wball/perga
1
0
Fork 0
Code Issues 15 Pull requests Projects Releases Packages Wiki Activity Actions

HUGE update

Browse source
This commit is contained in:
William Ball 2024-11-17 18:33:14 -08:00
parent c1ccd50644
commit 8c5311a2f6
16 changed files with 689 additions and 344 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

158
README.md
View file

@ -1,88 +1,108 @@
A basic implementation of a dependently typed lambda calculus (calculus of constructions) based on the exposition in Nederpelt and Geuvers *Type Theory and Formal Proof*. Right now it is *technically* a perfectly capable higher order logic proof checker, though there is lots of room for improved ergonomics and usability, which I intend to work on.
`perga` is a basic proof assistant based on a dependently typed lambda calculus (calculus of constructions). This implementation is based on the exposition in Nederpelt and Geuvers’ *Type Theory and Formal Proof*. Right now it is a perfectly capable higher order logic proof checker, though there is lots of room for improved ergonomics and usability, which I intend to work on. At the moment, `perga` is comparable to Automath in terms of power and ease of use, being slightly more powerful than Automath (though lacks a *primitive notion* system for the moment), and a touch less ergonomic.
# Syntax
The syntax is fairly flexible and should work as you expect. Identifiers can be Unicode as long as `megaparsec` calls them alphanumeric. Lambda and Pi abstractions can be written in many obvious ways that should be clear from the examples below. Additionally, arrows can be used as an abbreviation for a Π type where the parameter doesn’t appear in the body as usual.
The syntax is fairly flexible and should work as you expect. Identifiers can be Unicode as long as `megaparsec` calls them alphanumeric. `λ` and `\Pi` abstractions can be written in many obvious ways that should be clear from the examples below. Additionally, arrows can be used as an abbreviation for a Π type where the parameter doesn’t appear in the body as usual.
All of the following examples correctly parse, and should look familiar if you are used to standard lambda calculus notation or Coq syntax.
All of the following example terms correctly parse, and should look familiar if you are used to standard lambda calculus notation or Coq syntax.
λ (α : *) . λ (β : *) . λ (x : α) . λ (y : β) . x
fun (A B C : *) (g : → C) (f : A → B) (x : A) ⇒ g (f x)
fun (S : *) (P Q : S -> *) (H : Π (x : S) . P x -> Q x) (HP : forall (x : S), P x) => fun (x : S) => H x (HP x)
I mostly stick to Coq syntax throughout this file and the examples, as that is what I’m most used to and is easiest to type. I will probably make the syntax more strict in the future, as this level of flexibility is really not necessary.
# Basic Usage
Definitions work similarly, having abstract syntax as shown below.
For the moment, running the program drops you into a repl where you can enter terms in the calculus of construction, which the system will then type check and report the type of the entered term, or any errors if present.
<ident> (<ident> : <type>)* : <type>? := <term>;
(The distinction between `<type>` and `<term>` is for emphasis; they are the exact same syntactic category.) Here&rsquo;s a couple definitions of the `const` function from above showing the options with the syntax.
const := λ (α : *) . λ (β : *) . λ (x : α) . λ (y : β) . x;
const : forall (α β : *), α → β → α := fun (α β : *) (x : α) (y : β) ⇒ x;
const (α β : *) (x : α) (y : β) : α := x;
Type ascriptions are optional. If included, `perga` will check to make sure your definition matches the ascription, and, if so, will remember the way your wrote the type when printing inferred types, which is particularly handy when using abbreviations for complex types. `perga` has no problem inferring the types of top-level definitions, as they are completely determined by the term, but I recommend including ascriptions most of the time, as they serve as a nice piece of documentation, help guide the implementation process, and make sure you are implementing the type you think you are.
Line comments are `--` like in Haskell, and block comments are `(* *)` like ML (and nest properly). There is no significant whitespace, so you are free to format code as you wish.
## Sample Session
# Usage
Here&rsquo;s a sample session. Suppose your goal is to prove that for every set $S$ and pair of propositions $P$ and $Q$ on $S$, if $\forall x \in S, P x$ holds, and $\forall x \in S, P x \Rightarrow Q x$, then $\forall x \in S, Q x$. A set $S$ corresponds to a type `S`, so our term will begin
Running `perga` without any arguments drops you into a basic repl. From here, you can type in definitions which `perga` will typecheck. Previous definitions are accessible in future definitions. The usual readline keybindings are available, including navigating history, which is saved between sessions (in `~/.cache/perga/history`). In the repl, you can enter &ldquo;:q&rdquo;, press C-c, or press C-d to quit. Entering &ldquo;:e&rdquo; shows everything that has been defined along with their types. Entering &ldquo;:t <ident>&rdquo; prints the type of a particular identifier.
> fun (S : *)
Likewise propositions are functions from `S → *`, so we can continue
> fun (S : *) (P Q : S → *)
Since $\forall$ corresponds with `Π`, the hypothesis $\forall x \in S, P x$ would correspond to the type `Π (x : S) . P x` and the hypothesis $\forall x \in S, P x \Rightarrow Q x$ would correspond with `Π (x : S) . P x → Q x`. Thus we can continue
> fun (S : *) (P Q : S → *) (HP : forall (x : S), P x) (H : forall (x : S), P x → Q x)
Since we are trying to prove a universally quantified proposition, we need to introduce an `x : S`, so
> fun (S : *) (P Q : S → *) (HP : forall (x : S), P x) (H : forall (x : S), P x → Q x) (x : S)
Finally, all we have left to do is find something of type `Q x`. We can get to `Q x` using `H x` if we can find something of type `P x`. Fortunately, `HP x` is type `P x`, so our final term is
> fun (S : *) (P Q : S → *) (HP : forall (x : S), P x) (H : forall (x : S), P x → Q x) (x : S) ⇒ H x (HP x)
Pressing enter to send this term to the system, it responds with
λ (S : *) (P Q : S -> *) (HP : ∏ (x : S) . P x) (H : ∏ (x : S) . P x -> Q x) (x : S) . H x (HP x) : ∏ (S : *) (P Q : S -> *) . (∏ (x : S) . P x) -> (∏ (x : S) . P x -> Q x) -> ∏ (x : S) . Q x
This output is a bit hard to read, but it is ultimately in the form `term : type`. The `term` is, up to minor syntactic differences, the term we entered, and the `type` is the type of the term inferred by the system. As a nice sanity check, we can verify that the `type` indeed corresponds to the theorem we wanted to prove.
More complex and interesting proofs and theorems are technically possible (in fact, *all* interesting theorems and proofs are possible, for a certain definition of *interesting*, *theorem*, and *proof*), though practically infeasible without definitions.
You can also give `perga` a filename as an argument, in which case it will typecheck every definition in the file. Upon finishing, which should be nearly instantaneous, it will print out all the definitions it parsed along with their types (like you had typed &ldquo;:e&rdquo; in the repl) so you can verify that it worked.
# Goals
# Simple Example
A much larger, commented example is located in <./examples/example.pg>. Here is an example file defining Leibniz equality and proving that it is reflexive, symmetric, and transitive.
-- file: equality.pg
-- Defining Leibniz equality
-- Note that we can leave the ascription off
eq (A : *) (x y : A) := forall (P : A -> *), P x -> P y;
-- Equality is reflexive, which is easy to prove
-- Here we give an ascription so that when `perga` reports the type,
-- it references `eq` rather than inferring the type.
eq_refl (A : *) (x : A) : eq A x x := fun (P : A -> *) (Hx : P x) => Hx;
-- Equality is symmetric. This one's a little harder to prove.
eq_sym (A : *) (x y : A) (Hxy : eq A x y) : eq A y x :=
fun (P : A -> *) (Hy : P y) =>
Hxy (fun (z : A) => P z -> P x) (fun (Hx : P x) => Hx) Hy;
-- Equality is transitive.
eq_trans (A : *) (x y z : A) (Hxy : eq A x y) (Hyz : eq A y z) : eq A x z :=
fun (P : A -> *) (Hx : P x) => Hyz P (Hxy P Hx);
Running `perga equality.pg` yields the following output.
eq : ∏ (A : *) . A -> A -> *
eq_refl : ∏ (A : *) (x : A) . eq A x x
eq_sym : ∏ (A : *) (x y : A) . eq A x y -> eq A y x
eq_trans : ∏ (A : *) (x y z : A) . eq A x y -> eq A y z -> eq A x z
This means our proofs were accepted. Furthermore, as a sanity check, we can see that the types correspond exactly to what we wanted to prove.
# Future Goals
## Substantive
### TODO DEFINITIONS
### TODO Let-expressions
Some kind of definition system would be a difficult and substantial addition, but man is it necessary to do anything. Likewise, I&rsquo;d probably want a way to define primitive notions/axioms, but that should be an easy extension of the definition system. Further following *Type Theory and Formal Proof* would additionally yield a nice context system, which would be handy, though I disagree with the choice to differentiate between parameterized definitions and functions. That distinction doesn&rsquo;t really make sense in full calculus of constructions.
Sidenote: With a definition system, I would greatly prefer Haskell-style type annotations to ML-style type annotations, though the latter are likely way easier to implement. It serves as a nice bit of documentation, de-clutters the function definition, and follows the familiar mathematical statement-proof structure.
1. TODO Add support for named free variables to `Expr`
