Re: Library list for IFS folders?

Ok.. I'll see if I can make the picture a little clearer.

We built a product inquiry screen (intranet only). Screen1 prompts for
SKU and store.  Screen2 displays the image, basic info, qty on hand,
current promotions, rebates and finance offers for the SKU.  Each promo,
rebate and finance description is a link to other screens to view more
details about that promo, rebate, finance offer.
The legacy code for each screen/page has a separate ile/free pgm that
creates the html code line by line (I didn't create this, I inherited
it)  
We're converting this into CGI based code where the ile/free code exists
in qrpglesrc and the html src is in the root file system as a .html
file. (compiled objects are in DEVLIB and/or QALIB and/or PRODLIB)
When loading the html page with data we're using the following code: 
   GetHtmlIfs('/compcgi/dev/cgihtml/CGI_SKU411a.html');
      UpdHTMLvar('STRno':STRno);
      UpdHTMLvar('out_strname':out_strname);
      UpdHTMLvar('out_sku':out_sku);
      UpdHTMLvar('out_skudesc':out_skudesc);
      UpdHTMLvar('out_qoh':out_qoh);
      UpdHTMLvar('out_price':out_price);
      WrtSection('Page1');                                       

The 'dev' in the path name will be replaced with 'QA' or 'prod'
depending on which environment it was called from.  Normally there
should only be a copy of the .html file in dev if we're doing current
work on it, or a separate copy in qa if they are currently testing
changes we've made.  If neither is the case, then the only copy of that
.html file should be in production.. but I'm running this page from DEV
to test another page in the stream and so my path is defaulting to DEV
and it crashes because it doesn't look for and find the copy in PROD.
If this were a normal 'CALL' command, the system would look for the
program in the libraries listed in your library list - but we don't have
that built in ability in this case. 

So I'm looking for a way to work with this situation.  Do I have to keep
a copy of every object in every environment so it will only have to look
in the current directory? Should I search for an object through multiple
paths in a predetermined order and use the path of the first copy I
find?

Help! 



-----Original Message-----
From: web400-bounces@xxxxxxxxxxxx [mailto:web400-bounces@xxxxxxxxxxxx]
On Behalf Of Matt.Haas@xxxxxxxxxxx
Sent: Tuesday, February 20, 2007 8:44 AM
To: web400@xxxxxxxxxxxx
Subject: Re: [WEB400] Library list for IFS folders?

Charlie,

The root file system (which is what most people incorrectly call the
IFS, the IFS includes all file systems. QSYS is an example of a file
system to library lists do apply) does not have a library list. Having a
little more information about how you are using the stuff in the root
file system will help with getting suggestions.

Matt

-----Original Message-----
From: web400-bounces@xxxxxxxxxxxx [mailto:web400-bounces@xxxxxxxxxxxx]
On Behalf Of Charlie Ramsdale
Sent: Monday, February 19, 2007 12:27 PM
To: web400@xxxxxxxxxxxx
Subject: [WEB400] Library list for IFS folders?

We have one system that shares 3 environments - dev, QA and production,
each with its own copy of data files and possibly different versions of
programs depending on active dev and testing (we use Turnover to
promote). Accessing the correct copy or the production copy of a program
or file is easy with the library list.  When a user runs a pgm he'll hit
the dev or qa version (depending on liblist), if there is no dev or qa
version, then it will eventually find the production version.  Now we're
adding CGI development with the HTML code in the IFS folders and we've
abruptly discovered that we know of no 'library list' functionality with
objects in the IFS folders.  We obviously don't want to hard code paths
and maintaining a soft version of paths to all objects for the 3
environments seems ludicrous.  Is there a 'library list' style access
for IFS folders that we're not aware of or maybe some other method that
we as noobies need to be aware of before we go too far in our current
development?

 

Charlie

 

PS: We do have separate browser paths for the 3 environments so we know
which environ the user is running the program(s) from, but that limits
access to objects in that environment and it will not find an object in
the production environ when needed (unless the path is provided).