Assumed Knowledge:
* Can do basic proofs by induction and work with inductive data types as well as polymorphism.
Learning Outcomes:
* Have a broader range of Coq tactics in your proof-toolbox.
This chapter introduces several more proof strategies and
tactics that allow us to prove more interesting properties of
functional programs. We will see:
- how to use auxiliary lemmas in both "forward-style" and
"backward-style" proofs;
- how to reason about data constructors (in particular, how to use
the fact that they are injective and disjoint);
- how to create a strong induction hypothesis (and when such
strengthening is required); and
- more details on how to reason by case analysis.
The apply Tactic
We often encounter situations where the goal to be proved is
exactly the same as some hypothesis in the context or some
previously proved lemma.
Theorem silly1 :
∀(
n m o p :
nat),
n =
m →
[
n;
o] = [
n;
p]
→
[
n;
o] = [
m;
p].
Proof.
intros n m o p eq1 eq2.
rewrite ← eq1.
At this point, we could finish with "rewrite → eq2.
reflexivity." as we have done several times before. We can
achieve the same effect in a single step by using the apply
tactic instead:
apply eq2. Qed.
The apply tactic also works with conditional hypotheses
and lemmas: if the statement being applied is an implication, then
the premises of this implication will be added to the list of
subgoals needing to be proved.
Theorem silly2 :
∀(
n m o p :
nat),
n =
m →
(
∀(
q r :
nat),
q =
r → [
q;
o] = [
r;
p])
→
[
n;
o] = [
m;
p].
Proof.
intros n m o p eq1 eq2.
apply eq2.
apply eq1.
Qed.
You may find it instructive to experiment with this proof
and see if there is a way to complete it using just
rewrite
instead of
apply.
Typically, when we use
apply H, the statement
H will
begin with a
∀ that binds some
universal variables. When
Coq matches the current goal against the conclusion of
H, it
will try to find appropriate values for these variables. For
example, when we do
apply eq2 in the following proof, the
universal variable
q in
eq2 gets instantiated with
n and
r
gets instantiated with
m.
Theorem silly2a :
∀(
n m :
nat),
(
n,
n) = (
m,
m)
→
(
∀(
q r :
nat), (
q,
q) = (
r,
r)
→ [
q] = [
r])
→
[
n] = [
m].
Proof.
intros n m eq1 eq2.
apply eq2.
apply eq1.
Qed.
Exercise: 2 stars, optional (silly_ex)
Complete the following proof without using
simpl.
☐
To use the
apply tactic, the (conclusion of the) fact
being applied must match the goal exactly — for example,
apply
will not work if the left and right sides of the equality are
swapped.
In this case we can use the symmetry tactic, which switches the
left and right sides of an equality in the goal.
Exercise: 3 stars (apply_exercise1)
(
Hint: You can use
apply with previously defined lemmas, not
just hypotheses in the context. Remember that
SearchAbout is
your friend.)
☐
Exercise: 1 star, optional (apply_rewrite)
Briefly explain the difference between the tactics
apply and
rewrite. What are the situations where both can usefully be
applied?
☐
The apply ... with ... Tactic
The following silly example uses two rewrites in a row to
get from
[a,b] to
[e,f].
Example trans_eq_example :
∀(
a b c d e f :
nat),
[
a;
b] = [
c;
d]
→
[
c;
d] = [
e;
f]
→
[
a;
b] = [
e;
f].
Proof.
intros a b c d e f eq1 eq2.
rewrite → eq1.
rewrite → eq2.
reflexivity.
Qed.
Since this is a common pattern, we might like to pull it out
as a lemma recording, once and for all, the fact that equality is
transitive.
Theorem trans_eq :
∀(
X:
Type) (
n m o :
X),
n =
m → m =
o → n =
o.
Proof.
intros X n m o eq1 eq2.
rewrite → eq1.
rewrite → eq2.
reflexivity.
Qed.
Now, we should be able to use trans_eq to prove the above
example. However, to do this we need a slight refinement of the
apply tactic.
If we simply tell Coq apply trans_eq at this point, it can
tell (by matching the goal against the conclusion of the lemma)
that it should instantiate X with [nat], n with [a,b], and
o with [e,f]. However, the matching process doesn't determine
an instantiation for m: we have to supply one explicitly by
adding with (m:=[c,d]) to the invocation of apply.
apply trans_eq with (
m:=[
c;
d]).
apply eq1.
apply eq2.
