[
https://issues.apache.org/jira/browse/IO-337?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17800316#comment-17800316
]
Niall Pemberton commented on IO-337:
------------------------------------
I will -1 this change if it is ever committed since it would severely limit
what is provided by the current implementation and prevent it being used for
its original purpose.
As I said in 2015 and if you look at the Wikidepdia page, there are more than 5
listed BOMs. The original poster is incorrect in stating that it is mis-used in
XMLReader - from memory this code came from Tika and was the original purpose
for being developed (XML Guessing).
In its current implementation this functionality can be used for other
purposes, not just BOMs and so its a regressive step to make a change that
would prevent that and limit functionality. This works perfectly well as it is
and I see zero benefit and lots of downsides to this change, so we should close
this as WONTFIX
> ByteOrderMark can be refactored as an enum
> ------------------------------------------
>
> Key: IO-337
> URL: https://issues.apache.org/jira/browse/IO-337
> Project: Commons IO
> Issue Type: New Feature
> Reporter: Yaniv Kunda
> Priority: Minor
> Fix For: 3.x
>
> Attachments: ByteOrderMark-enum.patch
>
>
> ByteOrderMark is used primarily for encapsulating the data regarding to
> Unicode BOMs. Since those are fixed, it would make sense to make it an enum,
> instead of a public class.
> I have attached a patch that only covers ByteOrderMark itself, including an
> additional getCharset() utility method, and toString() simplification.
> This patch does not cover:
> 1) Its incorrect use (per its perceived purpose) in
> XmlStreamReader.XML_GUESS_BYTES
> 2) The now-unnecessary tests
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
