Mercurial > cpdt > repo

annotate src/Coinductive.v @ 406:fc03a67810e8

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Typesetting pass over DataStruct
author Adam Chlipala <adam@chlipala.net>
date Fri, 08 Jun 2012 14:31:17 -0400
parents df7cd24383ae
children a730378789f5
rev   line source
adam@398 1 (* Copyright (c) 2008-2012, Adam Chlipala
adamc@62 2 *
adamc@62 3 * This work is licensed under a
adamc@62 4 * Creative Commons Attribution-Noncommercial-No Derivative Works 3.0
adamc@62 5 * Unported License.
adamc@62 6 * The license text is available at:
adamc@62 7 * http://creativecommons.org/licenses/by-nc-nd/3.0/
adamc@62 8 *)
adamc@62 9
adamc@62 10 (* begin hide *)
adamc@62 11 Require Import List.
adamc@62 12
adam@314 13 Require Import CpdtTactics.
adamc@62 14
adam@402 15 Definition bad := 0.
adam@402 16
adamc@62 17 Set Implicit Arguments.
adamc@62 18 (* end hide *)
adamc@62 19
adamc@62 20
adamc@62 21 (** %\chapter{Infinite Data and Proofs}% *)
adamc@62 22
adam@395 23 (** In lazy functional programming languages like %\index{Haskell}%Haskell, infinite data structures are everywhere%~\cite{whyfp}%. Infinite lists and more exotic datatypes provide convenient abstractions for communication between parts of a program. Achieving similar convenience without infinite lazy structures would, in many cases, require acrobatic inversions of control flow.
adamc@62 24
adam@346 25 %\index{laziness}%Laziness is easy to implement in Haskell, where all the definitions in a program may be thought of as mutually recursive. In such an unconstrained setting, it is easy to implement an infinite loop when you really meant to build an infinite list, where any finite prefix of the list should be forceable in finite time. Haskell programmers learn how to avoid such slip-ups. In Coq, such a laissez-faire policy is not good enough.
adamc@62 26
adam@398 27 We spent some time in the last chapter discussing the %\index{Curry-Howard correspondence}%Curry-Howard isomorphism, where proofs are identified with functional programs. In such a setting, infinite loops, intended or otherwise, are disastrous. If Coq allowed the full breadth of definitions that Haskell did, we could code up an infinite loop and use it to prove any proposition vacuously. That is, the addition of general recursion would make CIC _inconsistent_. For an arbitrary proposition [P], we could write:
adamc@202 28 [[
adamc@202 29 Fixpoint bad (u : unit) : P := bad u.
adamc@205 30 ]]
adamc@205 31
adamc@202 32 This would leave us with [bad tt] as a proof of [P].
adamc@62 33
adamc@62 34 There are also algorithmic considerations that make universal termination very desirable. We have seen how tactics like [reflexivity] compare terms up to equivalence under computational rules. Calls to recursive, pattern-matching functions are simplified automatically, with no need for explicit proof steps. It would be very hard to hold onto that kind of benefit if it became possible to write non-terminating programs; we would be running smack into the halting problem.
adamc@62 35
adam@354 36 One solution is to use types to contain the possibility of non-termination. For instance, we can create a %``%#"#non-termination monad,#"#%''% inside which we must write all of our general-recursive programs; several such approaches are surveyed in Chapter 7. This is a heavyweight solution, and so we would like to avoid it whenever possible.
adamc@62 37
adam@402 38 Luckily, Coq has special support for a class of lazy data structures that happens to contain most examples found in Haskell. That mechanism,%\index{co-inductive types}% _co-inductive types_, is the subject of this chapter. *)
adamc@62 39
adamc@62 40
adamc@62 41 (** * Computing with Infinite Data *)
adamc@62 42
adam@398 43 (** Let us begin with the most basic type of infinite data, _streams_, or lazy lists.%\index{Vernacular commands!CoInductive}% *)
adamc@62 44
adamc@62 45 Section stream.
adam@351 46 Variable A : Type.
adamc@62 47
adam@351 48 CoInductive stream : Type :=
adamc@62 49 | Cons : A -> stream -> stream.
adamc@62 50 End stream.
adamc@62 51
adamc@62 52 (** The definition is surprisingly simple. Starting from the definition of [list], we just need to change the keyword [Inductive] to [CoInductive]. We could have left a [Nil] constructor in our definition, but we will leave it out to force all of our streams to be infinite.
adamc@62 53
adam@402 54 How do we write down a stream constant? Obviously simple application of constructors is not good enough, since we could only denote finite objects that way. Rather, whereas recursive definitions were necessary to _use_ values of recursive inductive types effectively, here we find that we need%\index{co-recursive definitions}% _co-recursive definitions_ to _build_ values of co-inductive types effectively.
adamc@62 55
adam@346 56 We can define a stream consisting only of zeroes.%\index{Vernacular commands!CoFixpoint}% *)
adamc@62 57
adamc@62 58 CoFixpoint zeroes : stream nat := Cons 0 zeroes.
adamc@62 59
adam@346 60 (* EX: Define a stream that alternates between [true] and [false]. *)
adam@346 61 (* begin thide *)
adam@346 62
adamc@62 63 (** We can also define a stream that alternates between [true] and [false]. *)
adamc@62 64
adam@346 65 CoFixpoint trues_falses : stream bool := Cons true falses_trues
adam@346 66 with falses_trues : stream bool := Cons false trues_falses.
