Please Make a Donation:
Support This Project

Hosted by:
Get Python Knowledge Engine (PyKE) at SourceForge.net. Fast, secure and Free Open Source software downloads

Matching Two Patterns

You've now seen all of the different kinds of patterns and how they are matched to data. Not too bad so far...

Patterns are used to generalize statements. One situation where you need to generalize a statement is when you want to ask a question. That's been covered above.

The other situation where you need to generalize a statement is when you write rules, which are explained later.

But rules also need to be able to match one pattern to another pattern, and not just match patterns to data as we have discussed so far.

So before we can move on to rules, we need to examine how one pattern is matched to another pattern.

The short answer is that it all comes down to pattern variables and that pattern variables may not only be bound to data, but may also be bound to other patterns.

Binding to a Literal Pattern

Binding a pattern variable to a literal pattern is just like binding it to the data within that literal pattern. Nothing fancy here!

Binding to Another Pattern Variable

When pattern variable A is bound to pattern variable B, they essentially become the same pattern variable. Basically, pattern variable A becomes pattern variable B (or, you might say, defers to pattern variable B).

Let's say that pattern variable A has been bound to pattern variable B and that pattern variable B is still unbound.

  1. Prior to binding pattern variable B to a value, it doesn't matter whether you ask if pattern variable A is bound or pattern variable B is bound, the answer is False for both.
  2. It doesn't matter whether you match pattern variable A to a value or pattern variable B to a value. In both cases, it is pattern variable B that gets bound to this value.
  3. And once pattern variable B is bound to a value, it doesn't matter whether you ask for the bound value of pattern variable A or pattern variable B, you will get the same value.

So for all intents and purposes pattern variable A and pattern variable B become the same pattern variable.

Binding to a Tuple Pattern

Because pattern variables may be bound to tuple patterns, the term fully bound is introduced. Asking whether the pattern variable is fully bound means that not only is it bound to a value (the tuple pattern), but that all of the subordinate patterns (recursively) within the tuple pattern are also bound to values.

Being fully bound means that the bound value of the pattern variable can be converted into standard Python data without any pattern variables in it. This is important when Pyke wants to talk to Python because Python has no concept of storing variables within its data structures.

Pathological Question

What is the bound value of pattern variable $y after matching the following two tuple patterns:

Tuple pattern A:
((ho, $_, ($a, $a)), ($a, $a, $b), ($a, *$b))
Tuple pattern B:
($x, $x, $y)

The answer is here (but no peeking!).

More:

Literal Patterns

Explanation of literal patterns.

Pattern Variables

Explanation of pattern variables.

Tuple Patterns

Explanation of tuple patterns.

Matching Two Patterns

Explanation of matching two patterns together, vs matching a pattern to data.

Page last modified Mon, Oct 27 2008.