I&rsquo;m taking the locally nameless approach. Dealing with free variables as de Bruijn indices greater than the context sounds terrible. This requires adding support for &delta; reductions, as well as performing reductions and type checking operations in an environment. NOTE: I&rsquo;ll probably want to use the [Reader monad](https://hackage.haskell.org/package/mtl-2.3.1/docs/Control-Monad-Reader.html).
I vaguely remember from reading Coq&rsquo;art that Coq does something special with `let` expressions, so I&rsquo;ll maybe want to check that out. I tried implementing `let` as syntax sugar for an immediately called function, but that proved to be a massive mess with how I&rsquo;m handling things. `let` expressions would definitely be handy for factoring out complex sub expressions.
### TODO Inference
Obviously not fully decidable, but I might be able to implement some basic unification algorithm. This isn&rsquo;t super necessary though, I find leaving off the types of arguments to generally be a bad idea, but in some cases it can be worth it.
Obviously not fully decidable, but I might be able to implement some basic unification algorithm. This isn&rsquo;t super necessary though, I find leaving off the types of arguments to generally be a bad idea, but in some cases it can be handy, especially not at the top level.
### TODO Implicits
Much, much more useful than [inference](#orgb32d6a9), implicit arguments would be amazing. It also seems a lot more complicated, but any system for dealing with implicit arguments is far better than none.
Much, much more useful than [inference](#org79d490c), implicit arguments would be amazing. It also seems a lot more complicated, but any system for dealing with implicit arguments is far better than none. Getting rid of stuff like [lines 213-215 of the example file](./examples/example.pg) would be amazing.
### TODO Module System
A proper module system would be wonderful. To me, ML style modules with structures, signatures, and functors seems like the right way to handle algebraic structures for a relatively simple language, rather than records (or a bunch of `and`&rsquo;s like I currently have) or type classes (probably much harder, but could be nicer), but any way of managing scope, importing files, etc. is a necessity.
### TODO Universes?
Not really necessary without [inductive definitions](#orgc763d54), but could be fun.
Not really all that necessary, especially without [inductive definitions](#org357916d), but could be fun.
### TODO Inductive Definitions
This is definitely a stretch goal. It would be cool though, and would turn this proof checker into a much more competent programming language. It&rsquo;s not necessary for the math, but inductive definitions let you leverage computation in proofs, which is amazing. They also make certain definitions easier, by avoiding needing to manually stipulate elimination rules, including induction principles.
This is definitely a stretch goal. It would be cool though, and would turn this proof checker into a much more competent programming language. It&rsquo;s not necessary for the math, but inductive definitions let you leverage computation in proofs, which is amazing. They also make certain definitions way easier, by avoiding needing to manually stipulate elimination rules, including induction principles, and let you keep more math constructive and understandable to the computer.
## Cosmetic/usage/technical
@ -93,36 +113,56 @@ This is definitely a stretch goal. It would be cool though, and would turn this
Right now, everything defaults to one line, which can be a problem with how large the proof terms get. Probably want to use [prettyprinter](https://hackage.haskell.org/package/prettyprinter) to be able to nicely handle indentation and line breaks.
### DONE Better usage
### TODO Better repl
Read input from a file if a filename is given, otherwise drop into a repl. Perhaps use something like [haskeline](https://hackage.haskell.org/package/haskeline) to improve the repl (though `rlwrap` is fine, so not a huge priority).
For even better usage in the future, check out [optparse](https://hackage.haskell.org/package/optparse-applicative) for command line arguments and [this really cool blogpost](https://abhinavsarkar.net/posts/repling-with-haskeline/) for some haskeline magic.
The repl is decent, but implementing something like [this](https://abhinavsarkar.net/posts/repling-with-haskeline/) would be awesome. I&rsquo;d also at least like to add a new command `":l <filename>"` to load the definitions from a file.
### TODO Improve error messages
The error messages currently aren&rsquo;t terrible, but it would be nice to have, e.g. line numbers. `megaparsec` allows for semantic errors, so somehow hook into that like what I&rsquo;m doing for unbound variables?
Error messages are decent, but a little buggy. Syntax error messages are pretty ok, but could have better labeling. The type check error messages are decent, but could do with better location information. Right now, the location defaults to the end of the current definition, which is often good enough, but more detail can&rsquo;t hurt.
### DONE Improve β-equivalence check
### TODO Better testing
The check for β-equivalence could certainly be a lot smarter. Right now it just fully normalizes both terms and checks if the normal forms are equal, which isn&rsquo;t the best strategy. This is much less of an issue without [inductive definitions](#orgc763d54)/recursion, as far less computation gets done in general, let alone at the type level. That being said, it&rsquo;s certainly not a bad idea.
Improved the β-equivalence check to first reduce both terms to WHNF then walk the syntax tree and recursively compare the terms. I figured this would be a necessary precondition to adding definitions, since fully normalizing everything with definitions (recursively unfolding all definitions) seems absolutely horrendous. Also implementing this wasn&rsquo;t nearly as hard as I would have thought.
I would like to avoid regressions as I keep working on this, and a suite of unit tests would make me feel much more comfortable. I made unit tests, then added a ton of stuff. Most of the unit tests are kind of pointless now. For now, I think running the code on the example file is pretty sufficient.
### DONE Better testing
### TODO Alternate syntax
Using some kind of testing framework, like [QuickCheck](https://hackage.haskell.org/package/QuickCheck) and/or [HUnit](https://hackage.haskell.org/package/HUnit) seems like a good idea. I would like to avoid regressions as I keep working on this, and a suite of unit tests would make me feel much more comfortable.
I&rsquo;ve had a bunch of ideas for a more mathematician-friendly syntax bouncing around my head for a while. Implementing one of them would be awesome, but probably quite tricky.
Something like
Theorem basic (S : *) (P : S → *) :
(∀ (x : S), P x → Q x) → (∀ (x : S), P x) → ∀ (x : S), Q x.
Proof
1. Suppose ∀ (x : S), P x → Q x
2. Suppose ∀ (x : S), P x
3. Let x : S
4. P x by [2 x]
5. Q x by [1 x 4]
Qed
I think could be reliably translated into
basic (S : *) (P : S → *) : (Π (x : S), P x → Q x) → (Π (x : S), P x) → Π (x : S), Q x :=
fun (a1 : Π (x : S), P x → Q x) ⇒
fun (a2 : Π (x : S), P x) ⇒
fun (x : S) ⇒
a1 x (a2 x);
and is more intuitively understandable to a mathematician not familiar with type theory, while the latter would be utter nonsense.
I&rsquo;m imagining the parser could be chosen based on the file extension or something. Some way to mix the syntaxes could be nice too.
### DONE Switch to `Text`
### TODO treesitter parser and/or emacs mode
Currently I use `String` everywhere. Switching to `Text` should be straightforward, and would improve performance and Unicode support.
Really not necessary, especially while the syntax is in a bit of flux, but would eventually be nice. The syntax is simple enough that a treesitter grammar shouldn&rsquo;t be too hard to write. An emacs mode would especially be nice if I ever get end up implementing an [alternate syntax](#orgf503794), to better handle indentation, automatically adjust line numbers, etc.
### TODO TUI
This is definitely a stretch goal, but I&rsquo;m imagining a TUI split into two panels. On the left you can see the term you are building with holes in it. On the right you have the focused hole&rsquo;s type as well as the types of everything in scope (like Coq and Lean show while you&rsquo;re in the middle of a proof). Then you can interact with the system by entering commands (e.g. `intros`, `apply`, etc.) which changes the proof term on the left. You&rsquo;d also just be able to type in the left window as well, and edit the proof term directly. This way you&rsquo;d get the benefits of working with tactics, making it way faster to construct proof terms, and the benefits of working with proof terms directly, namely transparency and simplicity. I&rsquo;ll probably want to look into [brick](https://hackage.haskell.org/package/brick) in order to make this happen.
This is definitely a stretch goal, and I&rsquo;m not sure how good of an idea it would be, but I&rsquo;m imagining a TUI split into two panels. On the left you can see the term you are building with holes in it. On the right you have the focused hole&rsquo;s type as well as the types of everything in scope (like Coq and Lean show while you&rsquo;re in the middle of a proof). Then you can interact with the system by entering commands (e.g. `intros`, `apply`, etc.) which changes the proof term on the left. You&rsquo;d also just be able to type in the left window as well, and edit the proof term directly. This way you&rsquo;d get the benefits of working with tactics, making it way faster to construct proof terms, and the benefits of working with proof terms directly, namely transparency and simplicity. I&rsquo;ll probably want to look into [brick](https://hackage.haskell.org/package/brick) if I want to make this happen.