adam@346 67 (* end thide *)
adamc@62 68
adamc@62 69 (** Co-inductive values are fair game as arguments to recursive functions, and we can use that fact to write a function to take a finite approximation of a stream. *)
adamc@62 70
adam@348 71 (* EX: Define a function to calculate a finite approximation of a stream, to a particular length. *)
adam@346 72 (* begin thide *)
adam@346 73
adamc@211 74 Fixpoint approx A (s : stream A) (n : nat) : list A :=
adamc@62 75 match n with
adamc@62 76 | O => nil
adamc@62 77 | S n' =>
adamc@62 78 match s with
adamc@62 79 | Cons h t => h :: approx t n'
adamc@62 80 end
adamc@62 81 end.
adamc@62 82
adamc@62 83 Eval simpl in approx zeroes 10.
adamc@211 84 (** %\vspace{-.15in}% [[
adamc@62 85 = 0 :: 0 :: 0 :: 0 :: 0 :: 0 :: 0 :: 0 :: 0 :: 0 :: nil
adamc@62 86 : list nat
adam@302 87 ]]
adam@302 88 *)
adamc@211 89
adam@346 90 Eval simpl in approx trues_falses 10.
adamc@211 91 (** %\vspace{-.15in}% [[
adamc@62 92 = true
adamc@62 93 :: false
adamc@62 94 :: true
adamc@62 95 :: false
adamc@62 96 :: true :: false :: true :: false :: true :: false :: nil
adamc@62 97 : list bool
adam@346 98 ]]
adam@346 99 *)
adamc@62 100
adam@349 101 (* end thide *)
adamc@62 102
adam@402 103 (* begin hide *)
adam@402 104 Definition looper := 0.
adam@402 105 (* end hide *)
adam@402 106
adam@398 107 (** So far, it looks like co-inductive types might be a magic bullet, allowing us to import all of the Haskeller's usual tricks. However, there are important restrictions that are dual to the restrictions on the use of inductive types. Fixpoints _consume_ values of inductive types, with restrictions on which _arguments_ may be passed in recursive calls. Dually, co-fixpoints _produce_ values of co-inductive types, with restrictions on what may be done with the _results_ of co-recursive calls.
adamc@62 108
adam@402 109 The restriction for co-inductive types shows up as the%\index{guardedness condition}% _guardedness condition_. First, consider this stream definition, which would be legal in Haskell.
adamc@62 110 [[
adamc@62 111 CoFixpoint looper : stream nat := looper.
adam@346 112 ]]
adamc@205 113
adam@346 114 <<
adamc@62 115 Error:
adamc@62 116 Recursive definition of looper is ill-formed.
adamc@62 117 In environment
adamc@62 118 looper : stream nat
adamc@62 119
adamc@62 120 unguarded recursive call in "looper"
adam@346 121 >>
adamc@205 122
adam@398 123 The rule we have run afoul of here is that _every co-recursive call must be guarded by a constructor_; that is, every co-recursive call must be a direct argument to a constructor of the co-inductive type we are generating. It is a good thing that this rule is enforced. If the definition of [looper] were accepted, our [approx] function would run forever when passed [looper], and we would have fallen into inconsistency.
adamc@62 124
adam@346 125 Some familiar functions are easy to write in co-recursive fashion. *)
adamc@62 126
adamc@62 127 Section map.
adam@351 128 Variables A B : Type.
adamc@62 129 Variable f : A -> B.
adamc@62 130
adamc@62 131 CoFixpoint map (s : stream A) : stream B :=
adamc@62 132 match s with
adamc@62 133 | Cons h t => Cons (f h) (map t)
adamc@62 134 end.
adamc@62 135 End map.
adamc@62 136
adam@402 137 (* begin hide *)
adam@402 138 Definition filter := 0.
adam@402 139 (* end hide *)
adam@402 140
adam@402 141 (** This code is a literal copy of that for the list [map] function, with the [nil] case removed and [Fixpoint] changed to [CoFixpoint]. Many other standard functions on lazy data structures can be implemented just as easily. Some, like [filter], cannot be implemented. Since the predicate passed to [filter] may reject every element of the stream, we cannot satisfy the guardedness condition.
adamc@62 142
adam@398 143 The implications of the condition can be subtle. To illustrate how, we start off with another co-recursive function definition that _is_ legal. The function [interleave] takes two streams and produces a new stream that alternates between their elements. *)
adamc@62 144
adamc@62 145 Section interleave.
adam@351 146 Variable A : Type.
adamc@62 147
adamc@62 148 CoFixpoint interleave (s1 s2 : stream A) : stream A :=
adamc@62 149 match s1, s2 with
adamc@62 150 | Cons h1 t1, Cons h2 t2 => Cons h1 (Cons h2 (interleave t1 t2))
adamc@62 151 end.
adamc@62 152 End interleave.
adamc@62 153
adamc@62 154 (** Now say we want to write a weird stuttering version of [map] that repeats elements in a particular way, based on interleaving. *)
adamc@62 155
adamc@62 156 Section map'.
adam@351 157 Variables A B : Type.
adamc@62 158 Variable f : A -> B.
adamc@68 159 (* begin thide *)
adamc@62 160 (** [[
adamc@62 161 CoFixpoint map' (s : stream A) : stream B :=
adamc@62 162 match s with
adam@346 163 | Cons h t => interleave (Cons (f h) (map' t)) (Cons (f h) (map' t))
adamc@68 164 end.
adamc@205 165 ]]
adamc@211 166
adam@346 167 We get another error message about an unguarded recursive call. *)
adamc@62 168
adam@346 169 End map'.
