mammoth-hasher 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 5c93a9f01f7da01607911a7f7c5df3acc1ff0957
-  data.tar.gz: 1704ea09e5aeb7422ed007826d66363c2ec6737d
+  metadata.gz: e8a53d70781a6348d58af444f42747ffcb700b64
+  data.tar.gz: caa0fff7b6d9053006daf8b1bcb2cd4e4f0e5c5f
 SHA512:
-  metadata.gz: b6040ec90679e24ecd5e1bbdf08a6ac719ac1a0621ec4fa8767a092b1864b448aa47a0bc52c1940fa5060f825392bc060030d25b0ff14b17c3e0ea0827729cc2
-  data.tar.gz: 9c5802cf32e2743a6f99448f62e108880c3ad1963b6052de8442b52aee504c7e02237d088879beb2819854caa0c4d0ccc305f84d5dd0e936fb4f54cd1f35614e
+  metadata.gz: 4dcad29b2156bbe8343a2f5fd93c637b1ec447ffdb2f062b5fc6eb5dab3cc9857ab7c325aa29a3fe862c2ee2622f8f211e04d24f8e3da2204747bba2f2531947
+  data.tar.gz: 7b224f29d208466247a2a89b7e0e60190ab42a4b088bd01a72bbc982cdefbd6389f0369e3304c70838a6f5f6b591667e189f03cf4b73a3d6a45c91befb489a1d

data/Rakefile

@@ -0,0 +1,8 @@
+require 'rake/testtask'
+
+Rake::TestTask.new do |t|
+  t.libs << 'test'
+end
+
+desc "Run tests"
+task :default => :test

data/lib/mammoth-hasher.rb

@@ -13,7 +13,7 @@ class MammothHasher
     # algorithm parameters
     # WARNING: if you change them, the resulting hash will be different !
     number_of_chunks = 100
-    length_of_chunks = 100
+    length_of_chunks = 4
 
     # we get the file size (in bytes), used as PRNG (Pseudo Random Number Generator)
     filesize = File.size filename
@@ -22,18 +22,22 @@ class MammothHasher
     # the MD5 of the whole file than to apply our custom algorithm
     if filesize <= number_of_chunks*length_of_chunks
       file = File.open(filename, 'r')
-      final_hash = Digest::MD5.file(file).hexdigest
+      hash = Digest::MD5.file(file).hexdigest
       file.close
       puts (Time.now - time_start).to_s + " seconds" if debug
-      return final_hash
+      return hash
     end
 
     # we initialize the PRNG
     prng = Random.new filesize
 
-    # we get 1000 numbers between 0 and filesize-size_of_chunk
-    offsets = []
-    for i in 0..number_of_chunks
+    # we always get a chunk at the offset 0 (beginning of file)
+    # because that's where the magic number indicating the file type is
+    # so making sure that it's still the same may prevent from some attacks
+    offsets = [0]
+
+    # we get 99 other offsets between 0 and (filesize - length_of_chunk)
+    for i in 1..(number_of_chunks-1)
       offsets << prng.rand(filesize - length_of_chunks)
     end
 
@@ -41,24 +45,17 @@ class MammothHasher
     # (in order to optimize the way the file will be read (in only one direction))
     offsets.sort
 
-    # we compute the hashes of several parts of the file
-    hashes = ""
-    # first, we compute the hash of the first bytes of the file,
-    # because that's where the magic number indicating the file type is
-    # so making sure that it's still the same may be safer
-    hashes << Digest::MD5.new.hexdigest(File.read(filename, 100))
-    # for each offset, we compute the hash of the following bytes
-    # and we concatenate these hashes
+    # we concatenate all the bytes from all the chunks at the offset we choose
+    bytes = ""
     for offset in offsets
-      hashes += Digest::MD5.new.hexdigest(File.read(filename, length_of_chunks, offset))
+      bytes = "#{bytes}#{File.read(filename, length_of_chunks, offset)}"
     end
 
-    # we compute the final hash, which is the hash of the concatenation
-    # of the previous hashes
-    final_hash = Digest::MD5.new.hexdigest hashes
+    # we compute the final hash, which is the hash of the concatenation of all chunks
+    hash = Digest::MD5.new.hexdigest bytes
 
     puts (Time.now - time_start).to_s + " seconds" if debug
 
-    return final_hash
+    return hash
   end
 end

data/test/test_mammoth-hasher.rb

@@ -0,0 +1,24 @@
+require 'test/unit'
+require 'digest'
+require 'mammoth-hasher'
+
+class MammothHasherTest < Test::Unit::TestCase
+  # for small files, instead of using our custom algorithm,
+  # it's simpler to use the md5 hash directly,
+  # so here we test that MammothHasher hash is the same than md5 hash
+  def test_small_file_hash
+    filename = "test/fixtures/small.txt"
+    file = File.open(filename, 'r')
+    assert_equal MammothHasher.hash(filename), Digest::MD5.file(file).hexdigest
+  end
+
+  def test_hash_size
+    filename = "test/fixtures/large.txt"
+    assert_equal MammothHasher.hash(filename).length, 32
+  end
+
+  def test_hash_result
+    filename = "test/fixtures/large.txt"
+    assert_equal MammothHasher.hash(filename), "d5d198a347f02adafa6e1749ad594340"
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: mammoth-hasher
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors:
 - Vincent Marquet
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-02-21 00:00:00.000000000 Z
+date: 2015-06-07 00:00:00.000000000 Z
 dependencies: []
 description: A library to compute fingerprints for big files, when runnning usual
   algorithms as MD5 is too long.
@@ -17,10 +17,12 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- Rakefile
 - lib/mammoth-hasher.rb
+- test/test_mammoth-hasher.rb
 homepage: http://github.com/vmarquet/ruby-mammoth-hasher
 licenses:
-- WTFPL
+- MIT
 metadata: {}
 post_install_message: 
 rdoc_options: []
@@ -38,8 +40,9 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.4.5
+rubygems_version: 2.4.6
 signing_key: 
 specification_version: 4
 summary: A library to compute fingerprints for big files.
-test_files: []
+test_files:
+- test/test_mammoth-hasher.rb