154
README.org
View file

@ -1,97 +1,135 @@
#+title: Dependent Lambda
#+title: Perga
#+options: toc:nil
A basic implementation of a dependently typed lambda calculus (calculus of constructions) based on the exposition in Nederpelt and Geuvers /Type Theory and Formal Proof/. Right now it is /technically/ a perfectly capable higher order logic proof checker, though there is lots of room for improved ergonomics and usability, which I intend to work on.
=perga= is a basic proof assistant based on a dependently typed lambda calculus (calculus of constructions). This implementation is based on the exposition in Nederpelt and Geuvers' /Type Theory and Formal Proof/. Right now it is a perfectly capable higher order logic proof checker, though there is lots of room for improved ergonomics and usability, which I intend to work on. At the moment, =perga= is comparable to Automath in terms of power and ease of use, being slightly more powerful than Automath (though lacks a /primitive notion/ system for the moment), and a touch less ergonomic.
* Syntax
The syntax is fairly flexible and should work as you expect. Identifiers can be Unicode as long as =megaparsec= calls them alphanumeric. Lambda and Pi abstractions can be written in many obvious ways that should be clear from the examples below. Additionally, arrows can be used as an abbreviation for a Π type where the parameter doesn't appear in the body as usual.
The syntax is fairly flexible and should work as you expect. Identifiers can be Unicode as long as =megaparsec= calls them alphanumeric. =λ= and =\Pi= abstractions can be written in many obvious ways that should be clear from the examples below. Additionally, arrows can be used as an abbreviation for a Π type where the parameter doesn't appear in the body as usual.
All of the following examples correctly parse, and should look familiar if you are used to standard lambda calculus notation or Coq syntax.
All of the following example terms correctly parse, and should look familiar if you are used to standard lambda calculus notation or Coq syntax.
#+begin_src
λ (α : *) . λ (β : *) . λ (x : α) . λ (y : β) . x
fun (A B C : *) (g : → C) (f : A → B) (x : A) ⇒ g (f x)
fun (S : *) (P Q : S -> *) (H : Π (x : S) . P x -> Q x) (HP : forall (x : S), P x) => fun (x : S) => H x (HP x)
#+end_src
I mostly stick to Coq syntax throughout this file and the examples, as that is what I'm most used to and is easiest to type. I will probably make the syntax more strict in the future, as this level of flexibility is really not necessary.
* Basic Usage
For the moment, running the program drops you into a repl where you can enter terms in the calculus of construction, which the system will then type check and report the type of the entered term, or any errors if present.
Definitions work similarly, having abstract syntax as shown below.
#+begin_src
<ident> (<ident> : <type>)* : <type>? := <term>;
#+end_src
(The distinction between =<type>= and =<term>= is for emphasis; they are the exact same syntactic category.) Here's a couple definitions of the =const= function from above showing the options with the syntax.
#+begin_src
const := λ (α : *) . λ (β : *) . λ (x : α) . λ (y : β) . x;
const : forall (α β : *), α → β → α := fun (α β : *) (x : α) (y : β) ⇒ x;
const (α β : *) (x : α) (y : β) : α := x;
#+end_src
Type ascriptions are optional. If included, =perga= will check to make sure your definition matches the ascription, and, if so, will remember the way your wrote the type when printing inferred types, which is particularly handy when using abbreviations for complex types. =perga= has no problem inferring the types of top-level definitions, as they are completely determined by the term, but I recommend including ascriptions most of the time, as they serve as a nice piece of documentation, help guide the implementation process, and make sure you are implementing the type you think you are.
** Sample Session
Here's a sample session. Suppose your goal is to prove that for every set $S$ and pair of propositions $P$ and $Q$ on $S$, if $\forall x \in S, P x$ holds, and $\forall x \in S, P x \Rightarrow Q x$, then $\forall x \in S, Q x$. A set $S$ corresponds to a type =S=, so our term will begin
#+begin_src
> fun (S : *)
#+end_src
Likewise propositions are functions from =S → *=, so we can continue
#+begin_src
> fun (S : *) (P Q : S → *)
#+end_src
Since $\forall$ corresponds with =Π=, the hypothesis $\forall x \in S, P x$ would correspond to the type =Π (x : S) . P x= and the hypothesis $\forall x \in S, P x \Rightarrow Q x$ would correspond with =Π (x : S) . P x → Q x=. Thus we can continue
#+begin_src
> fun (S : *) (P Q : S → *) (HP : forall (x : S), P x) (H : forall (x : S), P x → Q x)
#+end_src
Since we are trying to prove a universally quantified proposition, we need to introduce an =x : S=, so
#+begin_src
> fun (S : *) (P Q : S → *) (HP : forall (x : S), P x) (H : forall (x : S), P x → Q x) (x : S)
#+end_src
Finally, all we have left to do is find something of type =Q x=. We can get to =Q x= using =H x= if we can find something of type =P x=. Fortunately, =HP x= is type =P x=, so our final term is
#+begin_src
> fun (S : *) (P Q : S → *) (HP : forall (x : S), P x) (H : forall (x : S), P x → Q x) (x : S) ⇒ H x (HP x)
#+end_src
Pressing enter to send this term to the system, it responds with
#+begin_src
λ (S : *) (P Q : S -> *) (HP : ∏ (x : S) . P x) (H : ∏ (x : S) . P x -> Q x) (x : S) . H x (HP x) : ∏ (S : *) (P Q : S -> *) . (∏ (x : S) . P x) -> (∏ (x : S) . P x -> Q x) -> ∏ (x : S) . Q x
#+end_src
This output is a bit hard to read, but it is ultimately in the form =term : type=. The =term= is, up to minor syntactic differences, the term we entered, and the =type= is the type of the term inferred by the system. As a nice sanity check, we can verify that the =type= indeed corresponds to the theorem we wanted to prove.
Line comments are =--= like in Haskell, and block comments are =(* *)= like ML (and nest properly). There is no significant whitespace, so you are free to format code as you wish.
More complex and interesting proofs and theorems are technically possible (in fact, /all/ interesting theorems and proofs are possible, for a certain definition of /interesting/, /theorem/, and /proof/), though practically infeasible without definitions.
* Usage
Running =perga= without any arguments drops you into a basic repl. From here, you can type in definitions which =perga= will typecheck. Previous definitions are accessible in future definitions. The usual readline keybindings are available, including navigating history, which is saved between sessions (in =~/.cache/perga/history=). In the repl, you can enter ":q", press C-c, or press C-d to quit. Entering ":e" shows everything that has been defined along with their types. Entering ":t <ident>" prints the type of a particular identifier.
* Goals
You can also give =perga= a filename as an argument, in which case it will typecheck every definition in the file. Upon finishing, which should be nearly instantaneous, it will print out all the definitions it parsed along with their types (like you had typed ":e" in the repl) so you can verify that it worked.
* Simple Example
A much larger, commented example is located in [[./examples/example.pg]]. Here is an example file defining Leibniz equality and proving that it is reflexive, symmetric, and transitive.
#+begin_src
-- file: equality.pg
-- Defining Leibniz equality
-- Note that we can leave the ascription off
eq (A : *) (x y : A) := forall (P : A -> *), P x -> P y;
-- Equality is reflexive, which is easy to prove
-- Here we give an ascription so that when `perga` reports the type,
-- it references `eq` rather than inferring the type.
eq_refl (A : *) (x : A) : eq A x x := fun (P : A -> *) (Hx : P x) => Hx;
-- Equality is symmetric. This one's a little harder to prove.
eq_sym (A : *) (x y : A) (Hxy : eq A x y) : eq A y x :=
fun (P : A -> *) (Hy : P y) =>
Hxy (fun (z : A) => P z -> P x) (fun (Hx : P x) => Hx) Hy;
-- Equality is transitive.
eq_trans (A : *) (x y z : A) (Hxy : eq A x y) (Hyz : eq A y z) : eq A x z :=
fun (P : A -> *) (Hx : P x) => Hyz P (Hxy P Hx);
#+end_src
Running =perga equality.pg= yields the following output.
#+begin_src
eq : ∏ (A : *) . A -> A -> *
eq_refl : ∏ (A : *) (x : A) . eq A x x
eq_sym : ∏ (A : *) (x y : A) . eq A x y -> eq A y x
eq_trans : ∏ (A : *) (x y z : A) . eq A x y -> eq A y z -> eq A x z
#+end_src
This means our proofs were accepted. Furthermore, as a sanity check, we can see that the types correspond exactly to what we wanted to prove.
* Future Goals
** Substantive
*** TODO DEFINITIONS
Some kind of definition system would be a difficult and substantial addition, but man is it necessary to do anything. Likewise, I'd probably want a way to define primitive notions/axioms, but that should be an easy extension of the definition system. Further following /Type Theory and Formal Proof/ would additionally yield a nice context system, which would be handy, though I disagree with the choice to differentiate between parameterized definitions and functions. That distinction doesn't really make sense in full calculus of constructions.
Sidenote: With a definition system, I would greatly prefer Haskell-style type annotations to ML-style type annotations, though the latter are likely way easier to implement. It serves as a nice bit of documentation, de-clutters the function definition, and follows the familiar mathematical statement-proof structure.
**** TODO Add support for named free variables to =Expr=
I'm taking the locally nameless approach. Dealing with free variables as de Bruijn indices greater than the context sounds terrible. This requires adding support for \delta reductions, as well as performing reductions and type checking operations in an environment. NOTE: I'll probably want to use the [[https://hackage.haskell.org/package/mtl-2.3.1/docs/Control-Monad-Reader.html][Reader monad]].
*** TODO Let-expressions
I vaguely remember from reading Coq'art that Coq does something special with =let= expressions, so I'll maybe want to check that out. I tried implementing =let= as syntax sugar for an immediately called function, but that proved to be a massive mess with how I'm handling things. =let= expressions would definitely be handy for factoring out complex sub expressions.
*** TODO Inference
Obviously not fully decidable, but I might be able to implement some basic unification algorithm. This isn't super necessary though, I find leaving off the types of arguments to generally be a bad idea, but in some cases it can be worth it.
Obviously not fully decidable, but I might be able to implement some basic unification algorithm. This isn't super necessary though, I find leaving off the types of arguments to generally be a bad idea, but in some cases it can be handy, especially not at the top level.
*** TODO Implicits
Much, much more useful than [[Inference][inference]], implicit arguments would be amazing. It also seems a lot more complicated, but any system for dealing with implicit arguments is far better than none.
Much, much more useful than [[Inference][inference]], implicit arguments would be amazing. It also seems a lot more complicated, but any system for dealing with implicit arguments is far better than none. Getting rid of stuff like [[./examples/example.pg::213][lines 213-215 of the example file]] would be amazing.
*** TODO Module System
A proper module system would be wonderful. To me, ML style modules with structures, signatures, and functors seems like the right way to handle algebraic structures for a relatively simple language, rather than records (or a bunch of =and='s like I currently have) or type classes (probably much harder, but could be nicer), but any way of managing scope, importing files, etc. is a necessity.
*** TODO Universes?
Not really necessary without [[Inductive Definitions][inductive definitions]], but could be fun.
Not really all that necessary, especially without [[Inductive Definitions][inductive definitions]], but could be fun.
*** TODO Inductive Definitions
This is definitely a stretch goal. It would be cool though, and would turn this proof checker into a much more competent programming language. It's not necessary for the math, but inductive definitions let you leverage computation in proofs, which is amazing. They also make certain definitions easier, by avoiding needing to manually stipulate elimination rules, including induction principles.
This is definitely a stretch goal. It would be cool though, and would turn this proof checker into a much more competent programming language. It's not necessary for the math, but inductive definitions let you leverage computation in proofs, which is amazing. They also make certain definitions way easier, by avoiding needing to manually stipulate elimination rules, including induction principles, and let you keep more math constructive and understandable to the computer.
** Cosmetic/usage/technical
*** TODO Prettier pretty printing
Right now, everything defaults to one line, which can be a problem with how large the proof terms get. Probably want to use [[https://hackage.haskell.org/package/prettyprinter][prettyprinter]] to be able to nicely handle indentation and line breaks.
*** DONE Better usage
Read input from a file if a filename is given, otherwise drop into a repl. Perhaps use something like [[https://hackage.haskell.org/package/haskeline][haskeline]] to improve the repl (though =rlwrap= is fine, so not a huge priority).
For even better usage in the future, check out [[https://hackage.haskell.org/package/optparse-applicative][optparse]] for command line arguments and [[https://abhinavsarkar.net/posts/repling-with-haskeline/][this really cool blogpost]] for some haskeline magic.
*** TODO Better repl
The repl is decent, but implementing something like [[https://abhinavsarkar.net/posts/repling-with-haskeline/][this]] would be awesome. I'd also at least like to add a new command =":l <filename>"= to load the definitions from a file.
*** TODO Improve error messages
The error messages currently aren't terrible, but it would be nice to have, e.g. line numbers. =megaparsec= allows for semantic errors, so somehow hook into that like what I'm doing for unbound variables?
Error messages are decent, but a little buggy. Syntax error messages are pretty ok, but could have better labeling. The type check error messages are decent, but could do with better location information. Right now, the location defaults to the end of the current definition, which is often good enough, but more detail can't hurt.
*** DONE Improve β-equivalence check
The check for β-equivalence could certainly be a lot smarter. Right now it just fully normalizes both terms and checks if the normal forms are equal, which isn't the best strategy. This is much less of an issue without [[Inductive Definitions][inductive definitions]]/recursion, as far less computation gets done in general, let alone at the type level. That being said, it's certainly not a bad idea.
*** TODO Better testing
I would like to avoid regressions as I keep working on this, and a suite of unit tests would make me feel much more comfortable. I made unit tests, then added a ton of stuff. Most of the unit tests are kind of pointless now. For now, I think running the code on the example file is pretty sufficient.
Improved the β-equivalence check to first reduce both terms to WHNF then walk the syntax tree and recursively compare the terms. I figured this would be a necessary precondition to adding definitions, since fully normalizing everything with definitions (recursively unfolding all definitions) seems absolutely horrendous. Also implementing this wasn't nearly as hard as I would have thought.
*** TODO Alternate syntax
I've had a bunch of ideas for a more mathematician-friendly syntax bouncing around my head for a while. Implementing one of them would be awesome, but probably quite tricky.
*** DONE Better testing
Using some kind of testing framework, like [[https://hackage.haskell.org/package/QuickCheck][QuickCheck]] and/or [[https://hackage.haskell.org/package/HUnit][HUnit]] seems like a good idea. I would like to avoid regressions as I keep working on this, and a suite of unit tests would make me feel much more comfortable.
Something like
#+begin_src
Theorem basic (S : *) (P : S → *) :
(∀ (x : S), P x → Q x) → (∀ (x : S), P x) → ∀ (x : S), Q x.
Proof
1. Suppose ∀ (x : S), P x → Q x
2. Suppose ∀ (x : S), P x
3. Let x : S
4. P x by [2 x]
5. Q x by [1 x 4]
Qed
#+end_src
I think could be reliably translated into
#+begin_src
basic (S : *) (P : S → *) : (Π (x : S), P x → Q x) → (Π (x : S), P x) → Π (x : S), Q x :=
fun (a1 : Π (x : S), P x → Q x) ⇒
fun (a2 : Π (x : S), P x) ⇒
fun (x : S) ⇒
a1 x (a2 x);
#+end_src
and is more intuitively understandable to a mathematician not familiar with type theory, while the latter would be utter nonsense.
*** DONE Switch to =Text=
Currently I use =String= everywhere. Switching to =Text= should be straightforward, and would improve performance and Unicode support.
I'm imagining the parser could be chosen based on the file extension or something. Some way to mix the syntaxes could be nice too.
*** TODO treesitter parser and/or emacs mode
Really not necessary, especially while the syntax is in a bit of flux, but would eventually be nice. The syntax is simple enough that a treesitter grammar shouldn't be too hard to write. An emacs mode would especially be nice if I ever get end up implementing an [[Alternate syntax][alternate syntax]], to better handle indentation, automatically adjust line numbers, etc.
*** TODO TUI
This is definitely a stretch goal, but I'm imagining a TUI split into two panels. On the left you can see the term you are building with holes in it. On the right you have the focused hole's type as well as the types of everything in scope (like Coq and Lean show while you're in the middle of a proof). Then you can interact with the system by entering commands (e.g. =intros=, =apply=, etc.) which changes the proof term on the left. You'd also just be able to type in the left window as well, and edit the proof term directly. This way you'd get the benefits of working with tactics, making it way faster to construct proof terms, and the benefits of working with proof terms directly, namely transparency and simplicity. I'll probably want to look into [[https://hackage.haskell.org/package/brick][brick]] in order to make this happen.
This is definitely a stretch goal, and I'm not sure how good of an idea it would be, but I'm imagining a TUI split into two panels. On the left you can see the term you are building with holes in it. On the right you have the focused hole's type as well as the types of everything in scope (like Coq and Lean show while you're in the middle of a proof). Then you can interact with the system by entering commands (e.g. =intros=, =apply=, etc.) which changes the proof term on the left. You'd also just be able to type in the left window as well, and edit the proof term directly. This way you'd get the benefits of working with tactics, making it way faster to construct proof terms, and the benefits of working with proof terms directly, namely transparency and simplicity. I'll probably want to look into [[https://hackage.haskell.org/package/brick][brick]] if I want to make this happen.

