ruby/spec/rubyspec/library/zlib/inflate/append_spec.rb
eregon 95e8c48dd3 Add in-tree mspec and ruby/spec
* For easier modifications of ruby/spec by MRI developers.
* .gitignore: track changes under spec.
* spec/mspec, spec/rubyspec: add in-tree mspec and ruby/spec.
  These files can therefore be updated like any other file in MRI.
  Instructions are provided in spec/README.
  [Feature #13156] [ruby-core:79246]

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58595 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2017-05-07 12:04:49 +00:00

60 lines
1.5 KiB
Ruby

require File.expand_path('../../../../spec_helper', __FILE__)
require 'zlib'
describe "Zlib::Inflate#<<" do
before :all do
@foo_deflated = [120, 156, 75, 203, 207, 7, 0, 2, 130, 1, 69].pack('C*')
end
before :each do
@z = Zlib::Inflate.new
end
after :each do
@z.close unless @z.closed?
end
it "appends data to the input stream" do
@z << @foo_deflated
@z.finish.should == 'foo'
end
it "treats nil argument as the end of compressed data" do
@z = Zlib::Inflate.new
@z << @foo_deflated << nil
@z.finish.should == 'foo'
end
it "just passes through the data after nil argument" do
@z = Zlib::Inflate.new
@z << @foo_deflated << nil
@z << "-after_nil_data"
@z.finish.should == 'foo-after_nil_data'
end
it "properly handles data in chunks" do
# add bytes, one by one
@foo_deflated.each_byte { |d| @z << d.chr}
@z.finish.should == "foo"
end
it "properly handles incomplete data" do
# add bytes, one by one
@foo_deflated[0, 5].each_byte { |d| @z << d.chr}
lambda { @z.finish }.should raise_error(Zlib::BufError)
end
it "properly handles excessive data, byte-by-byte" do
# add bytes, one by one
data = @foo_deflated * 2
data.each_byte { |d| @z << d.chr}
@z.finish.should == "foo" + @foo_deflated
end
it "properly handles excessive data, in one go" do
# add bytes, one by one
data = @foo_deflated * 2
@z << data
@z.finish.should == "foo" + @foo_deflated
end
end