risk and you waive any claim against Amazon
+#  Digital Services, Inc. or its affiliates with respect to your use of
+#  this software code. (c) 2006 Amazon Digital Services, Inc. or its
+#  affiliates.
+#  
+# This software code is made available "AS IS" without warranties of any        
+# kind.  You may copy, display, modify and redistribute the software            
+# code either by itself or as incorporated into your code; provided that        
+# you do not remove any proprietary notices.  Your use of this software         
+# code is at your own risk and you waive any claim against the author
+# with respect to your use of this software code. 
+# (c) 2007 s3sync.net
+#
+require 'S3'
+require 'HTTPStreaming'
+
+# The purpose of this file is to overlay the S3 library from AWS
+# to add some functionality
+# (without changing the file itself or requiring a specific version)
+# It still isn't perfectly robust, i.e. if radical changes are made
+# to the underlying lib this stuff will need updating.
+
+module S3
+	class AWSAuthConnection
+	
+      def make_http(bucket='', host='', proxy_host=nil, proxy_port=nil, proxy_user=nil, proxy_pass=nil)
+
+         # build the domain based on the calling format
+         server = ''
+         if host != ''
+           server = host           
+         elsif bucket.empty?
+           # for a bucketless request (i.e. list all buckets)
+           # revert to regular domain case since this operation
+           # does not make sense for vanity domains
+           server = @server
+         elsif @calling_format == CallingFormat::SUBDOMAIN
+           server = "#{bucket}.#{@server}" 
+         elsif @calling_format == CallingFormat::VANITY
+           server = bucket 
+         else
+           server = @server
+         end
+         # automatically does the right thing when no proxy
+         http = Net::HTTP::Proxy(proxy_host, proxy_port, proxy_user, proxy_pass).new(server, @port)
+         #http = Net::HTTP.new(server, @port)
+         http.use_ssl = @is_secure
+         http.verify_mode=@verify_mode
+         http.ca_file=@ca_file
+         http.ca_path=@ca_path
+         http.start
+         return http
+      end
+
+		# add support for streaming the response body to an IO stream
+		alias __make_request__ make_request
+      def make_request(method, bucket='', key='', path_args={}, headers={}, data='', metadata={}, streamOut=nil)
+         # build the path based on the calling format
+         path = ''
+         if (not bucket.empty?) and (@calling_format == CallingFormat::REGULAR)
+           path << "/#{bucket}"
+         end
+         # add the slash after the bucket regardless
+         # the key will be appended if it is non-empty
+         path << "/#{key}"
+   
+         # build the path_argument string
+         # add the ? in all cases since 
+         # signature and credentials follow path args
+         path << '?'
+         path << S3.path_args_hash_to_string(path_args) 
+         
+         req = method_to_request_class(method).new("#{path}")
+   
+         set_headers(req, headers)
+         set_headers(req, metadata, METADATA_PREFIX)
+         set_headers(req, {'Connection' => 'keep-alive', 'Keep-Alive' => '300'})
+         
+         set_aws_auth_header(req, @aws_access_key_id, @aws_secret_access_key, bucket, key, path_args)
+         
+         http = $S3syncHttp
+            
+         if req.request_body_permitted?
