You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@camel.apache.org by Alex Dettinger <al...@gmail.com> on 2018/02/15 21:33:22 UTC

Re: Potential bug in convertBodyTo

I would not say that it's a bug.
Actually, when convertBodyTo ends up decoding bytes to string, I would
expect the charset to be removed.
When convertBodyTo ends up encoding a string into bytes, I would expect the
charset to be overridden (I don't think that it is the current
implementation however).

On Thu, Feb 15, 2018 at 7:38 PM, Quinn Stevenson <
quinn@pronoia-solutions.com> wrote:

> I found a situation where Camel isn’t doing what I expected, and I wanted
> to ask the developers before I created the JIRA and changed the behavior
> that I believe is a bug.
>
> What I’m seeing is when I call convertBodyTo with a both type and a
> character set supplied and the exchange already has a character set
> specified, the character set is being removed by the convertBodyTo
> operation.  I would expect the original character set on the exchange would
> be restored, not removed.
>
> Is this functionality by design or is it a bug?
>
>
>
>
>

Re: Potential bug in convertBodyTo

Posted by Quinn Stevenson <qu...@pronoia-solutions.com>.
Hmmm….

I would expect that if the character set was already on the exchange before the call, it should still be there after the call - that is the issue I hit - it was removing the CamelCharsetName exchange property that was already there.  If the charset is specified on the convertBodyTo, I think it should use that charset - but is should either leave the existing CamelCharsetName there or replace it with the new value.

Any other thoughts?


> On Feb 15, 2018, at 2:33 PM, Alex Dettinger <al...@gmail.com> wrote:
> 
> I would not say that it's a bug.
> Actually, when convertBodyTo ends up decoding bytes to string, I would
> expect the charset to be removed.
> When convertBodyTo ends up encoding a string into bytes, I would expect the
> charset to be overridden (I don't think that it is the current
> implementation however).
> 
> On Thu, Feb 15, 2018 at 7:38 PM, Quinn Stevenson <
> quinn@pronoia-solutions.com> wrote:
> 
>> I found a situation where Camel isn’t doing what I expected, and I wanted
>> to ask the developers before I created the JIRA and changed the behavior
>> that I believe is a bug.
>> 
>> What I’m seeing is when I call convertBodyTo with a both type and a
>> character set supplied and the exchange already has a character set
>> specified, the character set is being removed by the convertBodyTo
>> operation.  I would expect the original character set on the exchange would
>> be restored, not removed.
>> 
>> Is this functionality by design or is it a bug?
>> 
>> 
>> 
>> 
>>