adam@346 170
adam@346 171 (** What is going wrong here? Imagine that, instead of [interleave], we had called some other, less well-behaved function on streams. Here is one simpler example demonstrating the essential pitfall. We start defining a standard function for taking the tail of a stream. Since streams are infinite, this operation is total. *)
adam@346 172
adam@346 173 Definition tl A (s : stream A) : stream A :=
adam@346 174 match s with
adam@346 175 | Cons _ s' => s'
adam@346 176 end.
adam@346 177
adam@346 178 (** Coq rejects the following definition that uses [tl].
adam@346 179 [[
adam@346 180 CoFixpoint bad : stream nat := tl (Cons 0 bad).
adam@346 181 ]]
adam@346 182
adam@346 183 Imagine that Coq had accepted our definition, and consider how we might evaluate [approx bad 1]. We would be trying to calculate the first element in the stream [bad]. However, it is not hard to see that the definition of [bad] %``%#"#begs the question#"#%''%: unfolding the definition of [tl], we see that we essentially say %``%#"#define [bad] to equal itself#"#%''%! Of course such an equation admits no single well-defined solution, which does not fit well with the determinism of Gallina reduction.
adam@346 184
adam@346 185 Since Coq can be considered to check definitions after inlining and simplification of previously defined identifiers, the basic guardedness condition rules out our definition of [bad]. Such an inlining reduces [bad] to:
adam@346 186 [[
adam@346 187 CoFixpoint bad : stream nat := bad.
adam@346 188 ]]
adam@346 189 This is the same looping definition we rejected earlier. A similar inlining process reveals the way that Coq saw our failed definition of [map']:
adam@346 190 [[
adam@346 191 CoFixpoint map' (s : stream A) : stream B :=
adam@346 192 match s with
adam@346 193 | Cons h t => Cons (f h) (Cons (f h) (interleave (map' t) (map' t)))
adam@346 194 end.
adam@346 195 ]]
adam@346 196 Clearly in this case the [map'] calls are not immediate arguments to constructors, so we violate the guardedness condition. *)
adamc@68 197 (* end thide *)
adamc@211 198
adam@402 199 (** A more interesting question is why that condition is the right one. We can make an intuitive argument that the original [map'] definition is perfectly reasonable and denotes a well-understood transformation on streams, such that every output would behave properly with [approx]. The guardedness condition is an example of a syntactic check for%\index{productivity}% _productivity_ of co-recursive definitions. A productive definition can be thought of as one whose outputs can be forced in finite time to any finite approximation level, as with [approx]. If we replaced the guardedness condition with more involved checks, we might be able to detect and allow a broader range of productive definitions. However, mistakes in these checks could cause inconsistency, and programmers would need to understand the new, more complex checks. Coq's design strikes a balance between consistency and simplicity with its choice of guard condition, though we can imagine other worthwhile balances being struck, too. *)
adamc@62 200
adamc@63 201
adamc@63 202 (** * Infinite Proofs *)
adamc@63 203
adamc@63 204 (** Let us say we want to give two different definitions of a stream of all ones, and then we want to prove that they are equivalent. *)
adamc@63 205
adamc@63 206 CoFixpoint ones : stream nat := Cons 1 ones.
adamc@63 207 Definition ones' := map S zeroes.
adamc@63 208
adamc@63 209 (** The obvious statement of the equality is this: *)
adamc@63 210
adamc@63 211 Theorem ones_eq : ones = ones'.
adamc@63 212
adamc@63 213 (** However, faced with the initial subgoal, it is not at all clear how this theorem can be proved. In fact, it is unprovable. The [eq] predicate that we use is fundamentally limited to equalities that can be demonstrated by finite, syntactic arguments. To prove this equivalence, we will need to introduce a new relation. *)
adamc@68 214 (* begin thide *)
adamc@211 215
adamc@63 216 Abort.
adamc@63 217
adam@398 218 (** Co-inductive datatypes make sense by analogy from Haskell. What we need now is a _co-inductive proposition_. That is, we want to define a proposition whose proofs may be infinite, subject to the guardedness condition. The idea of infinite proofs does not show up in usual mathematics, but it can be very useful (unsurprisingly) for reasoning about infinite data structures. Besides examples from Haskell, infinite data and proofs will also turn out to be useful for modelling inherently infinite mathematical objects, like program executions.
adamc@63 219
adam@346 220 We are ready for our first %\index{co-inductive predicates}%co-inductive predicate. *)
adamc@63 221
adamc@63 222 Section stream_eq.
adam@351 223 Variable A : Type.
adamc@63 224
adamc@63 225 CoInductive stream_eq : stream A -> stream A -> Prop :=
adamc@63 226 | Stream_eq : forall h t1 t2,
adamc@63 227 stream_eq t1 t2
adamc@63 228 -> stream_eq (Cons h t1) (Cons h t2).
adamc@63 229 End stream_eq.
adamc@63 230
adamc@63 231 (** We say that two streams are equal if and only if they have the same heads and their tails are equal. We use the normal finite-syntactic equality for the heads, and we refer to our new equality recursively for the tails.
adamc@63 232
adamc@63 233 We can try restating the theorem with [stream_eq]. *)
adamc@63 234
adamc@63 235 Theorem ones_eq : stream_eq ones ones'.
adamc@63 236 (** Coq does not support tactical co-inductive proofs as well as it supports tactical inductive proofs. The usual starting point is the [cofix] tactic, which asks to structure this proof as a co-fixpoint. *)
adamc@211 237
adamc@63 238 cofix.
adamc@63 239 (** [[
adamc@63 240 ones_eq : stream_eq ones ones'
adamc@63 241 ============================
adamc@63 242 stream_eq ones ones'
adamc@211 243
adamc@211 244 ]]
adamc@63 245
adamc@211 246 It looks like this proof might be easier than we expected! *)
adamc@63 247
adamc@63 248 assumption.
