fix: ensure serialization still works with pseudo-IO classes

Fixes #2773
sparklemotion · Jan 22, 2023 · 952ff44 · 952ff44
1 parent 9e137bb
commit 952ff44
Show file tree

Hide file tree

Showing 4 changed files with 52 additions and 2 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -4,6 +4,15 @@ Nokogiri follows [Semantic Versioning](https://semver.org/), please see the [REA
 
 ---
 
+## 1.14.1 / unreleased
+
+### Fixed
+
+* Serializing documents now works again with pseudo-IO objects that don't support IO's encoding API (like rubyzip's `Zip::OutputStream`). This was a regression in v1.14.0 due to the fix for [#752](https://github.com/sparklemotion/nokogiri/issues/752) in [#2434](https://github.com/sparklemotion/nokogiri/issues/2434), and was not completely fixed by [#2753](https://github.com/sparklemotion/nokogiri/issues/2753). [[#2773](https://github.com/sparklemotion/nokogiri/issues/2773)]
+
+2e260f53e6b84b8f9c1b115b0ded85eebc8155d7
+
+
 ## 1.14.0 / 2023-01-12
 
 ### Notable Changes

diff --git a/Gemfile b/Gemfile
@@ -24,6 +24,7 @@ group :development do
   gem "minitest-reporters", "= 1.5.0"
   gem "ruby_memcheck", "1.2.0" unless RUBY_PLATFORM == "java"
   gem "simplecov", "= 0.21.2"
+  gem "rubyzip", "~> 2.3.2"
 
   # rubocop
   if Gem::Requirement.new("~> 3.0").satisfied_by?(Gem::Version.new(RUBY_VERSION))

diff --git a/ext/nokogiri/nokogiri.c b/ext/nokogiri/nokogiri.c
@@ -112,8 +112,13 @@ noko_io_write(void *io, char *c_buffer, int c_buffer_len)
 {
   VALUE rb_args[2], rb_n_bytes_written;
   VALUE rb_io = (VALUE)io;
-  VALUE rb_enc = rb_funcall(rb_io, id_external_encoding, 0);
-  rb_encoding *io_encoding = RB_NIL_P(rb_enc) ? rb_ascii8bit_encoding() : rb_to_encoding(rb_enc);
+  VALUE rb_enc = Qnil;
+  rb_encoding *io_encoding;
+
+  if (rb_respond_to(rb_io, id_external_encoding)) {
+    rb_enc = rb_funcall(rb_io, id_external_encoding, 0);
+  }
+  io_encoding = RB_NIL_P(rb_enc) ? rb_ascii8bit_encoding() : rb_to_encoding(rb_enc);
 
   rb_args[0] = rb_io;
   rb_args[1] = rb_enc_str_new(c_buffer, (long)c_buffer_len, io_encoding);

diff --git a/test/xml/test_document_encoding.rb b/test/xml/test_document_encoding.rb
@@ -87,6 +87,41 @@ class TestDocumentEncoding < Nokogiri::TestCase
           assert_equal(Encoding::UTF_16, output.encoding)
           assert_equal(utf16_document.bytesize, output.bytesize)
         end
+
+        describe "pseudo-IO" do
+          it "serializes correctly with Zip::OutputStream objects" do
+            # https://github.com/sparklemotion/nokogiri/issues/2773
+            require "zip"
+
+            xml = <<~XML
+              <?xml version="1.0" encoding="UTF-8"?>
+              <root>
+                <bar>A</bar>
+              </root>
+            XML
+
+            Dir.mktmpdir do |tmpdir|
+              zipfile_path = File.join(tmpdir, "test.zip")
+
+              Zip::OutputStream.open(zipfile_path) do |io|
+                io.put_next_entry("test-utf8.xml")
+                Nokogiri::XML(xml).write_to(io, encoding: "UTF-8")
+              end
+
+              Zip::InputStream.open(zipfile_path) do |io|
+                entry = io.get_next_entry
+                assert_equal("test-utf8.xml", entry.name)
+                output = io.read
+
+                # no final newline on jruby. descriptive, not prescriptive.
+                expected_length = Nokogiri.jruby? ? xml.bytesize - 1 : xml.bytesize
+
+                assert_equal(Encoding::UTF_8, output.encoding)
+                assert_equal(expected_length, output.bytesize)
+              end
+            end
+          end
+        end
       end
     end
   end