IF data fetch produced multiple values

[lisa sater]

Does anyone have any tcl that I can plug into the Xlate to get an idea of which of the 40 if statements is the one having an issue with producing multiple values?

Thanks!

Lisa

Topic

[Tom Rioux]

Lisa,

One trick that I use is to put an echo statement in my xlate.  I put it in near the top and if I see the results of the echo statement, then I push it down a little further in my xlate until I find the culprit.

Also, look at your IF statements and look for any that may not have the subfield designation on it (i.e. ..PID.#3.[0]).   Those can be potential problems.

Hope this helps….

Tom Rioux

[Max Drown (Infor)]

Also, try copying the xlate to a test file and then removing pieces of the code until you narrow it down.

-- Max Drown (Infor)

[Jim Kosloskey]

Lisa,

I have used both of the techniques described by Tom and Max.

For echoing, I have a small proc that’s only function in life is to echo out what is in the inbound pane of the COPY Action. It has a switch to tell it to echo or not.

I use that a lot when I have complex nested ITERATEs and/or IFs. By having the switch, I can leave the proc in place and turn the switch off for normal execution and then when I need to make that next modification, I don’t need to remember where to lay my traps I just turn the switch back on at the appropriate places for the echoes I want.

The binary technique Max describes is made a little easier to deal with using the Insert File capability in the Xlate.

I think the ability to better cut and paste in Cloverleaf(R) 5.6 will make the binary technique easier to do as well.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Tom Rioux]

Another way instead of removing code, if you are on 5.6, just go down the line and deactivate/reactivate your IF statements until you find the one that is the culprit.  Then, once you have found it, reactivate it and make your corrections.  No need for copying, pasting, inserting or removing anything…including echo’s

Hope this helps….

Tom Rioux

[Chris Williams]

That IF data fetch error is coming from insufficient granularity in your address for the given data. One notorious culprit that we see in our organization is the use of an ampersand in an address or name field as part of the text without considering that it is a separator character (e.g., A & W Root Beer. 4th & Main St, etc). You could also have a repeating field and just be specifying the entire field instead of the specific repetition.

[Jim Kosloskey]

Tom,

I like that new feature in 5.6 – just getting started playing with 5.6 so I forgot about that.

Anyway Lisa, lot’s of ways to narrow it down – but no magic bullet.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Max Drown (Infor)]

A note on “deactivating” if statement. I think when you deactive the IF and not the statements in the body of the IF, the statements are still executed. You have to deactive each statement within the IF body, too. I think …

-- Max Drown (Infor)

[Tom Rioux]

I just tried to deactivate an IF statement.  It grays out the IF statement along with all of the statements included with the IF.

Tom Rioux

[Rob Abbott]

Tom is correct.  If you deactivate any node that has statements underneath (IF, iterate, etc) all the statements nested underneath are deactivated as well.

Rob Abbott
Cloverleaf Emeritus

[Max Drown (Infor)]

I know it shows as deactivated, but I think the statements in the body are still executed when a message flows through. I definitely could be wrong, but I think I ran into that problem awhile ago.

-- Max Drown (Infor)

[Tom Rioux]

Max and Rob,

I just did a simple test.  I had an IF node in an xlate with one statement underneath.  When I deactivated the IF node, the statement below it did not get executed.

Hope this helps…

Tom Rioux

[Rob Abbott]

Max Drown wrote:

I know it shows as deactivated, but I think the statements in the body are still executed when a message flows through. I definitely could be wrong, but I think I ran into that problem awhile ago.

If you can reproduce this please let us know.  That does not sound like it is working as designed.

Rob Abbott
Cloverleaf Emeritus

[Max Drown (Infor)]

I tried to reproduce it, but I can’t. Seems to be working fine. It might have been some weird situation or more likely I was mistaken.

-- Max Drown (Infor)

[Femina Jaffer]

Hello everyone,

I encountered the exact same error, and it happened to be exactly what Chris suggested – an AMPERSAND in the field (OBX 5).  The questions is how to I fix this to avoid this error and make the IF work?

All I want to do is check if OBX 5 is NULL, IF SO THEN REMOVE THE ENTIRE OBX SEGMENT:

e.g.  IF 1(0).1(0).1(%g1).OBX.00573.[0] eq @null

       PATHCOPY -> 1(0).1(0).1(%g1).OBX

VALUE FOUND IN OBX5 – “for Ova & Parasite Exam “

Thank you,

Femina Jaffer

[John Mercogliano]

Femina,

 In an If statement you can narrow your search to

Code:

1(0).1(0).1(%g1).OBX.00573.[0].[0]

The extra .[0] takes into account the extra subfield that might exists.

John Mercogliano
Sentara Healthcare
Hampton Roads, VA

[Femina Jaffer]

Thank you, John.  I will give this a try.

Thanks again.

Femina

[Ian Morris]

John’s fix did not work for me.  I’m looking through OBX segments for “PROCEDURE: …..           DESC:…..”.  I’m doing it with a Xlate if-statement:

Code:

(IF) 1(%g1).OBX(%s2).#5(0).[0] ct=PROCEDURE: && 1(%g1).OBX(%s2).#5(0).[0] ct=DESC:

The problem is that I get a “multiple values” error if I hit a segment with a “&” symbol.

Any tips on how to address this problem?

[Gary Atkinson]

add in pre-proc before translation and change the & to T.  Run message through translation.  In post-proc change T back to &.  I use a table to make this easier to maintain.  table of segs and fields to check for “&”

[Ian Morris]

Gary Atkinson wrote:

add in pre-proc before translation and change the & to T.

[Gary Atkinson]

You could also try to go back to the vendor and tell them “hey your not following the standard” and see where that gets ya.  I have been down that route before with no luck.  I have attached the proc I have used before when I did xlate work, its arg driven (use at your own risk, yada yada yada…) 8)

GL

[Ian Morris]

That is helpful.  Thank you for the proc.

[Gary Atkinson]

Not a problem.  Good Luck!

[Author]

Replies