+           return http.request(req, data, streamOut)
+         else
+           return http.request(req, nil, streamOut)
+         end
+      end
+
+		# a "get" operation that sends the body to an IO stream
+		def get_stream(bucket, key, headers={}, streamOut=nil)
+         return GetResponse.new(make_request('GET', bucket, CGI::escape(key), {}, headers, '', {}, streamOut))
+		end
+		
+		# a "get" operation that sends the body to an IO stream
+		def get_query_stream(bucket, key, path_args={}, headers={}, streamOut=nil)
+         return GetResponse.new(make_request('GET', bucket, CGI::escape(key), path_args, headers, '', {}, streamOut))
+		end
+      
+		def head(bucket, key=nil, headers={})
+         return GetResponse.new(make_request('HEAD', bucket, CGI::escape(key), {}, headers, '', {}))
+		end
+      undef create_bucket
+      def create_bucket(bucket, object)
+         object = S3Object.new(object) if not object.instance_of? S3Object
+         return Response.new(
+            make_request('PUT', bucket, '', {}, {}, object.data, object.metadata)
+         )
+      end
+      # no, because internally the library does not support the header,wait,body paradigm, so this is useless
+      #alias __put__ put
+      #def put(bucket, key, object, headers={})
+      #   headers['Expect'] = "100-continue"
+      #   __put__(bucket, key, object, headers)
+      #end
+
+               
+		# allow ssl validation
+      attr_accessor :verify_mode
+      attr_accessor :ca_path
+      attr_accessor :ca_file      
+
+	end
+   module CallingFormat
+      def CallingFormat.string_to_format(s)
+         case s
+         when 'REGULAR'
+           return CallingFormat::REGULAR
+         when 'SUBDOMAIN'
+           return CallingFormat::SUBDOMAIN
+         when 'VANITY'
+           return CallingFormat::VANITY
+         else
+           raise "Unsupported calling format #{s}"
+         end
+      end
+   end
+   
+end

data/lib/S3encoder.rb

@@ -0,0 +1,50 @@
+# This software code is made available "AS IS" without warranties of any        
+# kind.  You may copy, display, modify and redistribute the software            
+# code either by itself or as incorporated into your code; provided that        
+# you do not remove any proprietary notices.  Your use of this software         
+# code is at your own risk and you waive any claim against the author
+# with respect to your use of this software code. 
+# (c) 2007 s3sync.net
+#
+
+# The purpose of this file is to overlay the cgi class
+# to add some functionality
+# (without changing the file itself or requiring a specific version)
+# It still isn't perfectly robust, i.e. if radical changes are made
+# to the underlying lib this stuff will need updating.
+
+require 'cgi'
+require 'iconv' # for UTF-8 conversion
+
+# thanks to http://www.redhillconsulting.com.au/blogs/simon/archives/000326.html
+module S3ExtendCGI
+	def self.included(base)
+		base.extend(ClassMethods)
+		base.class_eval do
+			class << self
+				alias_method :S3Extend_escape_orig, :escape unless method_defined?(:S3Extend_escape_orig)
+				alias_method :escape, :S3Extend_escape
+			end
+		end
+	end
+	module ClassMethods
+		@@exemptSlashesInEscape = false
+		attr_writer :exemptSlashesInEscape
+		@@usePercent20InEscape = false
+		attr_writer :usePercent20InEscape
+		@@nativeCharacterEncoding = "ISO-8859-1"
+		attr_writer :nativeCharacterEncoding
+		@@useUTF8InEscape = false
+		attr_writer :useUTF8InEscape
+		
+		def S3Extend_escape(string)
+			result = string
+			result = Iconv.iconv("UTF-8", @nativeCharacterEncoding, string).join if @useUTF8InEscape 
+			result = S3Extend_escape_orig(result)
+			result.gsub!(/%2f/i, "/") if @exemptSlashesInEscape
+			result.gsub!("+", "%20") if @usePercent20InEscape
+			result
+		end
+	end
+end
+CGI.send(:include, S3ExtendCGI)

data/lib/s3config.rb

@@ -0,0 +1,27 @@
+#!/usr/bin/ruby
+# This software code is made available "AS IS" without warranties of any        
+# kind.  You may copy, display, modify and redistribute the software            
+# code either by itself or as incorporated into your code; provided that        
+# you do not remove any proprietary notices.  Your use of this software         
+# code is at your own risk and you waive any claim against the author
+# with respect to your use of this software code. 
+# (c) 2007 alastair brunton
+#
+# modified to search out the yaml in several places, thanks wkharold.
+require 'yaml'
+
+module S3Config
+  
+  confpath = ["#{ENV['S3CONF']}", "#{ENV['HOME']}/.s3conf", "/etc/s3conf"]
+  
+  confpath.each do |path|
+    if File.exists?(path) and File.directory?(path) and File.exists?("#{path}/s3config.yml")
+      config = YAML.load_file("#{path}/s3config.yml")
+      config.each_pair do |key, value|
+        eval("$#{key.upcase} = '#{value}'")
+      end
+      break
+    end
+  end
+    
+end

data/lib/s3try.rb

@@ -0,0 +1,177 @@
+#!/usr/bin/env ruby 
+# This software code is made available "AS IS" without warranties of any        
+# kind.  You may copy, display, modify and redistribute the software            
+# code either by itself or as incorporated into your code; provided that        
+# you do not remove any proprietary notices.  Your use of this software         
+# code is at your own risk and you waive any claim against the author
+# with respect to your use of this software code. 
