Forum Replies Created
-
Posts
-
The problem still remains, I can't access the values of the constructed message though they will appear as expected if I write them to disk. I tried to assign the values to the Orchestration variables in the message assignment, in an expression shape right after the message construction, I tried to use the values directly when calling the helper method. I also tried to write them to the debug console and to the event viewer…
They are always "Nothing" but if i send them thorugh the pass through pipeline and write them to disk the values are in there … Did anyone encounter a similar problem?
Adapter is same in both versions.
Seems ur missing somewhere.
Check what the error says.
If it is "Subscription failure" error then check whether you are dropping the right xml file.
Maybe this power point from Todd Van Nurden will help?
http://download.microsoft.com/download/2/4/6/24642d74-6ab0-41b9-9966-29dd8a46fcba/CON411.ppt
A number of blogs talk about a BAM Pipeline Interceptor being a new feature in BizTalk 2006. From http://geekswithblogs.net/synboogaloo/archive/2005/06/02/41562.aspx:
"No longer need to write custom pipelines components to call the BAM API a BAM pipeline interceptor is supplied out of the box"
Would this BAM pipeline interceptor solve this problem above?
Chris,
if att1 = "mostfavorite" them map att2 to <loveit>
if att1 = "leastfavorite" them map att2 to <hateit>
use the value mapper and logical test like I showed you earlier
– wa
George,
Let's start with this – To answer #1, BizTalk is writing the messages to the Message Box db in BizTalk. From what I've seen (and I'm no expert in BT) it consumes the input files first before it does any writing. In my app I ran into the same problem as I had been testing with batches of 20-100 orders and everything was cool. I fed it an XML with 5800 orders and it went into a coma.
You don't indicate how much ram you have, but our test server has 4gb and that makes all the difference in the world compared to my dev system with 2gb. Even though it takes longer, on the development side I will test with small batches on the dev side, then migrate the whole mess to the test server to test the big batches. Other than pegging the CPU it does reasonably well.
Hope this helps.
[quote user="Hermit"]
Hi,
I have a problem with a flat file that I need to disassemble into separate XML files. The problem is I need to do this without tagging each record as I have no control over the format fo the incoming data as its from a legacy system.
A sample of the file in simplified format is as follows:
ID12052007
Mr Jo Bloggs M
Mrs Jane Bloggs F
Mr Sam Smith M
3The header contains a Tag + the data, the body a series of fields followed by CR/LF and the footer a record count.
I have used a pipeline with a flat file disassembler and have generated three schemas, one each for Header, Body and Footer. I have set Max occurs on body to 1.
If I use a tag before each row of data everything works happily, and I have tried delimted records with a tag and that works too. What I need however is to do this without a tag – any ideas would be very welcome. A lot of samples come close to this but nothing does it exactly.
To be honest I don't need the Header and Footer for processing, just the Body so a solution that looses the extra data is fine.
Thanks for any help in advance.
[/quote]
Herman,
The key to your solution is that you said "I don't need the Header and Footer for processing, just the Body so a solution that looses the extra data is fine."
To me this means that if you parse the file and map from the body you are fine.
What you want to define is one schema containing:
a header record with a count of 1
a detail record, count 1 to *
a trailer record with a count of 1Then just map from the detail record.
You don't need tags, as there aren't any in the data and the structure is enough to parse the file. Also, it seems like you don't to parse to three XML files.
-wa
Seems u drink lots of Beer.
I find "Cheers" in each of your statements.
Cheers John…. 😉
Let me know if u need more from me.
Cheers Gents.
[quote user="Anonymous"]I have a map, where there is an attribte on the left side ("middleinitial") which may or may not be populated. I want to conditionally create the attrubute on the right, depending on whether or not the one on the left's length is > 0.
The "middleinitial" attribute on the right has two siblings, "firstname" and "lastname" which are always populated (makes sense). I just want to create the thrird attribute, only if the one on the left exists.
Any thoughts – seems straight forward.
Any help would be greatly appreciated.Chris
[/quote]Chris –
1. link the middle initial to a maroon "length" functoid [###}]
2. link the output of length finctoid to a green [>] "greater than" functoid, and add the second parameter of "0"
3. link [>] functoid to a purple [=>] value mapper functoid.
4. link middile initial to the valume mapper as the second input parameter
5. link the output of value mapper to the destination for middle initial.
This will cause the output attribute to only be created when the value mapper gets a true, meaning the length of the middle initial is greater than 0
-wa