adamc@63 249 (** [[
adamc@211 250 Proof completed.
adamc@211 251 ]]
adamc@63 252
adamc@211 253 Unfortunately, we are due for some disappointment in our victory lap.
adamc@211 254 [[
adamc@63 255 Qed.
adam@346 256 ]]
adamc@63 257
adam@346 258 <<
adamc@63 259 Error:
adamc@63 260 Recursive definition of ones_eq is ill-formed.
adamc@63 261
adamc@63 262 In environment
adamc@63 263 ones_eq : stream_eq ones ones'
adamc@63 264
adamc@205 265 unguarded recursive call in "ones_eq"
adam@346 266 >>
adamc@205 267
adamc@211 268 Via the Curry-Howard correspondence, the same guardedness condition applies to our co-inductive proofs as to our co-inductive data structures. We should be grateful that this proof is rejected, because, if it were not, the same proof structure could be used to prove any co-inductive theorem vacuously, by direct appeal to itself!
adamc@63 269
adam@347 270 Thinking about how Coq would generate a proof term from the proof script above, we see that the problem is that we are violating the guardedness condition. During our proofs, Coq can help us check whether we have yet gone wrong in this way. We can run the command [Guarded] in any context to see if it is possible to finish the proof in a way that will yield a properly guarded proof term.%\index{Vernacular commands!Guarded}%
adamc@63 271 [[
adamc@63 272 Guarded.
adamc@205 273 ]]
adamc@205 274
adam@398 275 Running [Guarded] here gives us the same error message that we got when we tried to run [Qed]. In larger proofs, [Guarded] can be helpful in detecting problems _before_ we think we are ready to run [Qed].
adamc@63 276
adam@281 277 We need to start the co-induction by applying [stream_eq]'s constructor. To do that, we need to know that both arguments to the predicate are [Cons]es. Informally, this is trivial, but [simpl] is not able to help us. *)
adamc@63 278
adamc@63 279 Undo.
adamc@63 280 simpl.
adamc@63 281 (** [[
adamc@63 282 ones_eq : stream_eq ones ones'
adamc@63 283 ============================
adamc@63 284 stream_eq ones ones'
adamc@211 285
adamc@211 286 ]]
adamc@63 287
adamc@211 288 It turns out that we are best served by proving an auxiliary lemma. *)
adamc@211 289
adamc@63 290 Abort.
adamc@63 291
adamc@63 292 (** First, we need to define a function that seems pointless on first glance. *)
adamc@63 293
adamc@63 294 Definition frob A (s : stream A) : stream A :=
adamc@63 295 match s with
adamc@63 296 | Cons h t => Cons h t
adamc@63 297 end.
adamc@63 298
adamc@63 299 (** Next, we need to prove a theorem that seems equally pointless. *)
adamc@63 300
adamc@63 301 Theorem frob_eq : forall A (s : stream A), s = frob s.
adamc@63 302 destruct s; reflexivity.
adamc@63 303 Qed.
adamc@63 304
adamc@63 305 (** But, miraculously, this theorem turns out to be just what we needed. *)
adamc@63 306
adamc@63 307 Theorem ones_eq : stream_eq ones ones'.
adamc@63 308 cofix.
adamc@63 309
adamc@63 310 (** We can use the theorem to rewrite the two streams. *)
adamc@211 311
adamc@63 312 rewrite (frob_eq ones).
adamc@63 313 rewrite (frob_eq ones').
adamc@63 314 (** [[
adamc@63 315 ones_eq : stream_eq ones ones'
adamc@63 316 ============================
adamc@63 317 stream_eq (frob ones) (frob ones')
adamc@211 318
adamc@211 319 ]]
adamc@63 320
adamc@211 321 Now [simpl] is able to reduce the streams. *)
adamc@63 322
adamc@63 323 simpl.
adamc@63 324 (** [[
adamc@63 325 ones_eq : stream_eq ones ones'
adamc@63 326 ============================
adamc@63 327 stream_eq (Cons 1 ones)
adamc@63 328 (Cons 1
adamc@63 329 ((cofix map (s : stream nat) : stream nat :=
adamc@63 330 match s with
adamc@63 331 | Cons h t => Cons (S h) (map t)
adamc@63 332 end) zeroes))
adamc@211 333
adamc@211 334 ]]
adamc@63 335
adam@346 336 Note that [cofix] notation for anonymous co-recursion, which is analogous to the [fix] notation we have already seen for recursion. Since we have exposed the [Cons] structure of each stream, we can apply the constructor of [stream_eq]. *)
adamc@63 337
adamc@63 338 constructor.
adamc@63 339 (** [[
adamc@63 340 ones_eq : stream_eq ones ones'
adamc@63 341 ============================
adamc@63 342 stream_eq ones
adamc@63 343 ((cofix map (s : stream nat) : stream nat :=
adamc@63 344 match s with
adamc@63 345 | Cons h t => Cons (S h) (map t)
adamc@63 346 end) zeroes)
adamc@211 347
adamc@211 348 ]]
adamc@63 349
adamc@211 350 Now, modulo unfolding of the definition of [map], we have matched our assumption. *)
adamc@211 351
adamc@63 352 assumption.
adamc@63 353 Qed.
adamc@63 354
adamc@63 355 (** Why did this silly-looking trick help? The answer has to do with the constraints placed on Coq's evaluation rules by the need for termination. The [cofix]-related restriction that foiled our first attempt at using [simpl] is dual to a restriction for [fix]. In particular, an application of an anonymous [fix] only reduces when the top-level structure of the recursive argument is known. Otherwise, we would be unfolding the recursive definition ad infinitum.