+# (c) 2007 s3sync.net
+#
+module S3sync
+
+	$AWS_ACCESS_KEY_ID = ENV["AWS_ACCESS_KEY_ID"]           
+	$AWS_SECRET_ACCESS_KEY = ENV["AWS_SECRET_ACCESS_KEY"]   
+	$AWS_S3_HOST = (ENV["AWS_S3_HOST"] or "s3.amazonaws.com")
+   $HTTP_PROXY_HOST = ENV["HTTP_PROXY_HOST"]
+   $HTTP_PROXY_PORT = ENV["HTTP_PROXY_PORT"]
+   $HTTP_PROXY_USER = ENV["HTTP_PROXY_USER"]
+   $HTTP_PROXY_PASSWORD = ENV["HTTP_PROXY_PASSWORD"]
+	$SSL_CERT_DIR = ENV["SSL_CERT_DIR"]
+	$SSL_CERT_FILE = ENV["SSL_CERT_FILE"]
+	$S3SYNC_RETRIES = (ENV["S3SYNC_RETRIES"] or 100).to_i # number of errors to tolerate 
+	$S3SYNC_WAITONERROR = (ENV["S3SYNC_WAITONERROR"] or 30).to_i # seconds
+	$S3SYNC_NATIVE_CHARSET = (ENV["S3SYNC_NATIVE_CHARSET"] or "ISO-8859-1")
+	$AWS_CALLING_FORMAT = (ENV["AWS_CALLING_FORMAT"] or "REGULAR")
+   
+	require 'S3'
+
+	require 'HTTPStreaming'
+	require 'S3encoder'
+	CGI::exemptSlashesInEscape = true
+	CGI::usePercent20InEscape = true
+	CGI::useUTF8InEscape = true
+	CGI::nativeCharacterEncoding = $S3SYNC_NATIVE_CHARSET
+	require 'S3_s3sync_mod'
+
+
+	$S3syncRetriesLeft = $S3SYNC_RETRIES.to_i
+	
+	def S3sync.s3trySetup 	
+		
+		# ---------- CONNECT ---------- #
+
+		$S3syncConnection = S3::AWSAuthConnection.new($AWS_ACCESS_KEY_ID, $AWS_SECRET_ACCESS_KEY, $S3syncOptions['--ssl'], $AWS_S3_HOST)
+      $S3syncConnection.calling_format = S3::CallingFormat::string_to_format($AWS_CALLING_FORMAT)
+		if $S3syncOptions['--ssl']
+			if $SSL_CERT_DIR
+				$S3syncConnection.verify_mode = OpenSSL::SSL::VERIFY_PEER
+				$S3syncConnection.ca_path = $SSL_CERT_DIR
+			elsif $SSL_CERT_FILE
+				$S3syncConnection.verify_mode = OpenSSL::SSL::VERIFY_PEER
+				$S3syncConnection.ca_file = $SSL_CERT_FILE
+			end
+		end
+	end
+	def S3sync.s3urlSetup 	
+		$S3syncGenerator = S3::QueryStringAuthGenerator.new($AWS_ACCESS_KEY_ID, $AWS_SECRET_ACCESS_KEY, $S3syncOptions['--ssl'], $AWS_S3_HOST)
+      $S3syncGenerator.calling_format = S3::CallingFormat::string_to_format($AWS_CALLING_FORMAT)
+      $S3syncGenerator.expires_in = $S3syncOptions['--expires-in']
+	end
+   
+   def S3sync.S3tryConnect(bucket, host='')
+         $S3syncHttp = $S3syncConnection.make_http(bucket, host, $HTTP_PROXY_HOST, $HTTP_PROXY_PORT, $HTTP_PROXY_USER, $HTTP_PROXY_PASSWORD)
+   end
+	
+	def S3sync.S3try(command, bucket, *args)
+      if(not $S3syncHttp or (bucket != $S3syncLastBucket))
+         $stderr.puts "Creating new connection" if $S3syncOptions['--debug']
+         $S3syncLastBucket = bucket
+         while $S3syncRetriesLeft > 0 do
+            begin
+               S3sync.S3tryConnect(bucket)
+               break
+            rescue Errno::ECONNRESET => e
+               $stderr.puts "Connection reset: #{e}" 
+            rescue Errno::ECONNABORTED => e
+               $stderr.puts "Connection aborted: #{e}" 
+            rescue Errno::ETIMEDOUT => e
+               $stderr.puts "Connection timed out: #{e}"
+            rescue Timeout::Error => e
+               $stderr.puts "Connection timed out: #{e}" 
+            end
+            $S3syncRetriesLeft -= 1
+            $stderr.puts "#{$S3syncRetriesLeft} retries left, sleeping for #{$S3SYNC_WAITONERROR} seconds"
+            Kernel.sleep $S3SYNC_WAITONERROR.to_i
+         end
+      end
+      
+		result = nil
+		delim = $,
+		$,=' '
+		while $S3syncRetriesLeft > 0 do