12
app/Main.hs
View file

@ -1,6 +1,9 @@
module Main where
import Control.Monad (void)
import qualified Data.Map as M
import qualified Data.Text.IO as T
import Eval (Env)
import Parser
import Repl
import System.Environment
@ -10,15 +13,18 @@ main :: IO ()
main = do
args <- getArgs
case args of
[] -> repl
[] -> void repl
[file] -> handleFile file
_ -> putStrLn "usage './perga' for repl and './perga <filename>' to get input from a file"
dumpEnv :: Env -> IO ()
dumpEnv = void . M.traverseWithKey ((putStrLn .) . showEnvEntry)
handleFile :: String -> IO ()
handleFile fileName =
do
fileH <- openFile fileName ReadMode
input <- T.hGetContents fileH
case pAll input of
case parseProgram input of
Left err -> putStrLn err
Right () -> putStrLn "success!"
Right env -> dumpEnv env

66
app/Repl.hs
View file

@ -1,52 +1,62 @@
module Repl (repl) where
module Repl (repl, showEnvEntry) where
import Check
import qualified Data.Text as T
import Data.Functor (void)
import Data.List (isPrefixOf)
import qualified Data.Map as M
import qualified Data.Text as T
import Eval
import Expr
import Parser
import System.Console.Haskeline
import System.Directory (createDirectoryIfMissing, getHomeDirectory)
import System.FilePath ((</>))
newtype ReplState = ReplState
{ debugMode :: Bool
}
data ReplCommand = Quit | ToggleDebug | Input String deriving (Show)
data ReplCommand = Quit | DumpEnv | TypeQuery String | Input String deriving (Show)
parseCommand :: Maybe String -> Maybe ReplCommand
parseCommand Nothing = Nothing
parseCommand (Just ":q") = Just Quit
parseCommand (Just ":d") = Just ToggleDebug
parseCommand (Just input) = Just (Input input)
parseCommand (Just ":e") = Just DumpEnv
parseCommand (Just input)
| ":t" `isPrefixOf` input = case words input of
[":t", rest] -> Just $ TypeQuery rest
_ -> Nothing
| otherwise = Just $ Input input
handleInput :: ReplState -> String -> InputT IO ()
handleInput state input = case pAll (T.pack input) of
Left err -> outputStrLn err
Right () -> pure ()
showEnvEntry :: T.Text -> Expr -> String
showEnvEntry k v = T.unpack k ++ " : " ++ prettyS v
printDebugInfo :: Expr -> Expr -> InputT IO ()
printDebugInfo expr ty = do
outputStrLn $ "expr\t" ++ show expr
outputStrLn $ "type\t" ++ show ty
outputStrLn $ "type\t" ++ prettyS ty
dumpEnv :: Env -> InputT IO ()
dumpEnv = void . M.traverseWithKey ((outputStrLn .) . showEnvEntry)
repl :: IO ()
handleInput :: GlobalState -> String -> InputT IO GlobalState
handleInput env input =
let (res, env') = parseDefEmpty env (T.pack input)
in case res of
Left err -> outputStrLn err >> pure env'
Right () -> pure env'
repl :: IO GlobalState
repl = do
home <- getHomeDirectory
let basepath = home </> ".cache" </> "perga"
let filepath = basepath </> "history"
createDirectoryIfMissing True basepath
runInputT (defaultSettings{historyFile = Just filepath}) $ loop (ReplState False)
runInputT (defaultSettings{historyFile = Just filepath}) (loop GS{_defs = M.empty, _types = M.empty})
where
loop :: ReplState -> InputT IO ()
loop state = do
loop :: GlobalState -> InputT IO GlobalState
loop env = do
minput <- getInputLine "> "
case parseCommand minput of
Nothing -> return ()
Just Quit -> return ()
Just ToggleDebug -> loop state{debugMode = not (debugMode state)}
Nothing -> pure env
Just Quit -> pure env
Just DumpEnv -> dumpEnv (_types env) >> loop env
Just (TypeQuery input) ->
( case M.lookup (T.pack input) (_types env) of
Nothing -> outputStrLn (input ++ " unbound")
Just expr -> outputStrLn $ prettyS expr
)
>> loop env
Just (Input input) -> do
handleInput state input
loop state
env' <- handleInput env input
loop env'