Qed.
Actually, we usually don't have to include the name
m in
the
with clause; Coq is often smart enough to figure out which
instantiation we're giving. We could instead write:
apply
trans_eq with [c;d].
Exercise: 3 stars, optional (apply_with_exercise)
☐
The inversion Tactic
Recall the definition of natural numbers:
Inductive nat :
Type :=
|
O :
nat
|
S :
nat → nat.
It is obvious from this definition that every number has one of
two forms: either it is the constructor
O or it is built by
applying the constructor
S to another number. But there is more
here than meets the eye: implicit in the definition (and in our
informal understanding of how datatype declarations work in other
programming languages) are two more facts:
- The constructor S is injective. That is, if S n = S m, it
must be the case that n = m.
- The constructors O and S are disjoint. That is, O is not
equal to S n for any n.
Similar principles apply to all inductively defined types: all
constructors are injective, and the values built from distinct
constructors are never equal. For lists, the
cons constructor
is injective and
nil is different from every non-empty list.
For booleans,
true and
false are different. (Since neither
true nor
false take any arguments, their injectivity is not an
issue.) And so on.
Coq provides a tactic called
inversion that allows us to
exploit these principles in proofs. To see how to use it, let's
show explicitly that the
S constructor is injective:
By writing inversion H at this point, we ask Coq to
generate all equations that it can infer from H as additional
hypotheses, replacing variables in the goal as it goes. In the
present example, this amounts to adding a new hypothesis H1 : n =
m and replacing n by m in the goal.
inversion H. reflexivity. Qed.
Here's a more interesting example that shows how multiple
equations can be derived at once.
Theorem inversion_ex1 :
∀(
n m o :
nat),
[
n;
m] = [
o;
o]
→
[
n] = [
m].
Proof.
intros n m o H.
inversion H.
reflexivity.
Qed.
It is possible to name the equations that inversion
generates with an as ... clause:
Theorem inversion_ex2 :
∀(
n m :
nat),
[
n] = [
m]
→
n =
m.
Proof.
intros n o H.
inversion H as [
Hno].
reflexivity.
Qed.
Exercise: 1 star (inversion_ex3)
☐
While the injectivity of constructors allows us to reason
that
∀ (n m : nat), S n = S m → n = m, the converse of
this implication is an instance of a more general fact about
constructors and functions, which we will find useful below:
Theorem f_equal :
∀(
A B :
Type) (
f:
A → B) (
x y:
A),
x =
y → f x =
f y.
Proof.
intros A B f x y eq.
rewrite eq.
reflexivity.
Qed.
When used on a hypothesis involving an equality between
different constructors (e.g., S n = O), inversion solves the
goal immediately. To see why this makes sense, consider the
following proof:
We can proceed by case analysis on n. The first case is
trivial.
destruct n as [| n'].
-
intros H. reflexivity.
However, the second one doesn't look so simple: assuming
beq_nat 0 (S n') = true, we must show S n' = 0, but the latter
clearly contradictory! The way forward lies in the assumption.
After simplifying the goal state, we see that beq_nat 0 (S n') =
true has become false = true:
-
simpl.
If we use inversion on this hypothesis, Coq notices that
the subgoal we are working on is impossible, and therefore removes
it from further consideration.
intros H. inversion H. Qed.
This is an instance of a general logical principle known as
the principle of explosion, which asserts that a contradiction
entails anything, even false things. For instance:
If you find the principle of explosion confusing, remember
that these proofs are not actually showing that the conclusion of
the statement holds. Rather, they are arguing that the situation
described by the premise can never arise, so the implication is
vacuous. We'll explore the principle of explosion of more detail
in the next chapter.
Exercise: 1 star (inversion_ex6)
☐
To summarize this discussion, suppose
H is a hypothesis in the
context or a previously proven lemma of the form
c a1 a2 ...
an =
d b1 b2 ...
bm
for some constructors
c and
d and arguments
a1 ... an and
b1 ... bm. Then
inversion H has the following effect:
- If c and d are the same constructor, then, by the
injectivity of this constructor, we know that a1 = b1, a2 =
b2, etc.; inversion H adds these facts to the context, and
tries to use them to rewrite the goal.
- If c and d are different constructors, then the hypothesis
H is contradictory, and the current goal doesn't have to be
considered. In this case, inversion H marks the current goal
as completed and pops it off the goal stack.
