Merge pull request #340 from alsor/force-entry-names-encoding-option

add option to force entry names encoding
rubyzip · Nov 8, 2017 · 1039b28 · 1039b28
2 parents fc83680 + deb6616
commit 1039b28
Show file tree

Hide file tree

Showing 4 changed files with 35 additions and 1 deletion.
diff --git a/README.md b/README.md
@@ -254,6 +254,14 @@ Zip.default_compression = Zlib::DEFAULT_COMPRESSION
 ```
 It defaults to `Zlib::DEFAULT_COMPRESSION`. Possible values are `Zlib::BEST_COMPRESSION`, `Zlib::DEFAULT_COMPRESSION` and `Zlib::NO_COMPRESSION`
 
+Sometimes file names inside zip contain non-ASCII characters. If you can assume which encoding was used for such names and want to be able to find such entries using `find_entry` then you can force assumed encoding like so:
+
+```ruby
+Zip.force_entry_names_encoding = 'UTF-8'
+```
+
+Allowed encoding names are the same as accepted by `String#force_encoding`
+
 You can set multiple settings at the same time by using a block:
 
 ```ruby

diff --git a/lib/zip.rb b/lib/zip.rb
@@ -34,7 +34,15 @@
 
 module Zip
   extend self
-  attr_accessor :unicode_names, :on_exists_proc, :continue_on_exists_proc, :sort_entries, :default_compression, :write_zip64_support, :warn_invalid_date, :case_insensitive_match
+  attr_accessor :unicode_names,
+                :on_exists_proc,
+                :continue_on_exists_proc,
+                :sort_entries,
+                :default_compression,
+                :write_zip64_support,
+                :warn_invalid_date,
+                :case_insensitive_match,
+                :force_entry_names_encoding
 
   def reset!
     @_ran_once = false

diff --git a/lib/zip/entry.rb b/lib/zip/entry.rb
@@ -240,6 +240,9 @@ def read_local_entry(io) #:nodoc:all
       extra = io.read(@extra_length)
 
       @name.tr!('\\', '/')
+      if ::Zip.force_entry_names_encoding
+        @name.force_encoding(::Zip.force_entry_names_encoding)
+      end
 
       if extra && extra.bytesize != @extra_length
         raise ::Zip::Error, 'Truncated local zip entry header'
@@ -364,6 +367,9 @@ def read_c_dir_entry(io) #:nodoc:all
       check_c_dir_entry_signature
       set_time(@last_mod_date, @last_mod_time)
       @name = io.read(@name_length)
+      if ::Zip.force_entry_names_encoding
+        @name.force_encoding(::Zip.force_entry_names_encoding)
+      end
       read_c_dir_extra_field(io)
       @comment = io.read(@comment_length)
       check_c_dir_entry_comment_size

diff --git a/test/unicode_file_names_and_comments_test.rb b/test/unicode_file_names_and_comments_test.rb
@@ -33,6 +33,18 @@ def test_unicode_file_name
         assert(filepath == entry_name)
       end
     end
+
+    ::Zip.force_entry_names_encoding = 'UTF-8'
+    ::Zip::File.open(FILENAME) do |zip|
+      file_entrys.each do |filename|
+        refute_nil(zip.find_entry(filename))
+      end
+      directory_entrys.each do |filepath|
+        refute_nil(zip.find_entry(filepath))
+      end
+    end
+    ::Zip.force_entry_names_encoding = nil
+
     ::File.unlink(FILENAME)
   end