The GROUP instruction

Those declarations sections which declare items that are listed in a window of values (sensors, controls, variables and smartcabs) can have one or more groups defined before the actual items themselves.

A group instruction defines the order in which the items are displayed in the appropriate window, and possibly groups items together on a single line.

The best way to explain this is with an example:

Lets suppose that our layout is made up of several track sections (and possibly other things we are not considering here). Each track section has the following variables associated with it:

state:: some indication of whether the track is in use, reserved for a train, or whatever.
train:: which train is using the section.
next:: which section the train will go to next.
prev:: which section the train has just come from.

We have sections called s1, s2, s3 etc.

Therefore we might have the following variables declaration:

Variables:
s1state, s2state,  s3state
s1train, s2train,  s3train
s1next,  s2next,   s3next
s1prev,  s2prev,   s3prev

We would naturally like the variables to be grouped in the variables window like this:

s1state	s1train	s1next	s1prev
s2state	s2train	s2next	s2prev
s3state	s3train	s3next	s3prev

To achieve this we would use the following group statement:

Variables:
group tracks (state, train, next, prev)=(s1, s2, s3)
s1state, s2state,

This would produce a window like this:

In the group statement the first list is a list of column names, and the second is a list of row names.

The mechanism for group is a bit wild and magical... it tries to look for variables that contain the bits you give it, in any way that works, so the following:

GROUP: xx (c1, c2)=(r1, r2)

Would find and display:

Variables: c1r1, c1r2, c2r1, c2r2

or:

Variables: r1c1, r2c1, r1c2, r2c2

or even:

Variables: cr11, cr12, cr21, cr22

but it would not match:

Variables: rc11, rc12, rc21, rc22

In general it looks through all the variables that have been declared (including all elements in all structures and all elements in arrays) and finds variables with names which contain items from the second list (the row names), and when that portion is removed from the name, the remainder matches something in the first list (the column names)

The identified items are displayed in the order defined in the group statement, with the row and column labels attached.

Note that the group statement merely tells Tcc how to arrange some variables (or sensors or controls) in their window. It does NOT declare the variables themselves. For a short form of variable declaration see the STRUCT statement.

The row name list could instead be an array name, such as in:

Variables:
Group: trains (cab, len, sp)=(tr[])
cabtr [6], lentr [6], sptr [6]

This declares six cab-for-train (cabtr) variables, six length of train variables and six speed of train variables, and is similar to:

Group: trains (cab, len, sp)=(tr1, tr2, tr3, tr4, tr5, tr6)
cabtr1, cabtr2, cabtr3, cabtr4, cabtr5, cabtr6
lentr1, lentr2, ...

Except that the first declaration makes an array that can be accessed as an array, and the second for would simple make 18 independent variables.

Any items not defined in any group are shown at the end of the window, in the order they appear in the rest of the declaration.

As a special case, if the first column name begins with a period then it is assumed to be an element in a structure. So the following declarations go together:

Group: trains (.cab, .len, .sp)=(tr1, tr2, tr3)
Struct (cab, len, sp)=(tr1, tr2, tr3)

Back to declarations page

Last updated 24 Feb 2009