Using Tactics on Hypotheses
By default, most tactics work on the goal formula and leave
the context unchanged. However, most tactics also have a variant
that performs a similar operation on a statement in the context.
For example, the tactic
simpl in H performs simplification in
the hypothesis named
H in the context.
Similarly,
apply L in H matches some conditional statement
L (of the form
L1 → L2, say) against a hypothesis
H in the
context. However, unlike ordinary
apply (which rewrites a goal
matching
L2 into a subgoal
L1),
apply L in H matches
H
against
L1 and, if successful, replaces it with
L2.
In other words,
apply L in H gives us a form of "forward
reasoning": from
L1 → L2 and a hypothesis matching
L1, it
produces a hypothesis matching
L2. By contrast,
apply L is
"backward reasoning": it says that if we know
L1→L2 and we are
trying to prove
L2, it suffices to prove
L1.
Here is a variant of a proof from above, using forward reasoning
throughout instead of backward reasoning.
Forward reasoning starts from what is
given (premises,
previously proven theorems) and iteratively draws conclusions from
them until the goal is reached. Backward reasoning starts from
the
goal, and iteratively reasons about what would imply the
goal, until premises or previously proven theorems are reached.
If you've seen informal proofs before (for example, in a math or
computer science class), they probably used forward reasoning. In
general, idiomatic use of Coq tends to favor backward reasoning,
but in some situations the forward style can be easier to think
about.
Exercise: 3 stars, recommended (plus_n_n_injective)
Practice using "in" variants in this exercise. (Hint: use
plus_n_Sm.)
☐
Varying the Induction Hypothesis
Sometimes it is important to control the exact form of the
induction hypothesis when carrying out inductive proofs in Coq.
In particular, we need to be careful about which of the
assumptions we move (using
intros) from the goal to the context
before invoking the
induction tactic. For example, suppose
we want to show that the
double function is injective — i.e.,
that it always maps different arguments to different results:
Theorem double_injective:
∀n m,
double n =
double m → n =
m.
The way we
start this proof is a bit delicate: if we begin with
all is well. But if we begin it with
we get stuck in the middle of the inductive case...
Theorem double_injective_FAILED :
∀n m,
double n =
double m →
n =
m.
Proof.
intros n m.
induction n as [|
n'].
-
simpl.
intros eq.
destruct m as [|
m'].
+
reflexivity.
+
inversion eq.
-
intros eq.
destruct m as [|
m'].
+
inversion eq.
+
apply f_equal.
At this point, the induction hypothesis, IHn', does not give us
n' = m' — there is an extra S in the way — so the goal is
not provable.
Abort.
What went wrong?
The problem is that, at the point we invoke the induction
hypothesis, we have already introduced
m into the context —
intuitively, we have told Coq, "Let's consider some particular
n
and
m..." and we now have to prove that, if
double n = double
m for
these particular n and
m, then
n = m.
The next tactic,
induction n says to Coq: We are going to show
the goal by induction on
n. That is, we are going to prove, for
all n, that the proposition
- P n = "if double n = double m, then n = m"
holds, by showing
- P O
(i.e., "if double O = double m then O = m") and
- P n → P (S n)
(i.e., "if double n = double m then n = m" implies "if
double (S n) = double m then S n = m").
If we look closely at the second statement, it is saying something
rather strange: it says that, for a
particular m, if we know
- "if double n = double m then n = m"
then we can prove
- "if double (S n) = double m then S n = m".
To see why this is strange, let's think of a particular
m —
say,
5. The statement is then saying that, if we know
- Q = "if double n = 10 then n = 5"
then we can prove
- R = "if double (S n) = 10 then S n = 5".
But knowing
Q doesn't give us any help at all with proving
R! (If we tried to prove
R from
Q, we would start with
something like "Suppose
double (S n) = 10..." but then we'd be
stuck: knowing that
double (S n) is
10 tells us nothing about
whether
double n is
10, so
Q is useless.)
To summarize: Trying to carry out this proof by induction on
n
when
m is already in the context doesn't work because we are
then trying to prove a relation involving
every n but just a
single m.
The good proof of
double_injective leaves
m in the goal
statement at the point where the
induction tactic is invoked on
n:
Theorem double_injective :
∀n m,
double n =
double m →
n =
m.
Proof.
intros n.
induction n as [|
n'].