250
examples/example.pg Normal file
View file

@ -0,0 +1,250 @@
-- False
false : * := forall (A : *), A;
-- no introduction rule
-- elimination rule
false_elim (A : *) (contra : false) : A := contra A;
-- --------------------------------------------------------------------------------------------------------------
-- Negation
not (A : *) : * := A -> false;
-- introduction rule (kinda just the definition)
not_intro (A : *) (h : A -> false) : not A := h;
-- elimination rule
not_elim (A B : *) (a : A) (na : not A) : B := na a B;
-- can introduce double negation (can't eliminate it as that isn't constructive)
double_neg_intro (A : *) (a : A) : not (not A) :=
fun (notA : not A) => notA a;
-- --------------------------------------------------------------------------------------------------------------
-- Conjunction
and (A B : *) : * := forall (C : *), (A -> B -> C) -> C;
-- introduction rule
and_intro (A B : *) (a : A) (b : B) : and A B :=
fun (C : *) (H : A -> B -> C) => H a b;
-- left elimination rule
and_elim_l (A B : *) (ab : and A B) : A :=
ab A (fun (a : A) (b : B) => a);
-- right elimination rule
and_elim_r (A B : *) (ab : and A B) : B :=
ab B (fun (a : A) (b : B) => b);
-- --------------------------------------------------------------------------------------------------------------
-- Disjunction
-- 2nd order disjunction
or (A B : *) : * := forall (C : *), (A -> C) -> (B -> C) -> C;
-- left introduction rule
or_intro_l (A B : *) (a : A) : or A B :=
fun (C : *) (ha : A -> C) (hb : B -> C) => ha a;
-- right introduction rule
or_intro_r (A B : *) (b : B) : or A B :=
fun (C : *) (ha : A -> C) (hb : B -> C) => hb b;
-- elimination rule (kinda just the definition)
or_elim (A B C : *) (ab : or A B) (ha : A -> C) (hb : B -> C) : C :=
ab C ha hb;
-- --------------------------------------------------------------------------------------------------------------
-- Existential
-- 2nd order existential
exists (A : *) (P : A -> *) : * := forall (C : *), (forall (x : A), P x -> C) -> C;
-- introduction rule
exists_intro (A : *) (P : A -> *) (a : A) (h : P a) : exists A P :=
fun (C : *) (g : forall (x : A), P x -> C) => g a h;
-- elimination rule (kinda just the definition)
exists_elim (A B : *) (P : A -> *) (ex_a : exists A P) (h : forall (a : A), P a -> B) : B :=
ex_a B h;
-- --------------------------------------------------------------------------------------------------------------
-- Universal
-- 2nd order universal (just ∏, including it for completeness)
all (A : *) (P : A -> *) : * := forall (a : A), P a;
-- introduction rule
all_intro (A : *) (P : A -> *) (h : forall (a : A), P a) : all A P := h;
-- elimination rule
all_elim (A : *) (P : A -> *) (h_all : all A P) (a : A) : P a := h_all a;
-- --------------------------------------------------------------------------------------------------------------
-- Equality
-- 2nd order Leibniz equality
eq (A : *) (x y : A) := forall (P : A -> *), P x -> P y;
-- equality is reflexive
eq_refl (A : *) (x : A) : eq A x x := fun (P : A -> *) (Hx : P x) => Hx;
-- equality is symmetric
eq_sym (A : *) (x y : A) (Hxy : eq A x y) : eq A y x := fun (P : A -> *) (Hy : P y) =>
Hxy (fun (z : A) => P z -> P x) (fun (Hx : P x) => Hx) Hy;
-- equality is transitive
eq_trans (A : *) (x y z : A) (Hxy : eq A x y) (Hyz : eq A y z) : eq A x z := fun (P : A -> *) (Hx : P x) =>
Hyz P (Hxy P Hx);
-- --------------------------------------------------------------------------------------------------------------
-- Some logic theorems
-- ~(A ∨ B) => ~A ∧ ~B
de_morgan1 (A B : *) (h : not (or A B)) : and (not A) (not B) :=
and_intro (not A) (not B)
(fun (a : A) => h (or_intro_l A B a))
(fun (b : B) => h (or_intro_r A B b));
-- ~A ∧ ~B => ~(A ∨ B)
de_morgan2 (A B : *) (h : and (not A) (not B)) : not (or A B) :=
fun (contra : or A B) =>
or_elim A B false contra
(and_elim_l (not A) (not B) h)
(and_elim_r (not A) (not B) h);
-- ~A ∨ ~B => ~(A ∧ B)
de_morgan3 (A B : *) (h : or (not A) (not B)) : not (and A B) :=
fun (contra : and A B) =>
or_elim (not A) (not B) false h
(fun (na : not A) => na (and_elim_l A B contra))
(fun (nb : not B) => nb (and_elim_r A B contra));
-- the last one (~(A ∧ B) => ~A ∨ ~B) is not possible constructively
-- A ∧ B => B ∧ A
and_comm (A B : *) (h : and A B) : and B A :=
and_intro B A
(and_elim_r A B h)
(and_elim_l A B h);
-- A ∨ B => B ∨ A
or_comm (A B : *) (h : or A B) : or B A :=
or_elim A B (or B A) h
(fun (a : A) => or_intro_r B A a)
(fun (b : B) => or_intro_l B A b);
-- A ∧ (B ∧ C) => (A ∧ B) ∧ C
and_assoc_l (A B C : *) (h : and A (and B C)) : and (and A B) C :=
and_intro (and A B) C
(and_intro A B
(and_elim_l A (and B C) h)
(and_elim_l B C (and_elim_r A (and B C) h)))
(and_elim_r B C (and_elim_r A (and B C) h));
-- (A ∧ B) ∧ C => A ∧ (B ∧ C)
and_assoc_r (A B C : *) (h : and (and A B) C) : and A (and B C) :=
and_intro A (and B C)
(and_elim_l A B (and_elim_l (and A B) C h))
(and_intro B C
(and_elim_r A B (and_elim_l (and A B) C h))
(and_elim_r (and A B) C h));
-- A ∨ (B ∨ C) => (A ∨ B) ∨ C
or_assoc_l (A B C : *) (h : or A (or B C)) : or (or A B) C :=
or_elim A (or B C) (or (or A B) C) h
(fun (a : A) => or_intro_l (or A B) C (or_intro_l A B a))
(fun (g : or B C) =>
or_elim B C (or (or A B) C) g
(fun (b : B) => or_intro_l (or A B) C (or_intro_r A B b))
(fun (c : C) => or_intro_r (or A B) C c));
-- (A ∨ B) ∨ C => A ∨ (B ∨ C)
or_assoc_r (A B C : *) (h : or (or A B) C) : or A (or B C) :=
or_elim (or A B) C (or A (or B C)) h
(fun (g : or A B) =>
or_elim A B (or A (or B C)) g
(fun (a : A) => or_intro_l A (or B C) a)
(fun (b : B) => or_intro_r A (or B C) (or_intro_l B C b)))
(fun (c : C) => or_intro_r A (or B C) (or_intro_r B C c));
-- A ∧ (B ∨ C) => A ∧ B ∨ A ∧ C
and_distrib_l_or (A B C : *) (h : and A (or B C)) : or (and A B) (and A C) :=
or_elim B C (or (and A B) (and A C)) (and_elim_r A (or B C) h)
(fun (b : B) => or_intro_l (and A B) (and A C)
(and_intro A B (and_elim_l A (or B C) h) b))
(fun (c : C) => or_intro_r (and A B) (and A C)
(and_intro A C (and_elim_l A (or B C) h) c));
-- --------------------------------------------------------------------------------------------------------------
-- very little very basic algebra
-- what it means for a set and operation to be a semigroup
semigroup (M : *) (op : M -> M -> M) : * :=
forall (a b c : M), eq M (op (op a b) c) (op a (op b c));
-- what it means for a set, operation, and identity element to be a monoid
monoid (M : *) (op : M -> M -> M) (e : M) : * :=
and (semigroup M op)
(and (forall (a : M), eq M (op a e) a)
(forall (a : M), eq M (op e a) a));
-- identity elements in monoids are unique
monoid_identity_unique
(M : *) (op : M -> M -> M) (e : M) (Hmonoid : monoid M op e)
(a : M) (H : forall (b : M), eq M (op a b) b) : eq M a e :=
eq_trans M a (op a e) e
-- a = a * e
(eq_sym M (op a e) a
(and_elim_l
(forall (a : M), eq M (op a e) a)
(forall (a : M), eq M (op e a) a)
(and_elim_r
(semigroup M op)
(and (forall (a : M), eq M (op a e) a) (forall (a : M), eq M (op e a) a))
Hmonoid) a))
-- a * e = e
(H e);
-- what it means for a set, operation, identity element, and inverse map to be a group
group (G : *) (op : G -> G -> G) (e : G) (inv : G -> G) : * :=
and (monoid G op e)
(and (forall (a : G), eq G (op a (inv a)) e)
(forall (a : G), eq G (op (inv a) a) e));
-- --------------------------------------------------------------------------------------------------------------
-- Church numerals
nat : * := forall (A : *), (A -> A) -> A -> A;
-- zero is the constant function
zero : nat := fun (A : *) (f : A -> A) (x : A) => x;
-- `suc n` composes one more `f` than `n`
suc : nat -> nat := fun (n : nat) (A : *) (f : A -> A) (x : A) => f ((n A f) x);
-- addition as usual
add : nat -> nat -> nat := fun (n m : nat) (A : *) (f : A -> A) (x : A) => (m A f) (n A f x);
-- multiplication as usual
mul : nat -> nat -> nat := fun (n m : nat) (A : *) (f : A -> A) (x : A) => (m A (n A f)) x;
-- unforunately, I believe it is impossible to prove induction on Church numerals,
-- which makes it really hard to prove standard theorems, or do anything really.
-- with a primitive notion system, we could stipulate the existence of a type of
-- natural numbers satisfying the Peano axioms, but I haven't implemented that yet.

