[webkit-gtk] webkitwebframe control

Niranjan Rao nhrdls at gmail.com
Tue Aug 19 10:32:24 PDT 2014


I understand your reasoning. I was hampered by same limitation of not 
able to access frames. My current workaround is to fire xpath query 
searching for "iframe" or "frame" nodes until I find required frame. The 
call is recursive many times as frame might be buried in frame hierarchy.

On 08/19/2014 06:57 AM, Victor Lucero wrote:
> Niranjan,
>
> I understand the need of a webextension in the new api, my only 
> problem is that the new api only exposes the first frame and I need a 
> way to (in the extension) got all the frames or a signal that emits on 
> frame creation so we can get all the frames. but looks like a Bug will 
> need to be filed asking for this feature.
>
>
> Regards.
> Victor Lucero.
>
>
>
> El 15-08-2014, a las 14:14, Niranjan Rao <nhrdls at gmail.com 
> <mailto:nhrdls at gmail.com>> escribió:
>
>> Victor,
>>
>> I don't know about your particular use case, but I had similar need. 
>> I had actually sent couple questions to the list querying about same 
>> or similar things - accessing frame data early this year.
>>
>> For us, it turned out to be more complex exercise than we initially 
>> anticipated to migrate from older version to newer version. It took 
>> some time to understand new web process model and reasoning behind it.
>>
>> We actually had to write our extension and build secondary 
>> communication channel between web and ui process as each process has 
>> its own data and many times you need to access either piece. Simplest 
>> example will be writing a small tool that visits a page, wait for 
>> page load (UI functionality) and then sets color of an element which 
>> can be only done in web process since DOM access is available only in 
>> web processes.
>>
>> After migration, I must admit I am happy with the results. Carols and 
>> other developers have done a great job in creating these releases.
>>
>> The difference I see is that there two ways to use webkit gtk - as a 
>> tool or as browser. Tools need complex interactions between both 
>> processes. Extensions gives you that functionality, but then 
>> extensions need to have communication with main process. May be we 
>> can have API that allows communication between ui and web process. 
>> Example might be send this message to this extension. Then extensions 
>> won't have to build their own functionality and many of these 
>> problems can be solved by simple messages.
>>
>> Regards,
>>
>> Niranjan
>>
>> On 08/14/2014 07:56 AM, Carlos Garcia Campos wrote:
>>> El jue, 14-08-2014 a las 10:01 -0400, Victor Lucero escribió:
>>>> El 14-08-2014, a las 3:23, Carlos Garcia Campos<cgarcia at igalia.com>
>>>> escribió:
>>>>
>>>>> El mié, 13-08-2014 a las 15:05 -0400, Victor Lucero escribió:
>>>>>> I just read the links you provided .
>>>>>>
>>>>>> I now need to setup a extension to get access to the WebPage so i
>>>>>> can get access to the web frame an his js context.
>>>>>> thats nice but i also need access to all the frames in the current
>>>>>> page. will this be exposed in any way in the near future?
>>>>> There are no plans to expose it unless there's a real need. So, file
>>>>> a
>>>>> bug report and explain your use case there.
>>>>>
>>>>>
>>>>>> To me is very important because the kinds of tasks we produce are
>>>>>> inside the context of every frame.
>>>>> If we are talking about the WebExtensions API, things are easier
>>>>> because
>>>>> we already have a WebKitFrame object there.
>>>>>
>>>> That means that if We file a new bug asking for signal to get all the
>>>> frames may be a possibility?.
>>> Yes, file a bug report *and* explain your use case :-)
>>>
>>>> because I don’t mind writing as a WebExtension at all if its for the
>>>> sake of performance.
>>> It's not a matter of performance, it's because JavaScript context of a
>>> frame can't be accessed from the UI process, it lives in the Web
>>> process.
>>>
>>>>>> El 13-08-2014, a las 13:27, Sami Wagiaalla<swagiaal at redhat.com>
>>>>>> escribió:
>>>>>>
>>>>>>>> in the prior version of webkitgtk api(version 1) i got all
>>>>>>>> these useful signals. but in api2 they are gone :(
>>>>>>>>
>>>>>>>> does they exist inside Webkit2 and just aren’t exposed by the
>>>>>>>> Gtk api?
>>>>>>>>
>>>>>>> Most are still present. Some have been renamed, and some have
>>>>>>> been moved to the core and can only be accessed through the
>>>>>>> extension architecture; details here [1].
>>>>>>>
>>>>>>> You can ask here, or look through the documentation index [2]
>>>>>>> for the specific signals you need.
>>>>>>>
>>>>>>> [1]
>>>>>>> http://blogs.igalia.com/carlosgc/2013/09/10/webkit2gtk-web-process-extensions/
>>>>>>> [2]
>>>>>>> http://webkitgtk.org/reference/webkit2gtk/stable/index-all.html
>>>>> -- 
>>>>> Carlos Garcia Campos
>>>>> http://pgp.rediris.es:11371/pks/lookup?op=get&search=0xF3D322D0EC4582C3
>>>
>>>
>>> _______________________________________________
>>> webkit-gtk mailing list
>>> webkit-gtk at lists.webkit.org
>>> https://lists.webkit.org/mailman/listinfo/webkit-gtk
>>
>> _______________________________________________
>> webkit-gtk mailing list
>> webkit-gtk at lists.webkit.org <mailto:webkit-gtk at lists.webkit.org>
>> https://lists.webkit.org/mailman/listinfo/webkit-gtk
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-gtk/attachments/20140819/11801e0c/attachment.html>


More information about the webkit-gtk mailing list