[WEB400] Re: QtmhRdStin and QzhbCgiParse Limitation

[Hans Boldt <boldt@xxxxxxxxxx>]

Bartell, Aaron L. (TC) wrote:

> <Hans>
> I was surprised it works because of problem #1 above. I would expect 
> that you shouldn't be able to read 11M of data into a 32K variable, 
> unless you're overwriting the subsequent locations in memory. You 
> may be just very lucky that you're not overwriting something 
> important. I think you still need to dynamically allocate all the 
> storage you need and read into that.
> </Hans>
>
>
> Actually I was assuming that QtmhRdStin was doing proper allocation for me.
> Notice what I am passing in as the second parm to qtmhRdStin:
> qtmhRdStin(output: contentLength: bytesRead: ErrorDs);
>
> Variable 'contentLength' has a value of 11,000,000 in my 11MB scenario.
>
> Here is the documentation for QtmhRdStin:
> http://publib.boulder.ibm.com/iseries/v5r2/ic2924/index.htm?info/rzaie/rzaie
> api_qtmhrdstin.htm 
>
> Given that information don't you think it should work?  

What specifically about the documentation for "QtmhRdStin" would 
lead you to believe that it would allocate the storage for you? Note 
that you're passing in the receiving field by reference, and so the 
data should get written to that variable. You're telling the API by 
the value in the second parameter that the receiving field is big 
enough for that amount of data, and the API has no reason to suspect 
you of lying to it.

Cheers! Hans