XLT Copy woes

[Seth Jackson]

Hey everybody,

I’m working on a HL7 translation that has me stumped.  Here’s a pseudo code of, what I’m attempting to accomplish:

Copy all components starting from the source’s OBX.#5.[1] to the destination, starting at the destination’s OBX.#5.[0]

Here are examples of how that should look:

|=^20.2| would copy as |20.2|

|=^1^+| would copy as |1^+|

|=^20^:^06| would copy as |20^:^06|

In other words, I don’t want to erase or @null the data in OBX.#5.[0], rather I need to shift the data… So OBX.#5.[1] becomes .[0], .[2] becomes .[1], .[3] becomes .[2], and so on.

The issue I’m struggling against is, I don’t necessarily know how many components will be in the source field. But since xlateOutVals is a list, I have to specify multiple destinations, right? Do I just guess by putting .[0] through .[99] in the destination, and hope for the best? 😆

It just feels like there should be a more efficient and reliable way of doing this.

Any help or suggestions are greatly appreciated.

Topic

[Robert Milfajt]

Try a straight copy in Xlate with the following TCL fragment

COPY OBX.#5 OBX.#5

Code:

set numItms [llength $xlateInVals]
set xlateOutVals “”
for {set x 1} {$x<$numItms} {incr x} {lappend xlateOutVals [lindex $xlateInVals $x]}

Oh, and if your using PATHCOPY of OBX segment before this, remember to null out the OBX-5 field before this copy.  I would use:

PATHCOPY @null OBX.#5

Hope this helps,

Robert Milfajt
Northwestern Medicine
Chicago, IL

[Seth Jackson]

Hey Robert,

Thanks for the reply. I gave your TCL fragment a try but, unfortunately, it doesn’t work. If you echo $xlateOutVals, the list is built correctly, but only the first element of the list appears in the message.

Command output:

4.29

12.4 12345

2(0).0(0).1(0).0(1).OBX(0)  :  >|1|NM|RBC^RBC|1|4.29|10*6/uL^million per microliter^L||N|||F|||201411110936<

2(0).0(0).1(0).0(2).OBX(0)  :  >|2|NM|HGB^HGB|2|12.4|g/dL^gram per deciliter^L||N|||F|||201411110936<

(The second value should be 12.4^12345)

[Shane Farney]

Double check your translate, and make sure you aren’t copying just a single subfield.  It definitely works (tried it myself).

[Seth Jackson]

Quadruple-checked 😆

Source

6(%g1).0(0).OBX(0).#5(0)

Destination

2(0).0(0).1(0).0(%g1).OBX(0).#5(0)

Very odd… The fact that this all takes place inside an iterate shouldn’t effect anything, should it?

[Seth Jackson]

Oh wait… tried the code on a different field and it worked.

The TCL fragment doesn’t work on OBX.#5 because the variant only has one component (0 String Data).  If you put the code in a field with multiple components, it works.

Well that makes things even more complicated.

[Shane Farney]

Instead of using “lappend”, maybe just create a string delimited by your separator carets?

[Tom Rioux]

Maybe I don’t understand what you are asking or am just thinking too simply about this, but couldn’t you just do a simple copy statement on the field and put this as your tcl fragment:

set xlateOutVals [lrange $xlateInVals 1 end]

I’m sure that I am missing something here….

Thanks….

[Seth Jackson]

Thomas,

That was actually the original code I used.  Technically it works (just like Robert’s code suggestion) but it ultimately gets truncated down to a single component due to the way the variant is built.

I’m going to try Shane’s suggestion, but I suspect that may encounter issues if the lab ever sends sub-components in OBX.#5

I wish Cloverleaf had something like xlateRawVals… something that would allow me to perform TCL fragment code against an un-parsed field, in its entirety.

[Jim Kosloskey]

You mean the vendor cannot tell you how many components it willl send in OBX-5?

They certainly should know the limit.

Moreover what’s with the unpopulated first component.

Additionally the varisntt reflects the HL/7 standard for OBX-5.

If you want to you can create a new variant with the only difference being the data type you assign to the OBX-5 field.

Personally I would wrangle a little with the source system.

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

[Charlie Bursell]

If I know Jim he will come up with a no Tcl way  😀

However if it were me, and it is not, I would use a CALL since I want to absolutely control the output

Within the iterate I would first null out all my OBX.5 subfields like:

{ { OP PATHCOPY }

   { ERR 0 }

   { IN @null }

   { OUT 2(0).1(0).1(0).0(%g1).OBX.#5 }

}

( Your addressing may be different )

Then I would implement a CALL with Tcl like:

{ { OP CALL }

   { ERR 0 }

   { IN 2(0).1(0).1(0).0(%g1).OBX.#5 }

   { OUT 2(0).1(0).1(0).0(%g1).OBX.#5 }

   { TCL {

       set outAddr [lindex $xlateOutList]

       lvarpop xlateInVals

       set cnt 0

       foreach el $xlateInVals {

           set addr $outAddr.[$cnt]

           xpmstore $xlateId $addr c $el -type st

           incr cnt

       }

   }}

}

Note each time we pop off the first element of xlateInVals.  We then get the outbound address and for each remaining element in xlateInVals we change the outbound address to end with .[0], .[1]. etc

“There may be better ways I am sure but this is what came to my mind.

There are only two people that write decent code, you and me, and sometimes I wonder about you”   😆

[Robert Milfajt]

For my way to work, you might have to do CONCAT instead of COPY and ^ as delimeter with same TCL code.

Bob

Robert Milfajt
Northwestern Medicine
Chicago, IL

[Seth Jackson]

Hey Charlie,

This is… AWESOME! I’m fairly new to Cloverleaf development so I didn’t know about the XPM commands. Looks like I’ve got some reading to do 😆 I think these will solve a lot of the problems we’ve been struggling against with our e*Gate conversions.

Thank you so much!! 😀

[Author]

Replies