Locating Value in a List

Jon,

For your last question is the field somewhere you are referring to as the target a target field in an Xlate?

If so I suspect your problem is you are not treating xlateOutVals as a list (possibly not xlateInVals either). So the space in the variable is treated as a liste lement terminator. The Xlate treats xlateOutVlas (and the other xlateInxxxx, xlateOutxxxx data items) as lists and if you don’t then the Xlate will not behave the way you want.

Jon,

As for your first question.

The field is not a list in itself but it appears you could split on ; to create a list of the individual categories.

Then I think you can use lsearch with a -glob option to get the specific terms you are looking for. If you also use the -all option you will get a list of all of the indices of the original list which matched your glob search.

If you want to be really sure you only search the left hand side of the : then using split I believe there is a way to get a compund list by specifying multiple ( :; ) split characters. Then using a combination of list commands (including the lsearch described above) you can only search the terms to the left of the :.

You can also populate a list of the terms you want to search on and use list commands to drive your search. However if that changes frequently you might find maintaing the list in Tcl a bit of a chore.

Another option is to use a Lookup Table to contain your terms. You can make the left side of the table entry something like term1, term2, etc with the right side the actual term. Then using a loop routine and Cloverleaf provided Tcl Table lookup command build the key (the left side of the Table) adding the counter to the end and incrementing it until you receive the default (means you have exhausted the Table and thus your terms). So now you can update your termis to search for via the Table Configurator. Then do a purge caches and your Tcl proc will get the updfated verion of table of terms.

If you want to discuss this further, contact me via email off line.

To expand on what Jim said:

Assume the string is in variable fld

# Make sure array is clear

array unset fldArray

foreach el [split $fld ;] {

   lassign [split $el :]  tag data

   set fldArray($tag) $data

}

Now you can simply access the array by tag name for what you want

Also, heed Jim’s warning about treating xlateIn and out vals as a list

For example assume the field contains:  “Jack and Jill went up the hill”

If you echo $xlateInVals you would see {Jack and Jill went up the hill} because of the spaces.  Note the braces

If you assigned to xlateOutVals and you only had one address you would see just “Jack”

set xlateOutVals

This behaved differently on our system . With [split $fld ;] it returned each word as a separate list element. But parsed as desired when adding quotes around the semicolon [split $fld “;”]. I didn’t expect it to make a difference, but it did. Perhaps it saw the semicolon as the end of the split statement and reverted to the default of whitespace. Charlie, any insight?

Jon, depending on your needs you may also want to [string trim] the variables before setting them in the array if you find the whitespace messing with you.

Cheers.

yeah, forgot you may have to quote the ; since it is a special character in Tcl

As for the array it is nothing special.  Treat as you would any other Tcl variable

Jon,

First, if you want to see the contents of the array use parray which will display the data on its own. You don’t have to echo its return value:

Don’t forget that xlateInVals and xlateOutVals ARE LISTS. If you want to read the contents of xlateInVals you have to index the value you want:

When writing to xlateOutVals, you are building a list. Make sure you make your output data into a list element first. Depending on the logic you’re using you can do

The offset values, by definition, are zero-based and for xlateInVals and xlateOutVals refer to the rows of items in the Source and Destination boxes. If you want to refer to the second row of Source values you would use [lindex $xlateInVals 1]. If you are writing Tcl code, that is the only way you can access those Source/Destination values.

It’s difficult to say. It would be helpful to know a bit more about what you are trying to accomplish. e.g., are you trying to create a separate OBX for each of these array items? Do you know ahead of time which elements in the array you want to use if they appear? Do you just want to assign the array values to specific @variables?

It might be helpful to note that your xlate is a single Tcl instance. If you write a bit of code in the first line of an xlate that sets a Tcl variable, that variable exists until the end of that xlate, not just for the life of the single xlate action. If you want to manipulate that variable throughout your xlate, it’s not necessary to pass it from one line to another via @variables.

I’m sure that some people will disagree, but sometimes it’s just easier, cleaner and faster to manipulate data in a Tcl snipped rather than trying to do it all with xlate actions.

As with all of this, there are many ways to accomplish your task. It’s your choice of which is best for you and your circumstances.

Sorry for being so chatty today. To get back to your question about getting the info out of the array. You don’t need to use those array commands. Assuming your array is named labelArray and the array element is “Interval/Occurrences”, just treat it like any other variable except that it has a subscript.

Just make sure the bit in parens matches exactly what was loaded into the array. You can use parray as a diagnostic tool to verify the spelling and any punctuation that may be there.