| Author |
Message
|
| MrSmith |
 Posted: Wed Apr 30, 2008 7:09 am Post subject: MRM guidance |
 |
|
Master
Joined: 20 Mar 2008
Posts: 215
|
i have been looking throught the manuals and online help and i am confused and can't find what i am looking for so any help or links to an examples especially would be gratefully received.
This is what i want:
it is an outgoing MRM message which is one long data stream but is made up of mutltiple occurrences of say 500 bytes for max 9999 occurs. these 500 bytes are formed from the individual elements defined under each occurrence. i get the creation of the individual elemtnts and have created the complex type to sequence them underneath so hence my 500 byte stream is now Ok but need some examples on creating the next tier for the repeats. i have tried some arrangements but they either don't work and deep confusion and frustration now resides 
Unfortunately some of the message model PDF's that i thought might help are iunavailable to me as they are on IBM's ftp server and i am blocked from ftp sites. thanks |
|
| Back to top |
|
 |
| madi |
| Posted: Wed Apr 30, 2008 7:13 am Post subject: |
|
|
Chevalier
Joined: 17 Jan 2006
Posts: 475
|
|
| Back to top |
|
|
| MrSmith |
| Posted: Wed Apr 30, 2008 7:21 am Post subject: |
|
|
Master
Joined: 20 Mar 2008
Posts: 215
|
Madi
No not blocked from this and have had a look through, and although its for V6 not V5 as stated its samples section doesn't go anywhere i.e ther page is not available. I have read the whole message section model etc and there is an explaination for most of the parameters but what the manuals don't show that i ahve seen is some examples of the structure i explained i.e where the occurences would be set for example or wheat level the delimiter is set etc etc just was lloking for somthing to relate to rather than interpretation. thanks tho. |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Apr 30, 2008 7:23 am Post subject: Re: MRM guidance |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| MrSmith wrote: |
i have tried some arrangements but they either don't work and deep confusion and frustration now resides
|
Either don't work or what? 
Can you be a bit more specific than "don't work"? Don't produce a message tree as expected, don't parse at all, throw a broker error, what?
Have you taken a user trace and/or added Trace nodes to the flow? This could give valuable information on any discrepancies between what you believe you've defined in the message set, and what the broker believes you intended with the definiton?
If nothing else, such a trace will provide us with valuable information if posted.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| madi |
| Posted: Wed Apr 30, 2008 7:26 am Post subject: |
|
|
Chevalier
Joined: 17 Jan 2006
Posts: 475
|
just trying to give you something to try until one of the gurus step in
you can import samples within the toolkit ........... go to help and then Samples gallery ....... the samples listed in the help page are usually in there
good luck 
--madi
_________________
IBM Certified Solutions Developer - WMB 6.0 |
|
| Back to top |
|
|
| MrSmith |
| Posted: Wed Apr 30, 2008 7:43 am Post subject: Re: MRM guidance |
|
|
Master
Joined: 20 Mar 2008
Posts: 215
|
| Vitor wrote: |
| MrSmith wrote: |
i have tried some arrangements but they either don't work and deep confusion and frustration now resides
|
Either don't work or what?
Can you be a bit more specific than "don't work"? Don't produce a message tree as expected, don't parse at all, throw a broker error, what?
Have you taken a user trace and/or added Trace nodes to the flow? This could give valuable information on any discrepancies between what you believe you've defined in the message set, and what the broker believes you intended with the definiton?
If nothing else, such a trace will provide us with valuable information if posted. |
Vidor - the number of occurences depends on the incoming cardinality of an xml message element hence unknown so when tried with cardinality of 1 then message is fine but i am not 100% sure where i need to set the multi occurences etc
so what i have is complex element (call it COMPB) with my children that make up my 500bytes and then set that local complex type COMPB to have max occurs of 9999 and min of 1 of seperation type fixed length which makes sense to me. then i created another complex type COMPA which referenced the complex type COMPB as an element. I set COMPB to be kin 1 max 1 and the anonymous bit underneath to be all elements delimited and a delimiter of <CR><LF> this to me says i have a repeating element under my message type with a delimiter of CRLF and within them repeats is a fixed length element made up of children that consrtuct my 500 bytes. the validation of the set says i "A structure must have seperation of all delimited elements when the parent data seperation is fixed length". the physical format is TDS and the error is confusing at best as to me the parent is the all delimited and the structure underneath is fixed opposite to what the actual errormessage says, hence was looking for examples. in my mind it makes sense in construction it obv doesn't. Strange thing is there is another message prted from 2.1 that has similar setup to create similar strucure and is fine, altho have treated as red herring. |
|
| Back to top |
|
|
| kimbert |
| Posted: Wed Apr 30, 2008 2:19 pm Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5542
Location: Southampton
|
| Quote: |
| just trying to give you something to try until one of the gurus step in |
OK, I'll try to be a guru. I'll probably step right in it, as well
Here's my understanding of the requirement:
- You're trying to build an output message which consists of lines of data.
- Each line is composed of fixed length fields with a total length of 500 bytes ( or possibly characters, though that may be the same thing in your world )
- Your input records are in XML, Each occurrence of the record translates to one line of output.
Here are my suggestions:
- Don't bother with the All Elements Delimited type. Just set the Group Terminator on the fixed-length complex type to <CR> or whatever you need.
- Do consider reducing heap usage using the technique outlined here: http://www.ibm.com/developerworks/websphere/library/techarticles/0505_storey/0505_storey.html
I'm a bit puzzled about the error that you received as well. Maybe it's a defect in the v5 toolkit. Either way, the suggestion above should take it out of play. |
|
| Back to top |
|
|
| MrSmith |
| Posted: Thu May 01, 2008 2:05 am Post subject: |
|
|
Master
Joined: 20 Mar 2008
Posts: 215
|
Kimbert firstly thanks for steepinginto the breach
Ok so let me make sure i get you right i create lets say for example 15 elements simple types which represent my fields (already done) then create one complex type that "lists these elements underneath it creating my 500 bytes of data. Question here should the complex type be local or global? apart from re-use is there any advantage to either specifically?
OK - so on my complex type i then have 1 min occurs and 9999 max occurs and set the group terminator thus
| Quote: |
| Just set the Group Terminator on the fixed-length complex type to <CR> |
. then inside my compute node just create the fields of the message with a index counter and that should be it it will give me x number of 500 byte data with a terminator all in one "message" length?
apart from the terminator bit i think i already have this but am willing to try something else to resolve this.
I know IBm pu some examples in the manuals but a selection of real life examples full blow in the manuals would save some of this forum space I think 
thank you. |
|
| Back to top |
|
|
| kimbert |
| Posted: Thu May 01, 2008 2:24 am Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5542
Location: Southampton
|
| Quote: |
| should the complex type be local or global? apart from re-use is there any advantage to either specifically? |
Doesn't matter. Your choice.
| Quote: |
| just create the fields of the message with a index counter and that should be it it will give me x number of 500 byte data with a terminator all in one "message" length? |
Try it and see! |
|
| Back to top |
|
|
|
|