+         $stderr.puts "Trying command #{command} #{bucket} #{args} with #{$S3syncRetriesLeft} retries left" if $S3syncOptions['--debug']
+			forceRetry = false
+         now = false
+         hush = false
+			begin
+				result = $S3syncConnection.send(command, bucket, *args)
+			rescue Errno::EPIPE => e
+				forceRetry = true
+				$stderr.puts "Broken pipe: #{e}" 
+			rescue Errno::ECONNRESET => e
+				forceRetry = true
+				$stderr.puts "Connection reset: #{e}" 
+			rescue Errno::ECONNABORTED => e
+				forceRetry = true
+				$stderr.puts "Connection aborted: #{e}" 
+			rescue Errno::ETIMEDOUT => e
+				forceRetry = true
+				$stderr.puts "Connection timed out: #{e}"
+			rescue Timeout::Error => e
+				forceRetry = true
+				$stderr.puts "Connection timed out: #{e}" 
+			rescue EOFError => e
+				# i THINK this is happening like a connection reset
+				forceRetry = true
+				$stderr.puts "EOF error: #{e}"
+			rescue OpenSSL::SSL::SSLError => e
+				forceRetry = true
+				$stderr.puts "SSL Error: #{e}"
+			rescue NoMethodError => e
+				# we get this when using --progress, and the local item is something unreadable
+				$stderr.puts "Null stream error: #{e}"
+				break
+			end
+         if forceRetry
+				# kill and reset connection when we receive a non-50x yet retry-able error
+				S3sync.S3tryConnect(bucket)
+         end
+			begin
+				debug("Response code: #{result.http_response.code}")
+				break if ((200...300).include? result.http_response.code.to_i) and (not forceRetry)
+            if result.http_response.code.to_i == 301
+               $stderr.puts "Permanent redirect received. Try setting AWS_CALLING_FORMAT to SUBDOMAIN"
+            elsif result.http_response.code.to_i == 307
+               # move to the other host
+               host = %r{https?://([^/]+)}.match(result.http_response['Location'])[1]
+               $stderr.puts("Temporary Redirect to: " + host)
+               debug("Host: " + host)
+               S3sync.S3tryConnect(bucket, host)
+               $S3syncRetriesLeft = $S3syncRetriesLeft+1 # don't use one up below
+               forceRetry = true
+               now = true
+               hush = true
+            else
+               $stderr.puts "S3 command failed:\n#{command} #{args}"
+               $stderr.puts "With result #{result.http_response.code} #{result.http_response.message}\n"
+               debug(result.http_response.body)
+            end
+				# only retry 500's, per amazon
+				break unless ((500...600).include? result.http_response.code.to_i) or forceRetry
+			rescue NoMethodError
+				debug("No result available")
+			end
+			$S3syncRetriesLeft -= 1
+			$stderr.puts "#{$S3syncRetriesLeft} retries left, sleeping for #{$S3SYNC_WAITONERROR} seconds" unless hush
+			Kernel.sleep $S3SYNC_WAITONERROR.to_i unless now
+		end
+      if $S3syncRetriesLeft <= 0
+         $stderr.puts "Ran out of retries; operations did not complete!"
+      end
+		$, = delim
+		result
+	end
+   
+	def S3sync.S3url(command, bucket, *args)
+      S3sync.s3urlSetup() unless $S3syncGenerator
+		result = nil
+		delim = $,
+		$,=' '
+      $stderr.puts "Calling command #{command} #{bucket} #{args}" if $S3syncOptions['--debug']
+      result = $S3syncGenerator.send(command, bucket, *args)
+		$, = delim
+		result
+	end
+	
+end #module
+