Re: kannel OTA/OMA dev ?

Yep. Libxml creates a parse tree using WML DTD and wml compiler
picks tokens from the tree. Perhaps you should try adding string
table code to ota_compiler.c ?

On 22 Oct 2008, at 13:02, P. A. Bagyenda wrote:

> ota_compiler.c
> On Oct 22, 2008, at 13:07, Guillaume Godart wrote:
>
>> are you absolutely sure it is the source code used to compile wbxml
>> content from xml content ?
>>
>>
>>
>>
>>
>>
>> Le mercredi 22 octobre 2008 à 11:31 +0200, Aarno Syvänen a écrit :
>>> Hi,
>>>
>>> wml_compiler.c does have string table code. Perhaps you can begin  
>>> with
>>> it.
>>>
>>> Aarno
>>>
>>> On 22 Oct 2008, at 10:03, Guillaume Godart wrote:
>>>
>>>> Thank you for your work !
>>>>
>>>> The solution is to add the three token (CODEC, MEDIASUBTYPE and
>>>> PRIORITYINDEX) in a string table at the begining of the wbxml  
>>>> content.
>>>> How can I patch kannel to do that ?
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> Le mercredi 22 octobre 2008 à 05:04 +0200, Stipe Tolj a écrit :
>>>>> Guillaume Godart schrieb:
>>>>>> solution seems to be described here :
>>>>>> http://discussion.forum.nokia.com/forum/showthread.php?t=124252
>>>>>>
>>>>>> they speak about inserting a string table in the begining of the
>>>>>> wbxml,
>>>>>> adding the 3 words CODEC, MEDIASUBTYPE and PRIORITYINDEX.
>>>>>
>>>>> ok, after reviewing a couple of documents and specs:
>>>>>
>>>>> These XML elements have NOT being defined by OMA to be tokenized.
>>>>> Which means if
>>>>> there is no defined WBXML token for them in the token tables, any
>>>>> XML->WBXML
>>>>> conversion will leave them as plain ASCII encoded bytes in the
>>>>> resulting WBXML
>>>>> string stream.
>>>>>
>>>>> Stipe
>>>>>
>>>>
>>>>
>>>
>