-
simpl.
intros m eq.
destruct m as [|
m'].
+
reflexivity.
+
inversion eq.
-
simpl.
Notice that both the goal and the induction hypothesis are
different this time: the goal asks us to prove something more
general (i.e., to prove the statement for every m), but the IH
is correspondingly more flexible, allowing us to choose any m we
like when we apply the IH.
intros m eq.
Now we've chosen a particular m and introduced the assumption
that double n = double m. Since we are doing a case analysis on
n, we also need a case analysis on m to keep the two "in sync."
destruct m as [| m'].
+ simpl.
The 0 case is trivial:
At this point, since we are in the second branch of the destruct
m, the m' mentioned in the context is the predecessor of the
m we started out talking about. Since we are also in the S
branch of the induction, this is perfect: if we instantiate the
generic m in the IH with the current m' (this instantiation is
performed automatically by the apply in the next step), then
IHn' gives us exactly what we need to finish the proof.
apply IHn'. inversion eq. reflexivity. Qed.
What you should take away from all this is that we need to be
careful about using induction to try to prove something too
specific: If we're proving a property of
n and
m by induction
on
n, we may need to leave
m generic.
The following exercise requires the same pattern.
Exercise: 2 stars (beq_nat_true)
☐
Exercise: 2 stars, advanced (beq_nat_true_informal)
Give a careful informal proof of
beq_nat_true, being as explicit
as possible about quantifiers.
☐
The strategy of doing fewer
intros before an
induction to
obtain a more general IH doesn't always work by itself; sometimes
a little
rearrangement of quantified variables is needed.
Suppose, for example, that we wanted to prove
double_injective
by induction on
m instead of
n.
Theorem double_injective_take2_FAILED :
∀n m,
double n =
double m →
n =
m.
Proof.
intros n m.
induction m as [|
m'].
-
simpl.
intros eq.
destruct n as [|
n'].
+
reflexivity.
+
inversion eq.
-
intros eq.
destruct n as [|
n'].
+
inversion eq.
+
apply f_equal.
Abort.
The problem here is that, to do induction on
m, we must first
introduce
n. (If we simply say
induction m without
introducing anything first, Coq will automatically introduce
n
for us!)
What can we do about this? One possibility is to rewrite the
statement of the lemma so that
m is quantified before
n. This
will work, but it's not nice: We don't want to have to twist the
statements of lemmas to fit the needs of a particular strategy for
proving them — we want to state them in the most clear and
natural way.
What we can do instead is to first introduce all the quantified
variables and then
re-generalize one or more of them,
selectively taking variables out of the context and putting them
back at the beginning of the goal. The
generalize dependent
tactic does this.
Theorem double_injective_take2 :
∀n m,
double n =
double m →
n =
m.
Proof.
intros n m.
generalize dependent n.
induction m as [|
m'].
-
simpl.
intros n eq.
destruct n as [|
n'].
+
reflexivity.
+
inversion eq.
-
intros n eq.
destruct n as [|
n'].
+
inversion eq.
+
apply f_equal.
apply IHm'.
inversion eq.
reflexivity.
Qed.
Let's look at an informal proof of this theorem. Note that
the proposition we prove by induction leaves
n quantified,
corresponding to the use of generalize dependent in our formal
proof.
Theorem: For any nats
n and
m, if
double n = double m, then
n = m.
Proof: Let
m be a
nat. We prove by induction on
m that, for
any
n, if
double n = double m then
n = m.
- First, suppose m = 0, and suppose n is a number such
that double n = double m. We must show that n = 0.
Since m = 0, by the definition of double we have double n =
0. There are two cases to consider for n. If n = 0 we are
done, since m = 0 = n, as required. Otherwise, if n = S n'
for some n', we derive a contradiction: by the definition of
double, we can calculate double n = S (S (double n')), but
this contradicts the assumption that double n = 0.
- Second, suppose m = S m' and that n is again a number such
that double n = double m. We must show that n = S m', with
the induction hypothesis that for every number s, if double s =
double m' then s = m'.
By the fact that m = S m' and the definition of double, we
have double n = S (S (double m')). There are two cases to
consider for n.
If n = 0, then by definition double n = 0, a contradiction.
