From 1b211bcc08ddff68d52a0c124bbbe5f81c8527fe Mon Sep 17 00:00:00 2001
From: Mike West <mike@mikewest.org>
Date: Tue, 19 Oct 2010 13:32:03 +0200
Subject: [PATCH] Specify encoding for Pygments

This closes issue #10, in theory, but I'm not completely happy with the
behavior.  The output for both UTF-8 and ISO-8859-1 sources is arguably
correct, but I think it'd be better to do some autodetecting of the file
encoding, and explicitly convert everything to UTF-8 on input.  One
option is the [`chardet` gem][gem], but I'm loath to add another
dependency to Rocco...

[gem]: http://rubygems.org/gems/chardet/versions/0.9.0
---
 lib/rocco.rb                        |  2 +-
 test/fixtures/issue10.iso-8859-1.rb |  1 +
 test/fixtures/issue10.rb            |  1 -
 test/fixtures/issue10.utf-8.rb      |  1 +
 test/rocco_test.rb                  | 16 +++++++++++++---
 5 files changed, 16 insertions(+), 5 deletions(-)
 create mode 100644 test/fixtures/issue10.iso-8859-1.rb
 delete mode 100644 test/fixtures/issue10.rb
 create mode 100644 test/fixtures/issue10.utf-8.rb

diff --git a/lib/rocco.rb b/lib/rocco.rb
index 7abda79..ec62b39 100644
--- a/lib/rocco.rb
+++ b/lib/rocco.rb
@@ -194,7 +194,7 @@ class Rocco
   # then fork off a child process to write the input.
   def highlight_pygmentize(code)
     code_html = nil
-    open("|pygmentize -l #{@options[:language]} -f html", 'r+') do |fd|
+    open("|pygmentize -l #{@options[:language]} -O encoding=utf-8 -f html", 'r+') do |fd|
       pid =
         fork {
           fd.close_read
diff --git a/test/fixtures/issue10.iso-8859-1.rb b/test/fixtures/issue10.iso-8859-1.rb
new file mode 100644
index 0000000..34d1a80
--- /dev/null
+++ b/test/fixtures/issue10.iso-8859-1.rb
@@ -0,0 +1 @@
+# hello wörld
diff --git a/test/fixtures/issue10.rb b/test/fixtures/issue10.rb
deleted file mode 100644
index 93e2940..0000000
--- a/test/fixtures/issue10.rb
+++ /dev/null
@@ -1 +0,0 @@
-hello ąćęłńóśźż
diff --git a/test/fixtures/issue10.utf-8.rb b/test/fixtures/issue10.utf-8.rb
new file mode 100644
index 0000000..e87e125
--- /dev/null
+++ b/test/fixtures/issue10.utf-8.rb
@@ -0,0 +1 @@
+# hello ąćęłńóśźż
diff --git a/test/rocco_test.rb b/test/rocco_test.rb
index ff86ed8..58c0820 100644
--- a/test/rocco_test.rb
+++ b/test/rocco_test.rb
@@ -101,10 +101,20 @@ class RoccoIssueTests < Test::Unit::TestCase
     def test_issue10_utf8_processing
         # Rocco has issues with strange UTF-8 characters: need to explicitly set the encoding for Pygments
         # http://github.com/rtomayko/rocco/issues#issue/10
-        r = Rocco.new( File.dirname(__FILE__) + "/fixtures/issue10.rb" )
+        r = Rocco.new( File.dirname(__FILE__) + "/fixtures/issue10.utf-8.rb" )
         assert_equal(
-            "<p> hello ąćęłńóśźż</p>\n",
-            r.sections[0][0]
+            "<p>hello ąćęłńóśźż</p>\n",
+            r.sections[0][0],
+            "UTF-8 input files ought behave correctly."
+        )
+        # and, just for grins, ensure that iso-8859-1 works too.
+        # @TODO:    Is this really the correct behavior?  Converting text
+        #           to UTF-8 on the way out is probably preferable.
+        r = Rocco.new( File.dirname(__FILE__) + "/fixtures/issue10.iso-8859-1.rb" )
+        assert_equal(
+            "<p>hello w\366rld</p>\n",
+            r.sections[0][0],
+            "ISO-8859-1 input should probably also behave correctly."
         )
     end
     def test_issue12_css_octothorpe_classname_change