| Author |
Message
|
| shalabh1976 |
 Posted: Thu Oct 18, 2007 12:47 pm Post subject: Message Set creation question |
 |
|
Partisan
Joined: 18 Jul 2002
Posts: 381
Location: Gurgaon, India
|
I wanted to use TDS to create a message with the following structure:
Row 1 Length 5 n occurances
Row 2 Length 10 m occurances
Row 3 Length 6 x occurances.
n,m,x > 1 and unbounded.
All lines end with <CR><LF>
I know that I can't use Tagged delimited because the length is the only parameter to distinguish Row 1,2 or 3.
Using fixed length wants a value on MaxOccurances other than -1
I can only think of variable length delimited and all elements delimited.
But their usage is not working either.
I know I am setting some properties incorrectly or looks like this cannot be done without using a data pattern.
I am sure this was doable on 2.1
My current environment is V5.
Please advise
_________________
Shalabh
IBM Cert. WMB V6.0
IBM Cert. MQ V5.3 App. Prog.
IBM Cert. DB2 9 DB Associate |
|
| Back to top |
|
 |
| jefflowrey |
| Posted: Thu Oct 18, 2007 12:52 pm Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
There's an option somewhere for "repeat til end of bitstream". That's what you want, I think.
Unless you mean that n is unbounded AND m is unbounded AND x is unbounded? That's probably going to be tricky.
Or is is that there will only be three <cr><lf>'s in there?
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| shalabh1976 |
| Posted: Thu Oct 18, 2007 12:54 pm Post subject: |
|
|
Partisan
Joined: 18 Jul 2002
Posts: 381
Location: Gurgaon, India
|
Jeff,
All m,n,x are unbounded.
Wasn't end of bitstream somewhere in CWF? and not TDS?
_________________
Shalabh
IBM Cert. WMB V6.0
IBM Cert. MQ V5.3 App. Prog.
IBM Cert. DB2 9 DB Associate |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Thu Oct 18, 2007 1:00 pm Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
I think you'll have to use data patterns for this. But again, you may have issues if you have [row1<cr><lf>][row1<cr><lf>]...[row2<cr><lf>]...
Because there's not going to be anything that can tell the parser that row 1 has stopped repeating and row 2 has started.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| shalabh1976 |
| Posted: Fri Oct 19, 2007 4:25 am Post subject: |
|
|
Partisan
Joined: 18 Jul 2002
Posts: 381
Location: Gurgaon, India
|
Shouldn't the length of each line help me in distinguishing that?
But I think the parser needs to know before hand what kind of data it is handling and doesn't get around to determining an element type from its length.
_________________
Shalabh
IBM Cert. WMB V6.0
IBM Cert. MQ V5.3 App. Prog.
IBM Cert. DB2 9 DB Associate |
|
| Back to top |
|
|
| shalabh1976 |
| Posted: Fri Oct 19, 2007 5:22 am Post subject: |
|
|
Partisan
Joined: 18 Jul 2002
Posts: 381
Location: Gurgaon, India
|
I haven't used data patterns in MB before.
Is there a tutorial or something available?
I need to be able to map the length though as each row can be uniquely identified that way only.
_________________
Shalabh
IBM Cert. WMB V6.0
IBM Cert. MQ V5.3 App. Prog.
IBM Cert. DB2 9 DB Associate |
|
| Back to top |
|
|
| kimbert |
| Posted: Sat Oct 20, 2007 2:54 am Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5542
Location: Southampton
|
| Quote: |
| the length is the only parameter to distinguish Row 1,2 or 3. |
You can set Data Element Separation to 'Use Data Pattern' and use a data pattern something like '.*(100)\r\n' where the 100 is the length of the line.
This assumes, of course, that each row type has a fixed length.
By the way - this is very strange message format. Why doesn't the source application identify the lines with a tag? What application is your flow replacing, and how does that application parse these records? |
|
| Back to top |
|
|
|
|