62
lib/Check.hs
View file

@ -1,61 +1,51 @@
module Check (TypeCheckError (..), CheckResult, checkType) where
{-# LANGUAGE LambdaCase #-}
module Check (checkType) where
import Control.Monad (unless)
import Control.Monad.Except (MonadError (throwError))
import Control.Monad.Reader
import Data.List ((!?))
import qualified Data.Map as M
import Data.Text (Text)
import qualified Data.Text as T
import Eval
import Expr
import Errors
import Eval (Env, betaEquiv, envLookup, isSort, subst, whnf)
import Expr (Expr (..), incIndices, occursFree)
type Context = [Expr]
data TypeCheckError = SquareUntyped | UnboundVariable Text | NotASort Expr Expr | ExpectedPiType Expr Expr | NotEquivalent Expr Expr Expr deriving (Eq, Ord)
matchPi :: Expr -> Expr -> ReaderT Env Result (Expr, Expr)
matchPi x mt =
whnf mt >>= \case
(Pi _ a b) -> pure (a, b)
t -> throwError $ ExpectedPiType x t
instance Show TypeCheckError where
show SquareUntyped = "□ does not have a type"
show (UnboundVariable x) = "Unbound variable: " ++ T.unpack x
show (NotASort x t) = "Expected " ++ prettyS x ++ " to have type * or □, instead found " ++ prettyS t
show (ExpectedPiType m a) = prettyS m ++ " : " ++ prettyS a ++ " is not a function"
show (NotEquivalent a a' e) = "Cannot unify " ++ prettyS a ++ " with " ++ prettyS a' ++ " when evaluating " ++ prettyS e
type CheckResult = Either TypeCheckError
matchPi :: Expr -> Expr -> ReaderT Env CheckResult (Expr, Expr)
matchPi _ (Pi _ a b) = pure (a, b)
matchPi m e = throwError $ ExpectedPiType m e
findType :: Context -> Expr -> ReaderT Env CheckResult Expr
findType :: Context -> Expr -> ReaderT Env Result Expr
findType _ Star = pure Square
findType _ Square = throwError SquareUntyped
findType g (Var n x) = do
t <- maybe (throwError $ UnboundVariable x) pure $ g !? fromInteger n
s <- findType g t
unless (isSort s) $ throwError $ NotASort t s
(sSort, s) <- findType g t >>= isSort
unless sSort $ throwError $ NotASort t s
pure t
findType g (Free n) = asks (M.lookup n) >>= maybe (throwError $ UnboundVariable n) (findType g)
findType g (Free n) = envLookup n >>= findType g
findType g e@(App m n) = do
(a, b) <- findType g m >>= matchPi m
a' <- findType g n
equiv <- asks $ runReader (betaEquiv a a')
equiv <- betaEquiv a a'
unless equiv $ throwError $ NotEquivalent a a' e
pure $ subst 0 n b
findType g (Abs x a m) = do
s1 <- findType g a
unless (isSort s1) $ throwError $ NotASort a s1
(s1Sort, s1) <- findType g a >>= isSort
unless s1Sort $ throwError $ NotASort a s1
b <- findType (incIndices a : map incIndices g) m
s2 <- findType g (Pi x a b)
unless (isSort s2) $ throwError $ NotASort (Pi x a b) s2
(s2Sort, s2) <- findType g (Pi x a b) >>= isSort
unless s2Sort $ throwError $ NotASort (Pi x a b) s2
pure $ if occursFree 0 b then Pi x a b else Pi "" a b
findType g (Pi _ a b) = do
s1 <- findType g a
unless (isSort s1) $ throwError $ NotASort a s1
s2 <- findType (incIndices a : map incIndices g) b
unless (isSort s2) $ throwError $ NotASort b s2
(s1Sort, s1) <- findType g a >>= isSort
unless s1Sort $ throwError $ NotASort a s1
(s2Sort, s2) <- findType (incIndices a : map incIndices g) b >>= isSort
unless s2Sort $ throwError $ NotASort b s2
pure s2
checkType :: Env -> Context -> Expr -> CheckResult Expr
checkType env g t = runReaderT (findType g t) env
checkType :: Env -> Expr -> Result Expr
checkType env t = runReaderT (findType [] t) env

22
lib/Errors.hs Normal file
View file

@ -0,0 +1,22 @@
module Errors where
import Data.Text (Text)
import qualified Data.Text as T
import Expr
data Error
= SquareUntyped
| UnboundVariable Text
| NotASort Expr Expr
| ExpectedPiType Expr Expr
| NotEquivalent Expr Expr Expr
deriving (Eq, Ord)
instance Show Error where
show SquareUntyped = "□ does not have a type"
show (UnboundVariable x) = "Unbound variable: " ++ T.unpack x
show (NotASort x t) = "Expected " ++ prettyS x ++ " to have type * or □, instead found " ++ prettyS t
show (ExpectedPiType x t) = prettyS x ++ " : " ++ prettyS t ++ " is not a function"
show (NotEquivalent a a' e) = "Cannot unify " ++ prettyS a ++ " with " ++ prettyS a' ++ " when evaluating " ++ prettyS e
type Result = Either Error

55
lib/Eval.hs
View file

@ -1,9 +1,12 @@
{-# LANGUAGE TupleSections #-}
module Eval where
import Control.Monad.Except (MonadError (..))
import Control.Monad.Reader
import qualified Data.Map as M
import Data.Maybe
import Data.Text (Text)
import Errors
import Expr
type Env = M.Map Text Expr
@ -21,19 +24,43 @@ subst k s (App m n) = App (subst k s m) (subst k s n)
subst k s (Abs x m n) = Abs x (subst k s m) (subst (k + 1) (incIndices s) n)
subst k s (Pi x m n) = Pi x (subst k s m) (subst (k + 1) (incIndices s) n)
whnf :: Expr -> Expr
whnf (App (Abs _ _ v) n) = whnf $ subst 0 n v
whnf e = e
envLookup :: Text -> ReaderT Env Result Expr
envLookup n = asks (M.lookup n) >>= maybe (throwError $ UnboundVariable n) pure
betaEquiv :: Expr -> Expr -> Reader Env Bool
-- reduce until β reducts are impossible in head position
whnf :: Expr -> ReaderT Env Result Expr
whnf (App (Abs _ _ v) n) = whnf $ subst 0 n v
whnf (App m n) = do
m' <- whnf m
if m == m'
then pure $ App m n
else whnf $ App m' n
whnf (Free n) = envLookup n
whnf e = pure e
betaEquiv :: Expr -> Expr -> ReaderT Env Result Bool
betaEquiv e1 e2
| e1 == e2 = pure True
| otherwise = case (whnf e1, whnf e2) of
(Var k1 _, Var k2 _) -> pure $ k1 == k2
(Free n, Free m) -> pure $ n == m
(Free n, e) -> fromMaybe False <$> (asks (M.lookup n) >>= traverse (`betaEquiv` e))
(e, Free n) -> fromMaybe False <$> (asks (M.lookup n) >>= traverse (`betaEquiv` e))
(Star, Star) -> pure True
(Abs _ t1 v1, Abs _ t2 v2) -> (&&) <$> betaEquiv t1 t2 <*> betaEquiv v1 v2 -- i want idiom brackets
(Pi _ t1 v1, Pi _ t2 v2) -> (&&) <$> betaEquiv t1 t2 <*> betaEquiv v1 v2
_ -> pure False -- remaining cases impossible or false
| otherwise = do
e1' <- whnf e1
e2' <- whnf e2
case (e1', e2') of
(Var k1 _, Var k2 _) -> pure $ k1 == k2
(Free n, Free m) -> pure $ n == m
(Free n, e) -> envLookup n >>= betaEquiv e
(e, Free n) -> envLookup n >>= betaEquiv e
(Star, Star) -> pure True
(Abs _ t1 v1, Abs _ t2 v2) -> (&&) <$> betaEquiv t1 t2 <*> betaEquiv v1 v2 -- i want idiom brackets
(Pi _ t1 v1, Pi _ t2 v2) -> (&&) <$> betaEquiv t1 t2 <*> betaEquiv v1 v2
_ -> pure False -- remaining cases impossible or false
checkBeta :: Env -> Expr -> Expr -> Result Bool
checkBeta env e1 e2 = runReaderT (betaEquiv e1 e2) env
isSortPure :: Expr -> Bool
isSortPure Star = True
isSortPure Square = True
isSortPure _ = False
isSort :: Expr -> ReaderT Env Result (Bool, Expr)
isSort s = (,s) . isSortPure <$> whnf s