adamc@63 356
adam@398 357 Fixpoints only reduce when enough is known about the _definitions_ of their arguments. Dually, co-fixpoints only reduce when enough is known about _how their results will be used_. In particular, a [cofix] is only expanded when it is the discriminee of a [match]. Rewriting with our superficially silly lemma wrapped new [match]es around the two [cofix]es, triggering reduction.
adamc@63 358
adamc@63 359 If [cofix]es reduced haphazardly, it would be easy to run into infinite loops in evaluation, since we are, after all, building infinite objects.
adamc@63 360
adamc@63 361 One common source of difficulty with co-inductive proofs is bad interaction with standard Coq automation machinery. If we try to prove [ones_eq'] with automation, like we have in previous inductive proofs, we get an invalid proof. *)
adamc@63 362
adamc@63 363 Theorem ones_eq' : stream_eq ones ones'.
adamc@63 364 cofix; crush.
adamc@63 365 (** [[
adamc@205 366 Guarded.
adam@302 367 ]]
adam@302 368 *)
adamc@63 369 Abort.
adam@346 370
adam@346 371 (** The standard [auto] machinery sees that our goal matches an assumption and so applies that assumption, even though this violates guardedness. One usually starts a proof like this by [destruct]ing some parameter and running a custom tactic to figure out the first proof rule to apply for each case. Alternatively, there are tricks that can be played with %``%#"#hiding#"#%''% the co-inductive hypothesis.
adam@346 372
adam@346 373 %\medskip%
adam@346 374
adam@402 375 Must we always be cautious with automation in proofs by co-induction? Induction seems to have dual versions of the same pitfalls inherent in it, and yet we avoid those pitfalls by encapsulating safe Curry-Howard recursion schemes inside named induction principles. It turns out that we can usually do the same with%\index{co-induction principles}% _co-induction principles_. Let us take that tack here, so that we can arrive at an [induction x; crush]-style proof for [ones_eq'].
adam@346 376
adam@398 377 An induction principle is parameterized over a predicate characterizing what we mean to prove, _as a function of the inductive fact that we already know_. Dually, a co-induction principle ought to be parameterized over a predicate characterizing what we mean to prove, _as a function of the arguments to the co-inductive predicate that we are trying to prove_.
adam@346 378
adam@346 379 To state a useful principle for [stream_eq], it will be useful first to define the stream head function. *)
adam@346 380
adam@346 381 Definition hd A (s : stream A) : A :=
adam@346 382 match s with
adam@346 383 | Cons x _ => x
adam@346 384 end.
adam@346 385
adam@346 386 (** Now we enter a section for the co-induction principle, based on %\index{Park's principle}%Park's principle as introduced in a tutorial by Gim%\'%enez%~\cite{IT}%. *)
adam@346 387
adam@346 388 Section stream_eq_coind.
adam@351 389 Variable A : Type.
adam@346 390 Variable R : stream A -> stream A -> Prop.
adam@392 391 (** This relation generalizes the theorem we want to prove, characterizing exactly which two arguments to [stream_eq] we want to consider. *)
adam@346 392
adam@346 393 Hypothesis Cons_case_hd : forall s1 s2, R s1 s2 -> hd s1 = hd s2.
adam@346 394 Hypothesis Cons_case_tl : forall s1 s2, R s1 s2 -> R (tl s1) (tl s2).
adam@402 395 (** Two hypotheses characterize what makes a good choice of [R]: it enforces equality of stream heads, and it is %``%#<i>#hereditary#</i>#%''% in the sense that an [R] stream pair passes on %``%#"#[R]-ness#"#%''% to its tails. An established technical term for such a relation is%\index{bisimulation}% _bisimulation_. *)
adam@346 396
adam@346 397 (** Now it is straightforward to prove the principle, which says that any stream pair in [R] is equal. The reader may wish to step through the proof script to see what is going on. *)
adam@392 398
adam@346 399 Theorem stream_eq_coind : forall s1 s2, R s1 s2 -> stream_eq s1 s2.
adam@346 400 cofix; destruct s1; destruct s2; intro.
adam@346 401 generalize (Cons_case_hd H); intro Heq; simpl in Heq; rewrite Heq.
adam@346 402 constructor.
adam@346 403 apply stream_eq_coind.
adam@346 404 apply (Cons_case_tl H).
adam@346 405 Qed.
adam@346 406 End stream_eq_coind.
adam@346 407
adam@346 408 (** To see why this proof is guarded, we can print it and verify that the one co-recursive call is an immediate argument to a constructor. *)
adam@392 409
adam@346 410 Print stream_eq_coind.
adam@346 411
adam@346 412 (** We omit the output and proceed to proving [ones_eq''] again. The only bit of ingenuity is in choosing [R], and in this case the most restrictive predicate works. *)
adam@346 413
adam@346 414 Theorem ones_eq'' : stream_eq ones ones'.
adam@346 415 apply (stream_eq_coind (fun s1 s2 => s1 = ones /\ s2 = ones')); crush.
adam@346 416 Qed.
adam@346 417
adam@346 418 (** Note that this proof achieves the proper reduction behavior via [hd] and [tl], rather than [frob] as in the last proof. All three functions pattern match on their arguments, catalyzing computation steps.
adam@346 419
adam@346 420 Compared to the inductive proofs that we are used to, it still seems unsatisfactory that we had to write out a choice of [R] in the last proof. An alternate is to capture a common pattern of co-recursion in a more specialized co-induction principle. For the current example, that pattern is: prove [stream_eq s1 s2] where [s1] and [s2] are defined as their own tails. *)
adam@346 421
adam@346 422 Section stream_eq_loop.
