jmeter-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Philippe Mouawad <philippe.moua...@gmail.com>
Subject Re: Issues 52618 and 52674
Date Sat, 03 Mar 2012 00:14:53 GMT
Hello Sebb,
Thanks very much for taking some time to review.

Regarding docs usage update, as there is no impact at all on usage is there
something to update ?

A place to update would be the document jmeter_tutorial.pdf  linked under
"Extending JMeter".
What's the official way to update it ? Using Open Office and editing
jmeter_tutorial_mike.sxw or is there a better way ?

Thanks
Regards
Philippe

On Fri, Mar 2, 2012 at 11:28 PM, sebb <sebbaz@gmail.com> wrote:

> On 2 March 2012 12:24, Philippe Mouawad <philippe.mouawad@gmail.com>
> wrote:
> > Hello,
> > In french there is a quotation that says "Qui ne dit mot consent" :-)
>
> [That's called "lazy consensus" here, at least when applied to votes.]
>
> > Does it mean you're OK for me committing the issue
> > 52674<https://issues.apache.org/bugzilla/show_bug.cgi?id=52674>or you
> > want to take some more time to review ?
>
> Just had another look.
>
> As far as I can tell, the changes only affect the Proxy code.
> In which case, it cannot affect existing test plans; the worst that
> can happen is that the Proxy behaves differently from before.
> I don't think that would matter much.
>
> I don't object to the code being committed.
>
> Please ensure that the Proxy usage docs are updated as necessary.
>
> > Thank you all.
> > Regards
> > Philippe
> >
> > On Wed, Feb 29, 2012 at 4:26 PM, Philippe Mouawad <
> > philippe.mouawad@gmail.com> wrote:
> >
> >> Hello,
> >> To clarify a little bit what this enhancement brings:
> >>
> >> Today the Proxy can record standard HTTP sessions that are textual, see
> >> the Mail of 6 feb 2012
> >> "It was designed for recording standard HTTP sessions; these are not
> >> binary"
> >> And particularly Issue 49039
> >>
> >>
> >> Although it is limited to Textual HTTP Sessions, lot of work in it can
> be
> >> reused to record AMF Sessions , Silverlight sessions or other binary
> >> protocols.
> >>
> >> The idea behind the enhancement is to propose the following:
> >>
> >>    - Plugin implementor will be able by implementing a SampleCreator
> >>    subclass to reuse 90% of proxy feature without duplicating lot of
> code:
> >>       - All Header work
> >>       - Sampler transmission to a Target once it's created ,
> >>       - ie, all the work done in patched HttpRequestHdr which is big +
> >>       the ability to customize each of the methods in
> DefaultSamplerCreator (as
> >>       Struts Base class was build for example)
> >>
> >> Another idea is to be able to customize the created Sampler, a direct
> use
> >> I see is for example is during a recording of a JSON, GWT or REST
> >> protocols, I as a user had to go on each sampler after recording and
> switch
> >> to RAW POST BODY (which means 40 clicks for 20 samplers), with current
> >> enhancement  I can just subclass DefaultSamplerCreator and register it
> for
> >> GWT Content type and just set the property to switch it to RAW POST
> BODY.
> >>
> >> Hope it's clearer.
> >> Regards
> >> Philippe
> >>
> >> Today the Proxy feature can only be used to record HTTP Textual
> >>
> >> On Sun, Feb 26, 2012 at 9:04 PM, Philippe Mouawad <
> >> philippe.mouawad@gmail.com> wrote:
> >>
> >>> Hello,
> >>> I submitted a patch for 52674<
> https://issues.apache.org/bugzilla/show_bug.cgi?id=52674>
> >>> Hope you can have a look at it soon.
> >>>
> >>> I didn't update Tests yet, if you think it's OK then I will update
> them.
> >>>
> >>> Regards
> >>> Philippe
> >>>
> >>>
> >>> On Wed, Feb 15, 2012 at 11:30 PM, Philippe Mouawad <
> >>> philippe.mouawad@gmail.com> wrote:
> >>>
> >>>> Hello Sebb, Milamber, Rainer,
> >>>> Did you have time to look at:
> >>>>
> >>>>    - 52618 <https://issues.apache.org/bugzilla/show_bug.cgi?id=52618>
> >>>>
> >>>> Do you think patch should be applied ?
> >>>>
> >>>>
> >>>> Also I would like to have your opinion regarding :
> >>>>
> >>>>    - 52674 <https://issues.apache.org/bugzilla/show_bug.cgi?id=52674>
> >>>>
> >>>> I started an implementation, should I provide a patch or commit it
> >>>> directly ?
> >>>>
> >>>> My idea is the following:
> >>>>
> >>>>    - Either introduce 2 interfaces:
> >>>>
> >>>>
> >>>>    1. SamplerFactory with following method:
> >>>>          1. createSampler(String contentType) => Called in
> >>>>          HttpRequestHdr#getSampler()
> >>>>       2. SamplerCustomizer with following method:
> >>>>          1. customizeSampler(HttpSamplerBase sampler) => Called in
> >>>>          HttpRequestHdr#getSampler()
> >>>>          2. fillBody(byte[] rawPostBody) => Called in
> >>>>          HttpRequestHdr#getSampler()
> >>>>
> >>>>
> >>>>    - Or introduce only one SamplerProvider:
> >>>>       - createSampler(String contentType) => Called in
> >>>>       HttpRequestHdr#getSampler()
> >>>>       - customizeSampler(HttpSamplerBase sampler) => Called in
> >>>>       HttpRequestHdr#getSampler()
> >>>>       - fillBody(byte[] rawPostBody) => Called in
> >>>>       HttpRequestHdr#getSampler():
> >>>>          - Default implementation would do what is done today inside
> if
> >>>>          ((!HTTPConstants.CONNECT.equals(getMethod())) &&
> >>>>          (!HTTPConstants.GET.equals(method))) {
> >>>>          - Other protocols would handle it another way
> >>>>
> >>>> --
> >>>> Regards.
> >>>> Philippe Mouawad.
> >>>>
> >>>>
> >>>>
> >>>>
> >>>
> >>>
> >>> --
> >>> Cordialement.
> >>> Philippe Mouawad.
> >>>
> >>>
> >>>
> >>>
> >>
> >>
> >> --
> >> Cordialement.
> >> Philippe Mouawad.
> >>
> >>
> >>
> >>
> >
> >
> > --
> > Cordialement.
> > Philippe Mouawad.
>



-- 
Cordialement.
Philippe Mouawad.

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message