ANNOUNCE: ecap-tcl v 0.1

Georgios Petasis petasisg at yahoo.gr
Mon Dec 12 14:39:15 UTC 2016


On 11/12/2016 23:49, Alex Rousskov wrote:
> On 12/11/2016 07:17 AM, Georgios Petasis wrote:
>> With ecap-tcl you can adapt content with code written in Tcl.
>> https://github.com/petasis/ecap-tcl
>
> An eCAP Tcl wrapper is a valuable contribution, thank you! If you do not
> mind, I will add a link to your Github project from the e-cap.org site.
Of course, please add the link.
>
> FWIW, it was not clear to me what the current limitations of the wrapper
> are: Can Tcl adapters utilize all APIs offered by the official C++ eCAP
> library? Please consider documenting that if you have not already.
Right now, adapters written in Tcl can receive 5 "events": |wantsUrl, 
||actionStart, |
|contentAdapt, ||contentDone, ||actionStop. From these they can 
see/change/remove headers.
|
>
> For example, it was not immediately obvious to me whether the Tcl
> wrapper already supports asynchronous message analysis documented in
> libecap/doc/async.txt and illustrated by the official adapter_async
> sample adapter.
I am sorry, I was not aware on these resources. When I started (some 
months ago), I found a sample adapter that was changing some strings in 
the content (had a victim and replacement). I based ecap-tcl on this.
So, no it does not support asynchronous message analysis.
(I have used the packages for Fedora 23/24, and the doc folder is not 
there).

What I tried to do, is to start a number of Tcl threads, and try to 
distribute load to these threads, ensuring that all requests related to 
the same action, end up in the same thread.

I will try to read the new sample adapter, and see if I can update ecap-tcl.

BTW, here is a ecap-tcl in action:
http://s864.photobucket.com/user/petasis/media/ecap-tcl/ecap-tcl-in-action.png.html
It applies some linguistic processing on pages using Tcl code...

George

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.e-cap.org/pipermail/users/attachments/20161212/32c64d3c/attachment.html>


More information about the Users mailing list