15
lib/Expr.hs
View file

@ -32,11 +32,6 @@ occursFree n (App a b) = on (||) (occursFree n) a b
occursFree n (Abs _ a b) = occursFree n a || occursFree (n + 1) b
occursFree n (Pi _ a b) = occursFree n a || occursFree (n + 1) b
isSort :: Expr -> Bool
isSort Star = True
isSort Square = True
isSort _ = False
shiftIndices :: Integer -> Integer -> Expr -> Expr
shiftIndices d c (Var k x)
| k >= c = Var (k + d) x
@ -62,6 +57,14 @@ collectLambdas (Abs x ty body) = ((x, ty) : params, final)
(params, final) = collectLambdas body
collectLambdas e = ([], e)
cleanNames :: Expr -> Expr
cleanNames (App m n) = App (cleanNames m) (cleanNames n)
cleanNames (Abs x ty body) = Abs x (cleanNames ty) (cleanNames body)
cleanNames (Pi x ty body)
| occursFree 0 body = Pi x (cleanNames ty) (cleanNames body)
| otherwise = Pi "" ty (cleanNames body)
cleanNames e = e
collectPis :: Expr -> ([(Text, Expr)], Expr)
collectPis p@(Pi "" _ _) = ([], p)
collectPis (Pi x ty body) = ((x, ty) : params, final)
@ -104,7 +107,7 @@ helper k e@(Abs{}) = if k >= 1 then parenthesize res else res
res = "λ " <> T.unwords grouped <> " . " <> pretty body
pretty :: Expr -> Text
pretty = helper 0
pretty = helper 0 . cleanNames
prettyS :: Expr -> String
prettyS = T.unpack . pretty

117
lib/Parser.hs
View file

@ -1,39 +1,52 @@
{-# LANGUAGE NamedFieldPuns #-}
module Parser (pAll) where
module Parser (parseProgram, parseDef, parseDefEmpty, GlobalState (..)) where
import Check
import Control.Monad
import Control.Monad.State.Strict
import Control.Monad.State.Strict (
MonadState (get, put),
State,
evalState,
modify,
runState,
)
import Data.Bifunctor (first)
import Data.Functor.Identity
import Data.List (elemIndex)
import Data.List.NonEmpty (NonEmpty ((:|)))
import qualified Data.List.NonEmpty as NE
import qualified Data.Map as M
import Data.Text (Text)
import qualified Data.Text as T
import Errors (Error (..))
import Eval
import Expr (Expr (..), incIndices)
import Text.Megaparsec hiding (State)
import Text.Megaparsec.Char
import qualified Text.Megaparsec.Char.Lexer as L
data InnerState = IS {_binds :: [Text], _defs :: Env}
data TypeDef = TD {_ident :: Text, _type :: Expr}
data DefinitionLine = DL {_td :: TypeDef, _body :: Expr}
newtype TypeError = TE TypeCheckError
data GlobalState = GS {_defs :: Env, _types :: Env}
data InnerState = IS {_binds :: [TypeDef], _gs :: GlobalState}
newtype TypeError = TE Error
deriving (Eq, Ord, Show)
type Parser = ParsecT TypeError Text (State InnerState)
instance ShowErrorComponent TypeError where
showErrorComponent (TE e) = show e
showErrorComponent (TE e) = show e
bindsToIS :: ([Text] -> [Text]) -> InnerState -> InnerState
bindsToIS :: ([TypeDef] -> [TypeDef]) -> InnerState -> InnerState
bindsToIS f x@(IS{_binds}) = x{_binds = f _binds}
defsToIS :: (Env -> Env) -> InnerState -> InnerState
defsToIS f x@(IS{_defs}) = x{_defs = f _defs}
defsToIS f x@(IS{_gs = y@GS{_defs}}) = x{_gs = y{_defs = f _defs}}
typesToIS :: (Env -> Env) -> InnerState -> InnerState
typesToIS f x@(IS{_gs = y@GS{_types}}) = x{_gs = y{_types = f _types}}
skipSpace :: Parser ()
skipSpace =
@ -54,7 +67,7 @@ pIdentifier = label "identifier" $ lexeme $ do
pVar :: Parser Expr
pVar = label "variable" $ lexeme $ do
var <- pIdentifier
binders <- _binds <$> get
binders <- map _ident . _binds <$> get
pure $ case elemIndex var binders of
Just i -> Var (fromIntegral i) var
Nothing -> Free var
@ -65,18 +78,21 @@ defChoice options = lexeme $ label (T.unpack $ NE.head options) $ void $ choice
pParamGroup :: Parser [(Text, Expr)]
pParamGroup = lexeme $ label "parameter group" $ between (char '(') (char ')') $ do
idents <- some pIdentifier
_ <- defChoice $ ":" :| []
_ <- defChoice $ pure ":"
ty <- pExpr
modify $ bindsToIS $ flip (foldl $ flip (:)) idents
modify $ bindsToIS $ flip (foldl $ flip (:)) (map (\idt -> TD{_ident = idt, _type = ty}) idents)
pure $ zip idents (iterate incIndices ty)
pParams :: Parser [(Text, Expr)]
pParams = concat <$> some pParamGroup
pSomeParams :: Parser [(Text, Expr)]
pSomeParams = lexeme $ concat <$> some pParamGroup
pManyParams :: Parser [(Text, Expr)]
pManyParams = lexeme $ concat <$> many pParamGroup
pLAbs :: Parser Expr
pLAbs = lexeme $ label "λ-abstraction" $ do
_ <- defChoice $ "λ" :| ["lambda", "fun"]
params <- pParams
params <- pSomeParams
_ <- defChoice $ "." :| ["=>", "⇒"]
body <- pExpr
modify $ bindsToIS $ drop $ length params
@ -85,7 +101,7 @@ pLAbs = lexeme $ label "λ-abstraction" $ do
pPAbs :: Parser Expr
pPAbs = lexeme $ label "Π-abstraction" $ do
_ <- defChoice $ "∏" :| ["Pi", "forall", "∀"]
params <- pParams
params <- pSomeParams
_ <- defChoice $ "." :| [","]
body <- pExpr
modify $ bindsToIS $ drop $ length params
@ -98,13 +114,45 @@ pArrow = lexeme $ label "->" $ do
Pi "" a . incIndices <$> pExpr
pApp :: Parser Expr
pApp = foldl1 App <$> some pTerm
pApp = lexeme $ foldl1 App <$> some pTerm
pStar :: Parser Expr
pStar = Star <$ defChoice ("*" :| [])
pStar = lexeme $ Star <$ defChoice (pure "*")
pSquare :: Parser Expr
pSquare = Square <$ defChoice ("□" :| ["[]"])
pSquare = lexeme $ Square <$ defChoice ("□" :| ["[]"])
checkAscription :: Text -> Expr -> Maybe Expr -> Parser DefinitionLine
checkAscription ident value massert = do
IS{_gs = GS{_defs, _types}} <- get
case (checkType _defs value, massert) of
(Left err, _) -> customFailure $ TE err
(Right ty, Nothing) -> pure DL{_td = TD{_ident = ident, _type = ty}, _body = value}
(Right ty, Just assert) -> case checkBeta _defs ty assert of
Left err -> customFailure $ TE err
Right equiv -> do
unless equiv $ customFailure $ TE $ NotEquivalent ty assert value
pure DL{_td = TD{_ident = ident, _type = assert}, _body = value}
updateStateDefinition :: DefinitionLine -> Parser ()
updateStateDefinition DL{_td, _body} = do
modify $ defsToIS (M.insert (_ident _td) _body)
modify $ typesToIS (M.insert (_ident _td) (_type _td))
pDef :: Parser DefinitionLine
pDef = lexeme $ label "definition" $ do
skipSpace
ident <- pIdentifier
params <- pManyParams
ascription <- fmap (flip (foldr (uncurry Pi)) params) <$> pAscription
_ <- defChoice $ pure ":="
value <- flip (foldr (uncurry Abs)) params <$> pExpr
res <- checkAscription ident value ascription
_ <- defChoice $ pure ";"
pure res
pDefUpdate :: Parser ()
pDefUpdate = pDef >>= updateStateDefinition
pTerm :: Parser Expr
pTerm =
@ -112,9 +160,9 @@ pTerm =
label "term" $
choice
[ between (char '(') (char ')') pExpr
, pVar
, pStar
, pSquare
, pVar
]
pAppTerm :: Parser Expr
@ -123,20 +171,21 @@ pAppTerm = lexeme $ pLAbs <|> pPAbs <|> pApp
pExpr :: Parser Expr
pExpr = lexeme $ try pArrow <|> pAppTerm
pDef :: Parser ()
pDef = lexeme $ label "definition" $ do
ident <- pIdentifier
_ <- defChoice $ ":=" :| []
value <- pExpr
_ <- defChoice $ ";" :| []
foo <- get
let ty = checkType (_defs foo) [] value
case ty of
Left err -> customFailure $ TE err
Right _ -> modify $ defsToIS $ M.insert ident value
pAscription :: Parser (Maybe Expr)
pAscription = lexeme $ optional $ try $ defChoice (pure ":") >> label "type" pExpr
pProgram :: Parser ()
pProgram = void $ many pDef
pProgram :: Parser Env
pProgram = lexeme $ skipSpace >> many pDefUpdate >> _types . _gs <$> get
pAll :: Text -> Either String ()
pAll input = first errorBundlePretty $ fst $ runIdentity $ runStateT (runParserT pProgram "" input) $ IS{_binds = [], _defs = M.empty}
parseDef :: Text -> State GlobalState (Either String ())
parseDef input = do
env <- get
let (output, IS{_gs}) = runState (runParserT pDefUpdate "" input) (IS{_binds = [], _gs = env})
put _gs
pure $ first errorBundlePretty output
parseDefEmpty :: GlobalState -> Text -> (Either String (), GlobalState)
parseDefEmpty env input = runState (parseDef input) env
parseProgram :: Text -> Either String Env
parseProgram input = first errorBundlePretty $ evalState (runParserT pProgram "" input) IS{_binds = [], _gs = GS{_defs = M.empty, _types = M.empty}}

