tcl procs and table data

For your first question, in Tcl just use ‘tbllookup tablename value’.

As far as your second question goes your are looking for a Tcl package.  Any good Tcl book will have details on how to create and access them.  I use them heavily for the same scenario you describe.

The 5 second intro to packages:

   Create a file with the procs you want in the library.  Use an extension other than .tcl, I use .pkg.  I typically store the package file either in the $HCIROOT/tclprocs or $HCIROOT/tcl/lib/ directory so all sites can access the same library.

   At the top of the file add the line ‘package provide 1.0′.

   From cmd line go to the location of the package file and type ‘pkg_mkindex . *.pkg’.  This should create a pkgIndex.tcl file.

   From any Tcl code (proc, script, etc.) simply add ‘package require ‘.  Now you can use your library procs without them showing in the NetConfig UPoC lists.

Depending on what your building you may also want to explore namespaces which can take packages to the next level, although at the expense of complexity.[/list]

Oops, should be pkg_mkIndex.

Doc.

We use TCL translations in most of our interfaces and use ‘tables’ extensively.  We use a TCL array …

proc QryTableTop {aCode} {

       # make the array global, so it is always available

   global QryTableTopArray

       #

       # if the array requires initialisation, do this now

       #

   if {[array exists QryTableTopArray] == 0} {

       set QryTableTopArray(A)  AKHS

…

       set QryTableTopArray(X)  3811

       set QryTableTopArray(Z)  3812

   }

       # set the return variable to the default value

   set myData $aCode

       # make sure we don’t fall over if there is NO data!

   catch {set myData $QryTableTopArray($aCode)}

       # return the data from the array

   return $myData

}

We also have a library of common functions that we use and the library code and common functions exist in a ‘LIB’ directory, which is referenced from a site tclIndex

eg

# Tcl autoload index file, version 2.0

# This file is generated by the “auto_mkindex” command

# and sourced to set up indexing information for one or

# more commands.  Typically each line is a command that

# sets an element in the auto_index array, where the

# element name is the name of a command and the value is

# a script that loads the command.

set auto_index(ccm_xlate_ris)

]

set auto_index(ccm_xlate_cap)

]

set auto_index(ccm_xlate_ult)

]

set auto_index(ccm_xlate_rad)

]

set auto_index(ccm_xlate_sud)

]

set auto_index(MakeMrn)

]

set auto_index(MakeAcct)

]

set auto_index(MakePtType)

]

set auto_index(MakeLabHospitalService)

]

set auto_index(MakeCtrlId)

]

set auto_index(MapHexChars)

]

set auto_index(TranEncChars)

]

…

set auto_index(gen_code_resend)

]

set auto_index(gen_code_printmsg)

]

set auto_index(gen_code_splitter)

]

set auto_index(gen_code_mkreply)

]

set auto_index(Print)

]

set auto_index(RetRepeat)

]

set auto_index(RetComp)

]

set auto_index(ChangeField)

]

set auto_index(SourceTablesFile)

]

We run on Unix, so we use ‘make’ and CVS to create the TCL files and tclIndex. We also use Unix links to keep one copy of the source and provide a simple method of managing code changes.

Question #2:

I’m sure this will horrify the purists, but one fast, down and dirty way to make those procs available in unix without having to go the package route is to put them into filename then add a line inside …/integrator/tcl/lib/localInit.tcl that says something like

source $env(HCIROOT)/…/filename

Chris