adam@351 423 Variable A : Type.
adam@346 424 Variables s1 s2 : stream A.
adam@346 425
adam@346 426 Hypothesis Cons_case_hd : hd s1 = hd s2.
adam@346 427 Hypothesis loop1 : tl s1 = s1.
adam@346 428 Hypothesis loop2 : tl s2 = s2.
adam@346 429
adam@346 430 (** The proof of the principle includes a choice of [R], so that we no longer need to make such choices thereafter. *)
adam@346 431
adam@346 432 Theorem stream_eq_loop : stream_eq s1 s2.
adam@346 433 apply (stream_eq_coind (fun s1' s2' => s1' = s1 /\ s2' = s2)); crush.
adam@346 434 Qed.
adam@346 435 End stream_eq_loop.
adam@346 436
adam@346 437 Theorem ones_eq''' : stream_eq ones ones'.
adam@346 438 apply stream_eq_loop; crush.
adam@346 439 Qed.
adamc@68 440 (* end thide *)
adamc@63 441
adam@346 442 (** Let us put [stream_eq_ind] through its paces a bit more, considering two different ways to compute infinite streams of all factorial values. First, we import the [fact] factorial function from the standard library. *)
adam@346 443
adam@346 444 Require Import Arith.
adam@346 445 Print fact.
adam@346 446 (** %\vspace{-.15in}%[[
adam@346 447 fact =
adam@346 448 fix fact (n : nat) : nat :=
adam@346 449 match n with
adam@346 450 | 0 => 1
adam@346 451 | S n0 => S n0 * fact n0
adam@346 452 end
adam@346 453 : nat -> nat
adam@346 454 ]]
adam@346 455 *)
adam@346 456
adam@346 457 (** The simplest way to compute the factorial stream involves calling [fact] afresh at each position. *)
adam@346 458
adam@346 459 CoFixpoint fact_slow' (n : nat) := Cons (fact n) (fact_slow' (S n)).
adam@346 460 Definition fact_slow := fact_slow' 1.
adam@346 461
adam@346 462 (** A more clever, optimized method maintains an accumulator of the previous factorial, so that each new entry can be computed with a single multiplication. *)
adam@346 463
adam@346 464 CoFixpoint fact_iter' (cur acc : nat) := Cons acc (fact_iter' (S cur) (acc * cur)).
adam@346 465 Definition fact_iter := fact_iter' 2 1.
adam@346 466
adam@346 467 (** We can verify that the streams are equal up to particular finite bounds. *)
adam@346 468
adam@346 469 Eval simpl in approx fact_iter 5.
adam@346 470 (** %\vspace{-.15in}%[[
adam@346 471 = 1 :: 2 :: 6 :: 24 :: 120 :: nil
adam@346 472 : list nat
adam@346 473 ]]
adam@346 474 *)
adam@346 475 Eval simpl in approx fact_slow 5.
adam@346 476 (** %\vspace{-.15in}%[[
adam@346 477 = 1 :: 2 :: 6 :: 24 :: 120 :: nil
adam@346 478 : list nat
adam@346 479 ]]
adam@346 480
adam@346 481 Now, to prove that the two versions are equivalent, it is helpful to prove (and add as a proof hint) a quick lemma about the computational behavior of [fact]. *)
adam@346 482
adam@346 483 (* begin thide *)
adam@346 484 Lemma fact_def : forall x n,
adam@346 485 fact_iter' x (fact n * S n) = fact_iter' x (fact (S n)).
adam@346 486 simpl; intros; f_equal; ring.
adam@346 487 Qed.
adam@346 488
adam@346 489 Hint Resolve fact_def.
adam@346 490
adam@346 491 (** With the hint added, it is easy to prove an auxiliary lemma relating [fact_iter'] and [fact_slow']. The key bit of ingenuity is introduction of an existential quantifier for the shared parameter [n]. *)
adam@346 492
adam@346 493 Lemma fact_eq' : forall n, stream_eq (fact_iter' (S n) (fact n)) (fact_slow' n).
adam@346 494 intro; apply (stream_eq_coind (fun s1 s2 => exists n, s1 = fact_iter' (S n) (fact n)
adam@346 495 /\ s2 = fact_slow' n)); crush; eauto.
adam@346 496 Qed.
adam@346 497
adam@346 498 (** The final theorem is a direct corollary of [fact_eq']. *)
adam@346 499
adam@346 500 Theorem fact_eq : stream_eq fact_iter fact_slow.
adam@346 501 apply fact_eq'.
adam@346 502 Qed.
adam@346 503
adam@346 504 (** As in the case of [ones_eq'], we may be unsatisfied that we needed to write down a choice of [R] that seems to duplicate information already present in a lemma statement. We can facilitate a simpler proof by defining a co-induction principle specialized to goals that begin with single universal quantifiers, and the strategy can be extended in a straightforward way to principles for other counts of quantifiers. (Our [stream_eq_loop] principle is effectively the instantiation of this technique to zero quantifiers.) *)
adam@346 505
adam@346 506 Section stream_eq_onequant.
adam@351 507 Variables A B : Type.
adam@346 508 (** We have the types [A], the domain of the one quantifier; and [B], the type of data found in the streams. *)
adam@346 509
adam@346 510 Variables f g : A -> stream B.
adam@346 511 (** The two streams we compare must be of the forms [f x] and [g x], for some shared [x]. Note that this falls out naturally when [x] is a shared universally quantified variable in a lemma statement. *)
adam@346 512
adam@346 513 Hypothesis Cons_case_hd : forall x, hd (f x) = hd (g x).
