1. Asynchronous HTTP
I'm commencing an application that will go get a multitude of web
pages. There's been a number of EU applications for HTTP requests.
I like the look of AsyncHTTP.ew. because I can get a number of
requests under way at the same time. But that leads to a limitation of
my understanding of how to process them.
>From the documentation: "Because it is asynchronous you do not simply
call a function which returns a value, you call a procedure to say
that you wish to retrieve a document and when it has been retrieved
you are notified".
That's straightforward, but if I send 10 requests, and they all come
back at about the same time, I presume my handler is going to be
interrupted up to 9 times. Do I somehow have to write re-entrant code
for the handler? Are there any guidelines on how to do this? Or have I
missed a point somewhere?
Thanks,
--
Craig
2. Re: Asynchronous HTTP
Async HTTP is based off of AsyncSelect which just sends messages to a
window, and they get processed by some functions for the http, and
then they call you. Not re-entrant the same code that gets run for the
mouse moving or the window resizing is the same kind of code for
asynchttp that you need to write.
Dan
On 5/5/05, Craig Welch <euphoria at welchaviation.org> wrote:
>
> I'm commencing an application that will go get a multitude of web
> pages. There's been a number of EU applications for HTTP requests.
>
> I like the look of AsyncHTTP.ew. because I can get a number of
> requests under way at the same time. But that leads to a limitation of
> my understanding of how to process them.
>
> >From the documentation: "Because it is asynchronous you do not simply
> call a function which returns a value, you call a procedure to say
> that you wish to retrieve a document and when it has been retrieved
> you are notified".
>
> That's straightforward, but if I send 10 requests, and they all come
> back at about the same time, I presume my handler is going to be
> interrupted up to 9 times. Do I somehow have to write re-entrant code
> for the handler? Are there any guidelines on how to do this? Or have I
> missed a point somewhere?
>
> Thanks,
>
> --
> Craig
>
>
>
>
>
3. Re: Asynchronous HTTP
On 6 May 2005, at 14:55, Craig Welch wrote:
>
>
> I'm commencing an application that will go get a multitude of web
> pages. There's been a number of EU applications for HTTP requests.
>
> I like the look of AsyncHTTP.ew. because I can get a number of
> requests under way at the same time. But that leads to a limitation of
> my understanding of how to process them.
>
> >From the documentation: "Because it is asynchronous you do not simply
> call a function which returns a value, you call a procedure to say
> that you wish to retrieve a document and when it has been retrieved
> you are notified".
>
> That's straightforward, but if I send 10 requests, and they all come
> back at about the same time, I presume my handler is going to be
> interrupted up to 9 times. Do I somehow have to write re-entrant code
> for the handler? Are there any guidelines on how to do this? Or have I
> missed a point somewhere?
I use this:
function getw(sequence filename)
sequence savefile
savefile = current_dir() & "\\wget2\\tmp.txt"
junk = system_exec(current_dir() & "\\wget2\\wget.exe -T60 -A
txt,html,htm -qO "&savefile&" "&filename,2)
junk = getf(savefile)
return junk
end function -- getw()
Modify it to not block, but to return immediately, tell it to save to different
files, poll them to see when they are done. Eu needs these things people
ask for.
Kat
