source_encoding_spec.rb

require_relative '../spec_helper'

describe "Source files" do

  describe "encoded in UTF-8 without a BOM" do
    it "can be parsed" do
      ruby_exe(fixture(__FILE__, "utf8-nobom.rb"), args: "2>&1").should == "hello\n"
    end
  end

  describe "encoded in UTF-8 with a BOM" do
    it "can be parsed" do
      ruby_exe(fixture(__FILE__, "utf8-bom.rb"), args: "2>&1").should == "hello\n"
    end
  end

  describe "encoded in UTF-16 LE without a BOM" do
    it "are parsed because empty as they contain a NUL byte before the encoding comment" do
      ruby_exe(fixture(__FILE__, "utf16-le-nobom.rb"), args: "2>&1").should == ""
    end
  end

  describe "encoded in UTF-16 LE with a BOM" do
    it "are invalid because they contain an invalid UTF-8 sequence before the encoding comment" do
      bom = "\xFF\xFE".b
      source = "# encoding: utf-16le\nputs 'hello'\n"
      source = bom + source.bytes.zip([0]*source.bytesize).flatten.pack('C*')
      path = tmp("utf16-le-bom.rb")

      touch(path, "wb") { |f| f.write source }
      begin
        ruby_exe(path, args: "2>&1", exit_status: 1).should =~ /invalid multibyte char/
      ensure
        rm_r path
      end
    end
  end

  describe "encoded in UTF-16 BE without a BOM" do
    it "are parsed as empty because they contain a NUL byte before the encoding comment" do
      ruby_exe(fixture(__FILE__, "utf16-be-nobom.rb"), args: "2>&1").should == ""
    end
  end

  describe "encoded in UTF-16 BE with a BOM" do
    it "are invalid because they contain an invalid UTF-8 sequence before the encoding comment" do
      bom = "\xFE\xFF".b
      source = "# encoding: utf-16be\nputs 'hello'\n"
      source = bom + ([0]*source.bytesize).zip(source.bytes).flatten.pack('C*')
      path = tmp("utf16-be-bom.rb")

      touch(path, "wb") { |f| f.write source }
      begin
        ruby_exe(path, args: "2>&1", exit_status: 1).should =~ /invalid multibyte char/
      ensure
        rm_r path
      end
    end
  end

end