adam@346 514 Hypothesis Cons_case_tl : forall x, exists y, tl (f x) = f y /\ tl (g x) = g y.
adam@346 515 (** These conditions are inspired by the bisimulation requirements, with a more general version of the [R] choice we made for [fact_eq'] inlined into the hypotheses of [stream_eq_coind]. *)
adam@346 516
adam@346 517 Theorem stream_eq_onequant : forall x, stream_eq (f x) (g x).
adam@346 518 intro; apply (stream_eq_coind (fun s1 s2 => exists x, s1 = f x /\ s2 = g x)); crush; eauto.
adam@346 519 Qed.
adam@346 520 End stream_eq_onequant.
adam@346 521
adam@346 522 Lemma fact_eq'' : forall n, stream_eq (fact_iter' (S n) (fact n)) (fact_slow' n).
adam@346 523 apply stream_eq_onequant; crush; eauto.
adam@346 524 Qed.
adam@346 525
adam@346 526 (** We have arrived at one of our customary automated proofs, thanks to the new principle. *)
adam@346 527 (* end thide *)
adamc@64 528
adamc@64 529
adamc@64 530 (** * Simple Modeling of Non-Terminating Programs *)
adamc@64 531
adam@402 532 (** We close the chapter with a quick motivating example for more complex uses of co-inductive types. We will define a co-inductive semantics for a simple imperative programming language and use that semantics to prove the correctness of a trivial optimization that removes spurious additions by 0. We follow the technique of%\index{co-inductive big-step operational semantics}% _co-inductive big-step operational semantics_ %\cite{BigStep}%.
adamc@64 533
adam@347 534 We define a suggestive synonym for [nat], as we will consider programs with infinitely many variables, represented as [nat]s. *)
adamc@211 535
adam@347 536 Definition var := nat.
adamc@64 537
adam@347 538 (** We define a type [vars] of maps from variables to values. To define a function [set] for setting a variable's value in a map, we import the [Arith] module from Coq's standard library, and we use its function [beq_nat] for comparing natural numbers. *)
adamc@64 539
adam@347 540 Definition vars := var -> nat.
adam@347 541 Require Import Arith.
adam@347 542 Definition set (vs : vars) (v : var) (n : nat) : vars :=
adam@347 543 fun v' => if beq_nat v v' then n else vs v'.
adamc@67 544
adam@347 545 (** We define a simple arithmetic expression language with variables, and we give it a semantics via an interpreter. *)
adamc@67 546
adam@347 547 Inductive exp : Set :=
adam@347 548 | Const : nat -> exp
adam@347 549 | Var : var -> exp
adam@347 550 | Plus : exp -> exp -> exp.
adamc@64 551
adam@347 552 Fixpoint evalExp (vs : vars) (e : exp) : nat :=
adam@347 553 match e with
adam@347 554 | Const n => n
adam@347 555 | Var v => vs v
adam@347 556 | Plus e1 e2 => evalExp vs e1 + evalExp vs e2
adam@347 557 end.
adamc@64 558
adam@347 559 (** Finally, we define a language of commands. It includes variable assignment, sequencing, and a %\texttt{%#<tt>#while#</tt>#%}% form that repeats as long as its test expression evaluates to a nonzero value. *)
adamc@64 560
adam@347 561 Inductive cmd : Set :=
adam@347 562 | Assign : var -> exp -> cmd
adam@347 563 | Seq : cmd -> cmd -> cmd
adam@347 564 | While : exp -> cmd -> cmd.
adamc@64 565
adam@398 566 (** We could define an inductive relation to characterize the results of command evaluation. However, such a relation would not capture _nonterminating_ executions. With a co-inductive relation, we can capture both cases. The parameters of the relation are an initial state, a command, and a final state. A program that does not terminate in a particular initial state is related to _any_ final state. *)
adamc@67 567
adam@347 568 CoInductive evalCmd : vars -> cmd -> vars -> Prop :=
adam@347 569 | EvalAssign : forall vs v e, evalCmd vs (Assign v e) (set vs v (evalExp vs e))
adam@347 570 | EvalSeq : forall vs1 vs2 vs3 c1 c2, evalCmd vs1 c1 vs2
adam@347 571 -> evalCmd vs2 c2 vs3
adam@347 572 -> evalCmd vs1 (Seq c1 c2) vs3
adam@347 573 | EvalWhileFalse : forall vs e c, evalExp vs e = 0
adam@347 574 -> evalCmd vs (While e c) vs
adam@347 575 | EvalWhileTrue : forall vs1 vs2 vs3 e c, evalExp vs1 e <> 0
adam@347 576 -> evalCmd vs1 c vs2
adam@347 577 -> evalCmd vs2 (While e c) vs3
adam@347 578 -> evalCmd vs1 (While e c) vs3.
adam@347 579
adam@347 580 (** Having learned our lesson in the last section, before proceeding, we build a co-induction principle for [evalCmd]. *)
adam@347 581
adam@347 582 Section evalCmd_coind.
adam@347 583 Variable R : vars -> cmd -> vars -> Prop.
adam@347 584
adam@347 585 Hypothesis AssignCase : forall vs1 vs2 v e, R vs1 (Assign v e) vs2
adam@347 586 -> vs2 = set vs1 v (evalExp vs1 e).
adam@347 587
adam@347 588 Hypothesis SeqCase : forall vs1 vs3 c1 c2, R vs1 (Seq c1 c2) vs3
adam@347 589 -> exists vs2, R vs1 c1 vs2 /\ R vs2 c2 vs3.
