* lib/csv.rb: Fixed a bug in read_to_char() that would slurp

whole files if the encoding was invalid.  It will now read
  up to 10 bytes ahead to find a valid character boundary or
  give up.  [ruby-core:19465]
* test/csv/test_features.rb, test/csv/test_table.rb, test/csv/test_row.rb:
  Loosened some tests to check for a compatible? Encoding instea
  of an exact Encoding.  [ruby-core:19470]



git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@19931 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,13 @@
+Sat Oct 25 09:54:10 2008  James Edward Gray II  <jeg2@ruby-lang.org>
+
+  * lib/csv.rb:  Fixed a bug in read_to_char() that would slurp
+    whole files if the encoding was invalid.  It will now read
+    up to 10 bytes ahead to find a valid character boundary or
+    give up.  [ruby-core:19465]
+  * test/csv/test_features.rb, test/csv/test_table.rb, test/csv/test_row.rb:
+    Loosened some tests to check for a compatible? Encoding instea
+    of an exact Encoding.  [ruby-core:19470]
+
 Sat Oct 25 07:42:49 2008  Eric Hodel  <drbrain@segment7.net>
 
 	* lib/rdoc*: Update to RDoc 2.2.2 r192.

lib/csv.rb

@@ -199,7 +199,7 @@ require "stringio"
 # 
 class CSV
   # The version of the installed library.
-  VERSION = "2.4.3".freeze
+  VERSION = "2.4.4".freeze
   
   # 
   # A CSV::Row is part Array and part Hash.  It retains an order for the fields
@@ -1551,7 +1551,7 @@ class CSV
                   end
     @encoding ||= Encoding.default_internal || Encoding.default_external
     # 
-    # prepare for build safe regular expressions in the target encoding,
+    # prepare for building safe regular expressions in the target encoding,
     # if we can transcode the needed characters
     # 
     @re_esc   =   "\\".encode(@encoding) rescue ""
@@ -2251,10 +2251,11 @@ class CSV
   end
 
   # 
-  # Reads at least +bytes+ from <tt>@io</tt>, but will read on until the data
-  # read is valid in the ecoding of that data.  This should ensure that it is
-  # safe to use regular expressions on the read data.  The read data will be
-  # returned in <tt>@encoding</tt>.
+  # Reads at least +bytes+ from <tt>@io</tt>, but will read up 10 bytes ahead if
+  # needed to ensure the data read is valid in the ecoding of that data.  This
+  # should ensure that it is safe to use regular expressions on the read data,
+  # unless it is actually a broken encoding.  The read data will be returned in
+  # <tt>@encoding</tt>.
   # 
   def read_to_char(bytes)
     return "" if @io.eof?
@@ -2264,10 +2265,12 @@ class CSV
       raise unless encoded.valid_encoding?
       return encoded
     rescue  # encoding error or my invalid data raise
-      if @io.eof?
+      if @io.eof? or data.size >= bytes + 10
         return data
       else
-        data += @io.read(1) until data.valid_encoding? or @io.eof?
+        data += @io.read(1) until data.valid_encoding? or
+                                  @io.eof?             or
+                                  data.size >= bytes + 10
         retry
       end
     end

test/csv/test_features.rb

@@ -250,9 +250,11 @@ class TestCSVFeatures < Test::Unit::TestCase
     end
   end
   
-  def test_inspect_is_ascii_8bit_encoded
+  def test_inspect_encoding_is_ascii_compatible
     CSV.new("one,two,three\n1,2,3\n".encode("UTF-16BE")) do |csv|
-      assert_equal("ASCII-8BIT", csv.inspect.encoding.name)
+      assert( Encoding.compatible?( Encoding.find("US-ASCII"),
+                                    csv.inspect.encoding ),
+              "inspect() was not ASCII compatible." )
     end
   end
   

test/csv/test_row.rb

@@ -296,8 +296,10 @@ class TestCSVRow < Test::Unit::TestCase
     end
   end
   
-  def test_inspect_is_ascii_8bit_encoded
-    assert_equal("ASCII-8BIT", @row.inspect.encoding.name)
+  def test_inspect_encoding_is_ascii_compatible
+    assert( Encoding.compatible?( Encoding.find("US-ASCII"),
+                                  @row.inspect.encoding ),
+            "inspect() was not ASCII compatible." )
   end
   
   def test_inspect_shows_symbol_headers_as_bare_attributes

test/csv/test_table.rb

@@ -400,7 +400,9 @@ class TestCSVTable < Test::Unit::TestCase
     assert(str.include?("mode:#{@table.mode}"), "Mode not shown.")
   end
   
-  def test_inspect_is_us_ascii_encoded
-    assert_equal("US-ASCII", @table.inspect.encoding.name)
+  def test_inspect_encoding_is_ascii_compatible
+    assert( Encoding.compatible?( Encoding.find("US-ASCII"),
+                                  @table.inspect.encoding ),
+            "inspect() was not ASCII compatible." )
   end
 end