hi Mike,

Mike wrote:
WriteResponse('Content-Type: application/xml' + cr);
WriteResponse('Content-Length: ' + %char(%len(p_UsrSpcData)));

You shouldn't be providing 'content-length' -- that part should be taken care of by the HTTP server. Apache might decide to used identity encoding as you've provided, or it might decide to use chunked encoding. In either case, how it works is up to Apache, not you. Don't provide content-length like this. (Worst case scenario, it might pass that keyword on to the browser, and then it'd be unpredictable...)

You didn't say what 'cr' is defined as. But, it needs to be a linefeed (LF), or else a carriage return/linefeed (CR/LF) pair. It cannot be just a carriage return as the variable implies.

Finally (and this is the big one) you need to have a blank line (a line with just LF, or just CRLF) to separate the keywords like content-type from the data. You're missing this altogether, and that's the likely cause of your 500 error. Apache will think that all of your data is HTTP keywords, and none of it is actual data.

Those aren't your only errors, however... there's also a more elusive and dangerous one here:

P WriteResponse B
D WriteResponse PI
D buffer 32767A options(*varsize) const

/free

QtmhWrStout(%trim(buffer):%len(%trim(buffer)):ErrCode);

You have 'buffer' defined with options(*varsize). That tells RPG that it's okay to pass a shorter variable here... and that could be an absolute disaster, because your code does NOTHING to determine the actual length passed. It just assumes that all 32767 will be passed every single time.

The frightening part is that this will work a lot of the time... maybe even most of the time. You won't detect this sort of bug by testing. Yet in production, your code might output tons of garbage, or might crash.

If you don't fully grok options(*varsize), then I strongly recommend that you don't use it. In this case, your code was written as if you hadn't used it. So just remove it, and keep the code like this:

P WriteResponse B
D WriteResponse PI
D buffer 32767A const

/free

QtmhWrStout(%trim(buffer):%len(%trim(buffer)):ErrCode);
return;

This'll work fine. It's not especially efficient (since you're trimming 32000+ blanks TWICE every time you run this routine... what a waste of CPU cycles!) but it'll work, and it'll be safe.

A better method is to use a VARYING variable instead of a fixed-length one so no blanks need to be trimmed...

As an Amazon Associate we earn from qualifying purchases.

This thread ...

Follow-Ups:
Replies:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].

Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.