[Enigmail] Invalid UTF-8 Characters

Cristian KLEIN cristiklein at gmail.com
Sat Nov 1 14:28:04 PDT 2008


Patrick Brunschwig a scris:
> Cristian KLEIN wrote:
>> Patrick Brunschwig a scris:
>>> Cristian KLEIN wrote:
>>>>> Hi Cristian,
>>>>>
>>>>>> I'm the maintainer of the ro_RO locale for Enigmail. Recently I observed
>>>>>> that the message for prefGood (
Semntur bun de la %S
) from
>>>>>> enigmail.properties, which contains UTF-8 characters, is rendered
>>>>>> incorrectly in Thunderbird running in Ubuntu Intrepid.
>>>>>>
>>>>>> More exactly, 

 is rendered as a box containing 0003. This message
>>>>>> appears above a message in Thunderbird's main window, or when
>>>>>> double-clicking the envelope icon.
>>>>>>
>>>>>> Interestingly, other messages, such as keyAndSigDateID (
Semnat în
) and
>>>>>> securityInfo (
Informaii de securitate OpenPGP
) are rendered
>>>>>> correctly.
>>>>> usually a box showing the Unicode number only appears when the font used
>>>>> doesn't contain the letter requested, instead the Unicode number is
>>>>> shown.
>>>>> However, without having checked, I doubt that the code of 

 is 0003.
>>>>> In short: no idea.
>>>> Am I really the only one experiencing this problem? I observed that
>>>> beginPgpPart is also rendered distorted. Also, when I replied to this
>>>> message while Enigmail was active, all the special characters (

î
>>>> etc.) where distorted? So, I don't think its a simple localisation issue.
>>>> Also, I have observed that the messages go to different objects
>>>> (statusText in one case, errorObjMsg in another). Is there some kind of
>>>> difference between the two?
>>> Where precisely in the code? Both variables are used quite frequently.
>> The following code is taken from components/enigmail.js:3207
>>
>> if (goodSignature) {
>> 	errorMsgObj.value = trustPrefix + EnigGetString("prefGood",userId)
>>
>> I tried to play around a bit, deleting trustPrefix and changing
>> „prefGood”. If I delete trustPrefix and add „ăâîșț” to prefGood, the
>> message is rendered correctly.
> 
> 
> The problem is actually that the string is decoded twice from UTF-8 to
> the Mozilla-internal format. It's definitely a bug in Enigmail. What is
> interesting though is that it doesn't occur on all my test messages; for
> several messages, I see no problem at all, but for some the text is not
> decoded properly.

Any progress on this one?



More information about the Enigmail mailing list