Changes between Version 3 and Version 4 of HashSourceCodes

Timestamp:

Sep 30, 2008, 5:36:15 PM (16 years ago)

Author:

guest

Comment:

Ruby code improved to handle bigger (?) files. From ​http://github.com/johanlunds/down_sub/tree/lib/down_sub/movie_hasher.rb

HashSourceCodes

@@ -634 +634 @@
 #!ruby
 
-class MovieHasher
-
-  HASH_CHUNK_SIZE = 64 * 1024 # in bytes
-  HASH_SIZE = 8 # in bytes
-  
+module MovieHasher
+
+  CHUNK_SIZE = 64 * 1024 # in bytes
+
   def self.compute_hash(filename)
-    file = File.open(filename, "rb")
     filesize = File.size(filename)
-    
     hash = filesize
-    
+
     # Read 64 kbytes, divide up into 64 bits and add each
     # to hash. Do for beginning and end of file.
-    
-    # Q = unsigned long long (64 bit = HASH_SIZE)
-    file.read(HASH_CHUNK_SIZE).unpack("Q*").each { |n| hash += n }
-    file.seek([0, filesize - HASH_CHUNK_SIZE].max, IO::SEEK_SET)
-    file.read(HASH_CHUNK_SIZE).unpack("Q*").each { |n| hash += n }
-    
-    file.close
-    
-    hash &= 0xffffffffffffffff # keep the first 8 bytes / 64 bits
-    
-    sprintf("%0" + (HASH_SIZE * 2).to_s +  "x", hash)
+    File.open(filename, 'rb') do |f|    
+      # Q = unsigned long long = 64 bit
+      f.read(CHUNK_SIZE).unpack("Q*").each do |n|
+        hash = hash + n & 0xffffffffffffffff # to remain as 64 bit number
+      end
+
+      f.seek([0, filesize - CHUNK_SIZE].max, IO::SEEK_SET)
+
+      # And again for the end of the file
+      f.read(CHUNK_SIZE).unpack("Q*").each do |n|
+        hash = hash + n & 0xffffffffffffffff
+      end
+    end
+
+    sprintf("%016x", hash)
   end
 end
@@ -668 +669 @@
       assert_equal("8e245d9679d31e12", MovieHasher::compute_hash('breakdance.avi'))
     end
+
+    def test_compute_hash_large_file
+      assert_equal("61f7751fc2a72bfb", MovieHasher::compute_hash('dummy.bin'))
+    end
   end
 end