Thus, we may assume that n = S n' for some n', and again by
the definition of double we have S (S (double n')) =
S (S (double m')), which implies by inversion that double n' =
double m'. Instantiating the induction hypothesis with n' thus
allows us to conclude that n' = m', and it follows immediately
that S n' = S m'. Since S n' = n and S m' = m, this is just
what we wanted to show. ☐
Before we close this section and move on to some exercises, let's
digress briefly and use
beq_nat_true to prove a similar property
about identifiers that we'll need in later chapters:
Exercise: 3 stars, recommended (gen_dep_practice)
Prove this by induction on
l.
☐
Exercise: 3 stars, optional (app_length_cons)
Prove this by induction on
l1, without using
app_length
from
Lists.
☐
Exercise: 4 stars, optional (app_length_twice)
Prove this by induction on
l, without using
app_length from
Lists.
☐
Exercise: 3 stars, optional (double_induction)
Prove the following principle of induction over two naturals.
☐
Unfolding Definitions
It sometimes happens that we need to manually unfold a Definition
so that we can manipulate its right-hand side. For example, if we
define...
... and try to prove a simple fact about square...
... we get stuck:
simpl doesn't simplify anything at this point,
and since we haven't proved any other facts about
square, there
is nothing we can
apply or
rewrite with.
To make progress, we can manually
unfold the definition of
square:
Now we have plenty to work with: both sides of the equality are
expressions involving multiplication, and we have lots of facts
about multiplication at our disposal. In particular, we know that
it is commutative and associative, and from these facts it is not
hard to finish the proof.
At this point, a deeper discussion of unfolding and simplification
is in order.
You may already have observed that tactics like
simpl,
reflexivity, and
apply will often unfold the definitions of
functions automatically when it allows them to make progress. For
example, if we define
foo m to be the constant
5,
then the simpl in the following proof (or the reflexivity, if
we omit the simpl) will unfold foo m to (fun x ⇒ 5) m and
then further simplify this expression to just 5.
However, this automatic unfolding is rather conservative. For
example, if we define a slightly more complicated function
involving a pattern match...
Definition bar x :=
match x with
|
O ⇒ 5
|
S _ ⇒ 5
end.
...then the analogous proof will get stuck:
The reason that
simpl doesn't make progress here is that it
notices that, after tentatively unfolding
bar m, it is left with
a match whose scrutinee,
m, is a variable, so the
match cannot
be simplified further. (It is not smart enough to notice that the
two branches of the
match are identical.) So it gives up on
unfolding
bar m and leaves it alone. Similarly, tentatively
unfolding
bar (m+1) leaves a
match whose scrutinee is a
function application (that, itself, cannot be simplified, even
after unfolding the definition of
+), so
simpl leaves it
alone.
At this point, there are two ways to make progress. One is to use
destruct m to break the proof into two cases, each focusing on a
more concrete choice of
m (
O vs
S _). In each case, the
match inside of
bar can now make progress, and the proof is
easy to complete.
Fact silly_fact_2 :
∀m,
bar m + 1 =
bar (
m + 1) + 1.
Proof.
intros m.
destruct m.
-
simpl.
reflexivity.
-
simpl.
reflexivity.
Qed.
This approach works, but it depends on our recognizing that the
match hidden inside
bar is what was preventing us from making
progress.
A more straightforward way to finish the proof is to explicitly
tell Coq to unfold
bar.
Now it is apparent that we are stuck on the match expressions on
both sides of the =, and we can use destruct to finish the
proof without thinking too hard.
destruct m.
- reflexivity.
- reflexivity.
Qed.
Using destruct on Compound Expressions
We have seen many examples where
destruct is used to
perform case analysis of the value of some variable. But
sometimes we need to reason by cases on the result of some
expression. We can also do this with
destruct.
Here are some examples:
After unfolding
sillyfun in the above proof, we find that
we are stuck on
if (beq_nat n 3) then ... else .... But either
n is equal to
3 or it isn't, so we can use
destruct (beq_nat
n 3) to let us reason about the two cases.
In general, the
destruct tactic can be used to perform case
analysis of the results of arbitrary computations. If
e is an
expression whose type is some inductively defined type
T, then,
for each constructor
c of
T,
destruct e generates a subgoal
in which all occurrences of
e (in the goal and in the context)
are replaced by
c.
