URL download and Cache problems

[Alex Tweedly]

At 12:28 15/12/2004 -0500, Rick Harrison wrote:

>Hi there,
>
>I'm getting some Revolution weirdness when trying to download
>files from the internet.

I don't think there's a real problem there, but I'm not sure, because it's 
hard to follow the description / script to be sure I'm correctly 
interpreting what you want.

The basic issue is that the "load" command is non-blocking. So when you do 
the first "load URL" it does not (and should not) wait. It will accept any 
number of load commands, and queues them up (it probably fetches multiple 
of them in parallel, with some limit on how many it has under way at a time).

>I'm using:
>
>put "" into field "CachedURLSListField1" of card 1
>put "" into field "ClearedCacheList1" of card 1
>
>put 1 into N2
>repeat while N2 < 4
>load URL field "JPEGURL" of card 1 with message "downloadComplete"
>--(Which sends this message to the stack which I handle by updating a 
>field with "Status - Download completed")
>
>   export image "JpegImage" of card 2 to file field "ImageFileName1" of 
> card 1 as JPEG -- Gets entire image
>   put field "CachedURLSListField1" of card 1 & the cachedURLs into field 
> "CachedURLSListField1" of card 1
>
>--to list what files are in my cache
>--then I supposedly delete the cache with
>
>   unload URL field "JPEGURL" of card 1
>
>--and to check if the file was deleted out of the cache
>
>   put field "ClearedCacheList1" of card 1 & the cachedURLs into field 
> "ClearedCacheList1" of card 1
>add 1 to N2
>end repeat
>
>The problems I'm running into are the following:
>
>1.  The program doesn't wait until the first download is complete before 
>moving on to the second download.

That's good :-)

>2. The cache isn't getting cleared

Perhaps because you're clearing it (with unload) before that URL has 
finished being loaded; perhaps because there are multiple in-progress 
fetches, and you can never clear one without the next one appearing.

>3. It's like Revolution is executing just too fast.

The purpose of the load command is to act fast, and to allow your stack to 
continue with other work while the URLs are being pre-fetched. If you go 
ahead and use the URL before it has completely loaded, you'll still get the 
benefit.

>If I simply put
>
>answer "N2 = " & N2
>
>-- before the end repeat everything downloads - but unfortunately, -- I 
>have to be around to press the stupid button everytime to -- let the 
>program move on, and I don't want to have to do that.
>
>Any ideas as to what is going on, and as to what I need to do to fix things?

It's not clear just what you want to do, overall. It looks to me as though 
the loop is loading the same URL each time round - or am I missing something ?

You might want to do something like

local lNumberDone

on gettheURLs
   put 0 into lNumberDone
   repeat while N2 < 4
      load ......
   end repeat
   send checkifdone in 50 millisecs
end gettheURLS

on checkIfDone
   if lNumberDone = 4 then
      ....finished ...
   else
      send checkIfDone in 50 millisecs
   end if
end checkIfDone

on downloadComplete pURL, pURLStatus
   if pURLStatus = "Download Complete" then
       put URL pURL into (wherever you want it)
   else
       put "problem with " && pURL && ":" && pURLStatus & CR after field 
"myStatusfield"
   end if
    add 1 to lNumberDone
end downloadComplete

In summary - load is non-blocking.
If this hasn't helped - send more of the script, or a fuller description of 
what you want.

-- Alex.