damog-goodies 0.2 → 0.3

data/lib/goodies/html.rb

@@ -2,7 +2,7 @@ require "rubygems"
 require "hpricot"
 
 # for 'get'
-require "#{File.join(File.expand_path(File.dirname(__FILE__)), "lwr")}"
+require "#{File.join(File.expand_path(File.dirname(__FILE__)), "lwr-simple")}"
 
 # for 'has?'
 require "#{File.join(File.expand_path(File.dirname(__FILE__)), "array")}"

data/lib/goodies/lwr-simple.rb

@@ -0,0 +1,179 @@
+# Little tribute to LWP::Simple
+require "open-uri"
+require "tempfile"
+
+module LWR
+  class LWR::Simple
+		
+    def self.normalize(url)
+      case url
+        when URI
+        when String
+          url = URI.parse(url)
+        else
+          raise ArgumentError, "URI or String expected, got: "
+      end
+
+      unless url.scheme
+        url = URI.parse("http://" << url.to_s)
+      end
+
+      if url.path.empty?
+        url.path = "/"
+      end
+
+      url
+    end
+  end
+
+end
+
+
+module Kernel
+	# get will fetch the document identified by the given URL and return it.
+	# The url argument can be either a simple string or a URI object.
+	#
+	# You will not be able to examine the response code or response headers
+	# like ('Content-Type') wgen you are accessing the web using this function.
+	# If you need that information you should use the full OO interface of Net::HTTP.
+	def get(uri)
+		url = LWR::Simple.normalize(uri)
+
+		begin
+			return open(url).read
+		rescue
+			nil
+		end
+	end
+
+	# Get document headers. Returns the following 5 values if successful:
+	# - content_type
+	# - document_length
+	# - modified_time
+	# - expires
+	# - server
+	#
+	# Returns an empty array if it fails.
+	def head(uri)
+		url = LWR::Simple.normalize(uri)
+
+		head = []
+
+		begin
+			Net::HTTP::start(url.host, url.port) do |c|
+				res = c.request_head(url.path)
+				head << res["content-type"]
+				head << res["content-length"]
+				head << res["last-modified"]
+				head << res["expires"]
+				head << res["server"]
+			end
+		ensure
+			return head
+		end
+	end
+
+	# Get and print a document identified by URL. If the request fails, then the
+	# status code and message are printed on $stderr. The return value is the
+	# HTTPResponse object. See also getputs.
+	def getprint(uri, limit = 10)
+		raise ArgumentError, 'HTTP redirect too deep' if limit == 0
+		
+		url = LWR::Simple.normalize(uri)
+
+		begin
+			res = Net::HTTP.get_response(url)
+		rescue => ex
+			$stderr.puts "#{ex.class}: #{ex.message}"
+		else
+			case res
+				when Net::HTTPSuccess
+					puts res.body
+				when Net::HTTPRedirection
+					getprint(res["location"], limit - 1)
+				else
+					$stderr.puts "#{res.code} #{res.message}"
+			end
+		ensure
+			return res
+		end
+	end
+
+	# Gets a document identified by a URL and stores it in the file. The URL
+	# can either be a string or a URI object. File can be either a string or
+	# a File object. If a problem ocurrs while writing to a file, nil is
+	# returned. If it succeeds, the return value is the HTTPResponse object.
+	def getstore(uri, file)
+		res = nil
+
+		file = file.path if file.is_a? File
+
+		begin
+			File.open(file, "w") do |f|
+				res = f.getprint(uri) # this is why Ruby is so sweet
+			end
+		rescue
+			return nil
+		end
+
+		res
+	end
+
+	def mirror(uri, file)
+		url = LWR::Simple.normalize(uri)
+		req = Net::HTTP::Get.new(url.path)
+
+		case file
+			when String
+				if File.exists? file
+					file_exists = true
+				else
+					file_exists = false
+				end
+				file = File.open(file, "w")
+			when File
+				if File.exists? file.path
+					file_exists = true
+				else
+					file_exists = false
+				end
+			else
+				raise ArgumentError, "Failed arguments"
+		end
+
+		if file_exists
+			req.add_field("If-Modified-Since", file.mtime.httpdate)
+		end
+
+		res = Net::HTTP.new(url.host, url.port).start do |http|
+			http.request(req)
+		end
+
+		case res
+			when Net::HTTPSuccess
+				tmpfile = Tempfile.new("lwr")
+				tmpfile.print res.body
+				
+				if res["content-length"] and tmpfile.size < res["content-length"].to_i
+					raise Exception, "Transfer truncated, only #{tmpfile.size} of #{res["content-length"]} bytes received"
+					File.unlink(tmpfile.path)
+				elsif res["content-length"] and tmpfile.size > res["content-length"].to_i
+					raise Exception, "Content-length mismatch, expected #{res["content-length"]} bytes, got #{tmpfile.size}"
+					File.unlink(tmpfile.path)
+				else # OK
+					File.unlink(file.path) if File.exists? file.path
+					File.rename(tmpfile.path, file.path)
+
+				end
+
+			else
+		end
+
+		res
+
+	end
+		
+	alias getputs getprint
+
+end
+

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: damog-goodies
 version: !ruby/object:Gem::Version 
-  version: "0.2"
+  version: "0.3"
 platform: ruby
 authors: 
 - David Moreno <david@axiombox.com>
@@ -9,11 +9,12 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2008-01-03 00:00:00 -08:00
+date: 2009-02-28 00:00:00 -08:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: hpricot
+  type: :runtime
   version_requirement: 
   version_requirements: !ruby/object:Gem::Requirement 
     requirements: 
@@ -34,7 +35,7 @@ files:
 - lib/goodies.rb
 - lib/goodies/array.rb
 - lib/goodies/html.rb
-- lib/goodies/lwr.rb
+- lib/goodies/lwr-simple.rb
 has_rdoc: true
 homepage: http://github.com/damog/goodies
 post_install_message: 

data/lib/goodies/lwr.rb

@@ -1,13 +0,0 @@
-# Little tribute to LWP
-require "open-uri"
-
-module Kernel
-	def get(url)
-		uri = URI.parse(url)
-		if uri.scheme
-			return open(url).read
-		else
-			return open("http://" + url).read
-		end
-	end
-end