Version: 4.1.4
6 Testing
All of the exports in this section are provided both by
redex/reduction-semantics (which includes
all non-GUI portions of Redex) and also exported by
redex (which includes all of Redex).
Tests to see if e1 is equal to e2.
(test--> reduction-relation e1 e2 ...) |
Tests to see if the value of e1 (which should be a term),
reduces to the e2s under reduction-relation.
Tests to see if the value of e matches the predicate p?.
Prints out how many tests passed and failed, and resets the
counters so that next time this function is called, it
prints the test results for the next round of tests.
Constructs a structure to contain the per-case test coverage of
the relation r. Use with relation-coverage
and covered-cases.
Returns #t for a value produced by make-coverage
and #f for any other.
When c is a coverage structure, rather than
#f (the default), procedures such as
apply-reduction-relation, traces, etc. count
the number applications of each case of the
reduction-relation, storing the results in c.
Extracts the coverage information recorded in c, producing
an association list mapping names to application counts.
Examples: |
> (define-language empty-lang) |
|
|
(("add" . 2) ("unnamed" . 1) ("zero" . 0)) |
Generates a random term matching pattern (in the given language).
The argument size-expr bounds the height of the generated term
(measured as the height of the derivation tree used to produce
the term).
The optional keyword argument attempt-num-expr
(default 1) provides coarse grained control over the random
decisions made during generation (e.g., the expected length of
pattern-sequences increases with attempt-num-expr).
(redex-check language pattern property-expr kw-arg ...) |
|
kw-arg | | = | | #:attempts attempts-expr | | | | | | #:source metafunction | | | | | | #:source relation-expr |
|
|
|
Searches for a counterexample to property-expr, interpreted
as a predicate universally quantified over its free
term-variables. redex-check chooses substitutions for
these free term-variables by generating random terms matching
pattern and extracting the sub-terms bound by the
names and non-terminals in pattern.
redex-check generates at most attempts-expr (default 100)
random terms in its search. The size and complexity of terms it generates
gradually increases with each failed attempt.
When the optional #:source argument is present, redex-check
generates 10% of its terms by randomly choosing a pattern from the
left-hand sides the definition of the supplied metafunction or relation.
redex-check raises an exception if a term generated from an alternate
pattern does not match the pattern.
Tests a relation as follows: for each case of relation,
check-reduction-relation generates attempts random
terms that match that case’s left-hand side and applies property
to each random term.
Like check-reduction-relation but for metafunctions.
Debugging PLT Redex Programs
It is easy to write grammars and reduction rules that are
subtly wrong and typically such mistakes result in examples
that just get stuck when viewed in a `traces’ window.
The best way to debug such programs is to find an expression
that looks like it should reduce but doesn’t and try to find
out what pattern is failing to match. To do so, use the
redex-match special form, described above.
In particular, first ceck to see if the term matches the
main non-terminal for your system (typically the expression
or program nonterminal). If it does not, try to narrow down
the expression to find which part of the term is failing to
match and this will hopefully help you find the problem. If
it does match, figure out which reduction rule should have
matched, presumably by inspecting the term. Once you have
that, extract a pattern from the left-hand side of the
reduction rule and do the same procedure until you find a
small example that shoudl work but doesn’t (but this time
you might also try simplifying the pattern as well as
simplifying the expression).