Exercise: 3 stars, optional (combine_split)
☐
However,
destructing compound expressions requires a bit of
care, as such
destructs can sometimes erase information we need
to complete a proof. For example, suppose we define a function
sillyfun1 like
this:
Now suppose that we want to convince Coq of the (rather
obvious) fact that sillyfun1 n yields true only when n is
odd. By analogy with the proofs we did with sillyfun above, it
is natural to start the proof like this:
We get stuck at this point because the context does not
contain enough information to prove the goal! The problem is that
the substitution performed by
destruct is too brutal — it threw
away every occurrence of
beq_nat n 3, but we need to keep some
memory of this expression and how it was destructed, because we
need to be able to reason that, since
beq_nat n 3 = true in this
branch of the case analysis, it must be that
n = 3, from which
it follows that
n is odd.
What we would really like is to substitute away all existing
occurences of
beq_nat n 3, but at the same time add an equation
to the context that records which case we are in. The
eqn:
qualifier allows us to introduce such an equation, giving it a
name that we choose.
Exercise: 2 stars (destruct_eqn_practice)
☐
Review
We've now seen many of Coq's most fundamental tactics. We'll
introduce a few more in the coming chapters, and later on we'll
see some more powerful
automation tactics that make Coq help us
with low-level details. But basically we've got what we need to
get work done.
Here are the ones we've seen:
- intros: move hypotheses/variables from goal to context
- reflexivity: finish the proof (when the goal looks like e =
e)
- apply: prove goal using a hypothesis, lemma, or constructor
- apply... in H: apply a hypothesis, lemma, or constructor to
a hypothesis in the context (forward reasoning)
- apply... with...: explicitly specify values for variables
that cannot be determined by pattern matching
- simpl: simplify computations in the goal
- simpl in H: ... or a hypothesis
- rewrite: use an equality hypothesis (or lemma) to rewrite
the goal
- rewrite ... in H: ... or a hypothesis
- symmetry: changes a goal of the form t=u into u=t
- symmetry in H: changes a hypothesis of the form t=u into
u=t
- unfold: replace a defined constant by its right-hand side in
the goal
- unfold... in H: ... or a hypothesis
- destruct... as...: case analysis on values of inductively
defined types
- destruct... eqn:...: specify the name of an equation to be
added to the context, recording the result of the case
analysis
- induction... as...: induction on values of inductively
defined types
- inversion: reason by injectivity and distinctness of
constructors
- assert (e) as H: introduce a "local lemma" e and call it
H
- generalize dependent x: move the variable x (and anything
else that depends on it) from the context back to an explicit
hypothesis in the goal formula
Additional Exercises
Exercise: 3 stars (beq_nat_sym)
☐
Exercise: 3 stars, advanced, optional (beq_nat_sym_informal)
Give an informal proof of this lemma that corresponds to your
formal proof above:
Theorem: For any
nats
n m,
beq_nat n m = beq_nat m n.
Proof:
☐
Exercise: 3 stars, optional (beq_nat_trans)
☐
Exercise: 3 stars, advanced (split_combine)
We proved, in an exercise above, that for all lists of pairs,
combine is the inverse of
split. How would you formalize the
statement that
split is the inverse of
combine? When is this
property true?
Complete the definition of
split_combine_statement below with a
property that states that
split is the inverse of
combine. Then, prove that the property holds. (Be sure to leave
your induction hypothesis general by not doing
intros on more
things than necessary. Hint: what property do you need of
l1
and
l2 for
split combine l1 l2 = (l1,l2) to be true?)
☐
Exercise: 3 stars, advanced (filter_exercise)
This one is a bit challenging. Pay attention to the form of your
induction hypothesis.
☐
Exercise: 4 stars, advanced, recommended (forall_exists_challenge)
Define two recursive
Fixpoints,
forallb and
existsb. The
first checks whether every element in a list satisfies a given
predicate:
forallb oddb [1;3;5;7;9] =
true
forallb negb [
false;
false] =
true
forallb evenb [0;2;4;5] =
false
forallb (
beq_nat 5) [] =
true
The second checks whether there exists an element in the list that
satisfies a given predicate:
existsb (
beq_nat 5) [0;2;3;6] =
false
existsb (
andb true) [
true;
true;
false] =
true
existsb oddb [1;0;0;0;0;3] =
true
existsb evenb [] =
false
Next, define a
nonrecursive version of
existsb — call it
existsb' — using
forallb and
negb.
Finally, prove a theorem
existsb_existsb' stating that
existsb' and
existsb have the same behavior.