[ https://issues.apache.org/jira/browse/IO-833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17847382#comment-17847382 ]
Jochen Wiedmann commented on IO-833: ------------------------------------ [~elharo] As far as I can tell, the implementation does use a Charset: {{ {color:#5a5a5a}{color:#7f0055}private{color}{color:#000000} Charset {color}{color:#0000c0}charset{color}{color:#000000} = Charset.{color}{color:#000000}defaultCharset{color}{color:#000000}();{color}{color}}} I admit, that the choice of Charset is questionable, I'd have recommended UTF_8, but that can't be changed without loosing compatibility. > Every origin needs a charset > ---------------------------- > > Key: IO-833 > URL: https://issues.apache.org/jira/browse/IO-833 > Project: Commons IO > Issue Type: Bug > Reporter: Elliotte Rusty Harold > Priority: Critical > > Every origin except possibly URIOrigin needs a charset. There is no reliable, > acceptable way to convert bytes to chars (ByteOrigin, PathOrigin) or chars to > bytes (CharSequenceOrigin) without it. > The only possible exception is URIOrigin which can have enough metadata to > usefully deduce the charset. > Methods like getBytes and getReader should throw an IllegalStateException if > charset is needed and not supplied. -- This message was sent by Atlassian Jira (v8.20.10#820010)