[ruby-core:69503] [Ruby trunk - Feature #10900] GzipReader does not define `#external_encoding`

Issue #10900 has been updated by Aleksandar Kostadinov.


Daniel Berger wrote:
> I'd say it should add #encoding and #internal_encoding methods as well, since the constructor supports those options.

I guess you have to file a separate feature request for that.

----------------------------------------
Feature #10900: GzipReader does not define `#external_encoding`
https://bugs.ruby-lang.org/issues/10900#change-52806

* Author: Aleksandar Kostadinov
* Status: Closed
* Priority: Normal
* Assignee: 
----------------------------------------
Hello,

I want to store/load some **yaml** and other text data in gz for space efficiency. I thought it's most efficient to read/write directly to/from a gzip stream. But seems like the **GzipReader** io object does not support everything that a regular file io object does. I spotted the **#external_encoding** method. So for example to load YAML directly from gzip I use the following method:

~~~
    def load_yaml_from_gz(gz)
      loaded = nil
      Zlib::GzipReader.open(gz) do |f|
        class << f
          def external_encoding
            Encoding::UTF_8
          end
        end
        loaded = YAML.load(f, gz)
      end
      return loaded
    end
~~~

I need the monkey-patching because otherwise the YAML.load method fails for no **#external_encoding** method defined.

To avoid monkey patching just to read some text from a gzip archive I propose that GzipReader implements **#external_encoding** the same way as **File** does.



-- 
https://bugs.ruby-lang.org/