adam@347 590
adam@347 591 Hypothesis WhileCase : forall vs1 vs3 e c, R vs1 (While e c) vs3
adam@347 592 -> (evalExp vs1 e = 0 /\ vs3 = vs1)
adam@347 593 \/ exists vs2, evalExp vs1 e <> 0 /\ R vs1 c vs2 /\ R vs2 (While e c) vs3.
adam@347 594
adam@402 595 (** The proof is routine. We make use of a form of %\index{tactics!destruct}%[destruct] that takes an%\index{intro pattern}% _intro pattern_ in an [as] clause. These patterns control how deeply we break apart the components of an inductive value, and we refer the reader to the Coq manual for more details. *)
adam@347 596
adam@347 597 Theorem evalCmd_coind : forall vs1 c vs2, R vs1 c vs2 -> evalCmd vs1 c vs2.
adam@347 598 cofix; intros; destruct c.
adam@347 599 rewrite (AssignCase H); constructor.
adam@347 600 destruct (SeqCase H) as [? [? ?]]; econstructor; eauto.
adam@347 601 destruct (WhileCase H) as [[? ?] | [? [? [? ?]]]]; subst;
adam@347 602 [ econstructor | econstructor 4 ]; eauto.
adam@347 603 Qed.
adam@347 604 End evalCmd_coind.
adam@347 605
adam@347 606 (** Now that we have a co-induction principle, we should use it to prove something! Our example is a trivial program optimizer that finds places to replace [0 + e] with [e]. *)
adam@347 607
adam@347 608 Fixpoint optExp (e : exp) : exp :=
adam@347 609 match e with
adam@347 610 | Plus (Const 0) e => optExp e
adam@347 611 | Plus e1 e2 => Plus (optExp e1) (optExp e2)
adam@347 612 | _ => e
adam@347 613 end.
adam@347 614
adam@347 615 Fixpoint optCmd (c : cmd) : cmd :=
adam@347 616 match c with
adam@347 617 | Assign v e => Assign v (optExp e)
adam@347 618 | Seq c1 c2 => Seq (optCmd c1) (optCmd c2)
adam@347 619 | While e c => While (optExp e) (optCmd c)
adam@347 620 end.
adam@347 621
adam@347 622 (** Before proving correctness of [optCmd], we prove a lemma about [optExp]. This is where we have to do the most work, choosing pattern match opportunities automatically. *)
adam@347 623
adam@347 624 (* begin thide *)
adam@347 625 Lemma optExp_correct : forall vs e, evalExp vs (optExp e) = evalExp vs e.
adam@347 626 induction e; crush;
adam@347 627 repeat (match goal with
adam@347 628 | [ |- context[match ?E with Const _ => _ | Var _ => _
adam@347 629 | Plus _ _ => _ end] ] => destruct E
adam@347 630 | [ |- context[match ?E with O => _ | S _ => _ end] ] => destruct E
adam@347 631 end; crush).
adamc@64 632 Qed.
adamc@64 633
adam@375 634 Hint Rewrite optExp_correct .
adamc@64 635
adam@384 636 (** The final theorem is easy to establish, using our co-induction principle and a bit of Ltac smarts that we leave unexplained for now. Curious readers can consult the Coq manual, or wait for the later chapters of this book about proof automation. At a high level, we show inclusions between behaviors, going in both directions between original and optimized programs. *)
adamc@64 637
adam@384 638 Ltac finisher := match goal with
adam@384 639 | [ H : evalCmd _ _ _ |- _ ] => ((inversion H; [])
adam@384 640 || (inversion H; [|])); subst
adam@384 641 end; crush; eauto 10.
adam@384 642
adam@384 643 Lemma optCmd_correct1 : forall vs1 c vs2, evalCmd vs1 c vs2
adam@347 644 -> evalCmd vs1 (optCmd c) vs2.
adam@347 645 intros; apply (evalCmd_coind (fun vs1 c' vs2 => exists c, evalCmd vs1 c vs2
adam@347 646 /\ c' = optCmd c)); eauto; crush;
adam@347 647 match goal with
adam@347 648 | [ H : _ = optCmd ?E |- _ ] => destruct E; simpl in *; discriminate
adam@347 649 || injection H; intros; subst
adam@384 650 end; finisher.
adam@384 651 Qed.
adam@384 652
adam@384 653 Lemma optCmd_correct2 : forall vs1 c vs2, evalCmd vs1 (optCmd c) vs2
adam@384 654 -> evalCmd vs1 c vs2.
adam@384 655 intros; apply (evalCmd_coind (fun vs1 c vs2 => evalCmd vs1 (optCmd c) vs2));
adam@384 656 crush; finisher.
adam@384 657 Qed.
adam@384 658
adam@384 659 Theorem optCmd_correct : forall vs1 c vs2, evalCmd vs1 (optCmd c) vs2
adam@384 660 <-> evalCmd vs1 c vs2.
adam@384 661 intuition; apply optCmd_correct1 || apply optCmd_correct2; assumption.
adam@347 662 Qed.
adam@347 663 (* end thide *)
adamc@64 664
adam@384 665 (** In this form, the theorem tells us that the optimizer preserves observable behavior of both terminating and nonterminating programs, but we did not have to do more work than for the case of terminating programs alone. We merely took the natural inductive definition for terminating executions, made it co-inductive, and applied the appropriate co-induction principle. Curious readers might experiment with adding command constructs like %\texttt{%#<tt>#if#</tt>#%}%; the same proof script should continue working, after the co-induction principle is extended to the new evaluation rules. *)