[
https://issues.apache.org/jira/browse/CAMEL-8346?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14320269#comment-14320269
]
Stefan Mandel commented on CAMEL-8346:
--------------------------------------
I know of this workaround (it works). I also tried with
.convertBodyTo(InputStream.class, "iso-8859-1") but this won't work.
I expect that the workaround is fine for small files, but not so robust on
large files since the complete file is loaded into memory. This may be a
theoretical problem, because I actually do not know how the Jsonpathengine acts
on streams.
I think a fix should consider first to the configured file charset, then to the
JSON-default-Encoding (as implemented in 2.15.x) and never to the
System-charset-encoding. Do you agree?
> JsonPathEngine skips file encoding
> ----------------------------------
>
> Key: CAMEL-8346
> URL: https://issues.apache.org/jira/browse/CAMEL-8346
> Project: Camel
> Issue Type: Bug
> Components: camel-jsonpath
> Affects Versions: 2.14.1, 2.15.0
> Reporter: Stefan Mandel
> Attachments: JsonPathSourceTest.java, germanbooks-iso-8859-1.json,
> germanbooks-utf8.json
>
>
> This occurs if one uses the camel file component with a charset parameter,
> e.g. file:/url?charset=UTF-8. It leads to strange effects:
> # testing the same utf8-file under linux (default encoding:UTF-8) and german
> windows (default encoding:windows-1252) leads to problems with german umlauts.
> # using a file in another encoding than the default (ISO-8859-1 on linux)
> leads also to problems with german umlauts.
> - Both problems can be reproduced with the attached unit test files in 2.14.1.
> - Problem 1 is hidden in 2.15-SNAPSHOT, because UTF-8 is hard-coded as file
> encoding (maybe ok, since JSON is almost never used with language specific
> encondings), Problem 2 also occurs in 2.15-SNAPSHOT
> I can provide or help with a fix (or pull request), if these problems are
> classified as true defects.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
