| Wen Chi 2005-01-28, 2:46 am |
| Thanks for your reply in advance. I posted some of
questions before. But didnt get a clear answer.
1. Files with line terminator 0D0A can't go through map
We converted BTS2002 maps to BTS2004 maps. When we tested
out the converted maps, we encountered a problem with line
terminators. As the messages we process have line
terminator 0D0A, we had set the line terminator as 0D0A in
the map; but translation using this map failed (which
shouldn't be the case).
Please advise how we should configure a map such that
messages with line terminator 0D0A can be successfully
translated using the map.
Please also advise if there are specific settings in the
schema that must be present for such a translation to be
successful.
//For your information
BTS02-0D Lineterm
BTS04-0D Lineterm BTS04-0D0A Lineterm
Files with 0D0A Can go
through Can't go through
Can't go through (!)
Files with 0D (not
tested) Can go through (*)
(not tested)
Files with 0A (not
tested) (not tested)
(not tested)
2. Partial output for translated files
In one of my implementations, I had associated a receive
location (file adapter) with a map. I tested it out and
had the following finding:
When I submitted one file at a time, the output was
complete.
When I submitted multiple files at a time, I found that
the output was complete for only some of the files.
I suspect that this phenomenon occurs when the map
contains a DB lookup functoid (because for those partial
outputs, the output is truncated from the place where a DB
lookup value was to have been populated) - though I cannot
be sure that that is the cause of it.
3. Some script functoids not working when placed in
maps converted using map conversion tool.
I had converted a map from BTS2002 into a map usable with
BTS2004 using the map conversion tool. I then added
script functoids bound to external assemblies (created by
wrapping VB6 dlls), as well as script functoids
implemented using inline C# code into the converted map.
When I tested the map out, I got the following findings:
Case 1: External assembly set to process first
Only the script functoids bound to external assemblies
worked - those that were implemented using inline C# code
did not work.
Case 2: Inline C# code set to process first
Only the script functoids implemented using inline C# code
worked; those that were bound to external assemblies did
not work.
I would have expected that in both cases, both types of
script functoids should have worked.
4. Behaviour upon validation failure of message
containing multiple datasets
I have a implementation for which messages arrive in
files, and each file can have multiple datasets. When I
pass the file into a receive pipeline and validate the
datasets contained therein with a schema (for a single
dataset), each dataset is individually validated, and we
get an output for each dataset. This is correct.
Consider a file with 10 datasets. The 5th dataset has a
problem (e.g.. missing a mandatory element) that would
cause it to fail schema validation (the other datasets are
fine). When I tested out such a file, what I found out
that I got outputs for the first 4 datasets only (so
presumably the processing of the file halted once the
error with the 5th dataset was encountered).
I suppose this is reasonable behaviour. However, I note
that in our implementation in BTS2002, if we encountered
such a problem file, I would get outputs for 9 datasets
(i.e. output for all the datasets other than the 5th).
Please advise: Is there anyway we can make the BTS2004
implementation behave in the same way?
5. Changed schema doesn't take effect if placed in
same project as flat file pipeline.
I had used a flat file pipeline in one of my
implementations. I had associated the flat file pipeline
with a schema, and placed the pipeline and the schema in
the same project.
Subsequently, I had to make a change to schema. After I
saved the change and tested it out by submitting a file to
the pipeline, I found that the change that I had made had
not taken effect. I then tried to set and commit the
schema property of the pipeline again - the change still
did not take effect.
In the end, I found that the only way I could get the
change to take effect was to put the schema in a project
that was separate from the project that contained the file
pipeline.
Please advise: Is there anyway of me making a change to
the schema and still keeping it in the same project as the
file pipeline?
|