13
dependent-lambda.cabal → perga.cabal
View file

@ -1,5 +1,5 @@
cabal-version: 3.0
name: dependent-lambda
name: perga
-- PVP summary: +-+------- breaking API changes
-- | | +----- non-breaking API additions
@ -27,12 +27,13 @@ extra-doc-files: CHANGELOG.md
common warnings
ghc-options: -Wall
library dependent-lambda-lib
library perga-lib
import: warnings
exposed-modules: Check
Parser
Expr
Eval
Errors
hs-source-dirs: lib
build-depends: base ^>=4.19.1.0
@ -45,18 +46,19 @@ library dependent-lambda-lib
default-extensions: OverloadedStrings
, GADTs
executable dependent-lambda
executable perga
import: warnings
main-is: Main.hs
other-modules: Repl
build-depends: base ^>=4.19.1.0
, dependent-lambda-lib
, perga-lib
, text
, containers
, haskeline
, directory
, filepath
, mtl
hs-source-dirs: app
default-language: Haskell2010
default-extensions: OverloadedStrings
@ -67,13 +69,12 @@ test-suite tests
type: exitcode-stdio-1.0
main-is: Tests.hs
other-modules: ExprTests
, ParserTests
, CheckTests
build-depends: base ^>=4.19.1.0
, HUnit
, text
, containers
, dependent-lambda-lib
, perga-lib
hs-source-dirs: tests
default-language: Haskell2010
default-extensions: OverloadedStrings

2
test.pg
View file

@ -1,2 +0,0 @@
id := fun (A : *) (x : A) . x ;
foo := fun (A B : *) (f : A -> B) (x : A) . id (A -> B) f (id A x) ;

26
tests/CheckTests.hs
View file

@ -6,20 +6,12 @@ import Expr (Expr (..))
import Test.HUnit
sort :: Test
sort = TestCase $ assertEqual "*" (Right Square) (checkType M.empty [] Star)
stlc :: Test
stlc =
TestCase $
assertEqual
"fun (x : A) (y : B) . x"
(Right $ Pi "" (Var 0 "A") (Pi "" (Var 2 "B") (Var 2 "A")))
(checkType M.empty [Star, Star] $ Abs "x" (Var 0 "A") (Abs "y" (Var 2 "B") (Var 1 "x")))
sort = TestCase $ assertEqual "*" (Right Square) (checkType M.empty Star)
freeVar :: Test
freeVar =
TestCase $
assertEqual "{x = *} , [] |- x : □" (Right Square) (checkType (M.singleton "x" Star) [] (Free "x"))
assertEqual "{x = *} , [] |- x : □" (Right Square) (checkType (M.singleton "x" Star) (Free "x"))
polyIdent :: Test
polyIdent =
@ -27,7 +19,7 @@ polyIdent =
assertEqual
"fun (A : *) (x : A) . x"
(Right $ Pi "A" Star (Pi "" (Var 0 "A") (Var 1 "A")))
(checkType M.empty [] (Abs "A" Star (Abs "x" (Var 0 "A") (Var 0 "x"))))
(checkType M.empty (Abs "A" Star (Abs "x" (Var 0 "A") (Var 0 "x"))))
typeCons :: Test
typeCons =
@ -35,7 +27,7 @@ typeCons =
assertEqual
"fun (A B : *) . A -> B"
(Right $ Pi "" Star (Pi "" Star Star))
(checkType M.empty [] (Abs "A" Star (Abs "B" Star (Pi "" (Var 1 "A") (Var 1 "B")))))
(checkType M.empty (Abs "A" Star (Abs "B" Star (Pi "" (Var 1 "A") (Var 1 "B")))))
useTypeCons :: Test
useTypeCons =
@ -43,7 +35,7 @@ useTypeCons =
assertEqual
"fun (C : * -> *) (A : *) (x : C A) . x"
(Right $ Pi "C" (Pi "" Star Star) (Pi "A" Star (Pi "" (App (Var 1 "C") (Var 0 "A")) (App (Var 2 "C") (Var 1 "A")))))
(checkType M.empty [] $ Abs "C" (Pi "" Star Star) (Abs "A" Star (Abs "x" (App (Var 1 "C") (Var 0 "A")) (Var 0 "x"))))
(checkType M.empty $ Abs "C" (Pi "" Star Star) (Abs "A" Star (Abs "x" (App (Var 1 "C") (Var 0 "A")) (Var 0 "x"))))
dependent :: Test
dependent =
@ -51,7 +43,7 @@ dependent =
assertEqual
"fun (S : *) (x : S) . S -> S"
(Right $ Pi "S" Star (Pi "" (Var 0 "S") Star))
(checkType M.empty [] $ Abs "S" Star (Abs "x" (Var 0 "S") (Pi "" (Var 1 "S") (Var 2 "S"))))
(checkType M.empty $ Abs "S" Star (Abs "x" (Var 0 "S") (Pi "" (Var 1 "S") (Var 2 "S"))))
useDependent :: Test
useDependent =
@ -59,7 +51,7 @@ useDependent =
assertEqual
"fun (S : *) (P : S -> *) (x : S) . P x"
(Right $ Pi "S" Star (Pi "" (Pi "" (Var 0 "S") Star) (Pi "" (Var 1 "S") Star)))
(checkType M.empty [] $ Abs "S" Star (Abs "P" (Pi "" (Var 0 "S") Star) (Abs "x" (Var 1 "S") (App (Var 1 "P") (Var 0 "x")))))
(checkType M.empty $ Abs "S" Star (Abs "P" (Pi "" (Var 0 "S") Star) (Abs "x" (Var 1 "S") (App (Var 1 "P") (Var 0 "x")))))
big :: Test
big =
@ -67,13 +59,13 @@ big =
assertEqual
"fun (S : *) (P Q : S -> *) (H : forall (x : S), P x -> Q x) (G : forall (x : S), P x) (x : S) . H x (G x)"
(Right $ Pi "S" Star (Pi "P" (Pi "" (Var 0 "S") Star) (Pi "Q" (Pi "" (Var 1 "S") Star) (Pi "" (Pi "x" (Var 2 "S") (Pi "" (App (Var 2 "P") (Var 0 "x")) (App (Var 2 "Q") (Var 1 "x")))) (Pi "" (Pi "x" (Var 3 "S") (App (Var 3 "P") (Var 0 "x"))) (Pi "x" (Var 4 "S") (App (Var 3 "Q") (Var 0 "x"))))))))
(checkType M.empty [] $ Abs "S" Star (Abs "P" (Pi "" (Var 0 "S") Star) (Abs "Q" (Pi "" (Var 1 "S") Star) (Abs "H" (Pi "x" (Var 2 "S") (Pi "" (App (Var 2 "P") (Var 0 "x")) (App (Var 2 "Q") (Var 1 "x")))) (Abs "G" (Pi "x" (Var 3 "S") (App (Var 3 "P") (Var 0 "x"))) (Abs "x" (Var 4 "S") (App (App (Var 2 "H") (Var 0 "x")) (App (Var 1 "G") (Var 0 "x")))))))))
(checkType M.empty $ Abs "S" Star (Abs "P" (Pi "" (Var 0 "S") Star) (Abs "Q" (Pi "" (Var 1 "S") Star) (Abs "H" (Pi "x" (Var 2 "S") (Pi "" (App (Var 2 "P") (Var 0 "x")) (App (Var 2 "Q") (Var 1 "x")))) (Abs "G" (Pi "x" (Var 3 "S") (App (Var 3 "P") (Var 0 "x"))) (Abs "x" (Var 4 "S") (App (App (Var 2 "H") (Var 0 "x")) (App (Var 1 "G") (Var 0 "x")))))))))
tests :: Test
tests =
TestList
[ TestLabel "sort" sort
, TestLabel "λ→" $ TestList [stlc, freeVar]
, TestLabel "λ→" $ TestList [freeVar]
, TestLabel "λ2" polyIdent
, TestLabel "λω" $ TestList [typeCons, useTypeCons]
, TestLabel "λP2" $ TestList [dependent, useDependent]

9
tests/ExprTests.hs
View file

@ -35,19 +35,10 @@ substE1 =
after
(subst 0 (Var 2 "B") inner)
whnfE1 :: Test
whnfE1 =
TestCase $
assertEqual
"e1 B"
after
(whnf $ App e1 $ Var 2 "B")
tests :: Test
tests =
TestList
[ TestLabel "fFree" fFree
, TestLabel "incE1" incE1
, TestLabel "substE1" substE1
, TestLabel "whnfE1" whnfE1
]

70
tests/ParserTests.hs
View file

@ -1,70 +0,0 @@
module ParserTests (tests) where
import Expr (Expr (..))
import Parser (pAll)
import Test.HUnit
ident :: Expr
ident =
Abs "A" Star $
Abs "x" (Var 0 "A") $
Var 0 "x"
ident1 :: Test
ident1 =
TestCase $
assertEqual
"ident 1"
(Right ident)
(pAll "lambda (A : *) . lambda (x : A) . x")
ident2 :: Test
ident2 =
TestCase $
assertEqual
"ident 2"
(Right ident)
(pAll "fun (A : *) (x : A) => x")
double :: Expr
double =
Abs "A" Star $
Abs "B" Star $
Abs "f" (Pi "" (Var 1 "A") (Var 1 "B")) $
Abs "g" (Pi "" (Var 2 "A") (Var 2 "B")) $
Abs "x" (Var 3 "A") $
App (Var 2 "f") (Var 0 "x")
doubleTest :: Test
doubleTest =
TestCase $
assertEqual
"double"
(Right double)
(pAll "fun (A B : *) (f g : A -> B) (x : A) => f x")
theorem :: Expr
theorem =
Abs "S" Star $
Abs "P" (Pi "" (Var 0 "S") Star) $
Abs "Q" (Pi "" (Var 1 "S") Star) $
Abs "HP" (Pi "x" (Var 2 "S") (App (Var 2 "P") (Var 0 "x"))) $
Abs "H" (Pi "x" (Var 3 "S") (Pi "" (App (Var 3 "P") (Var 0 "x")) (App (Var 3 "Q") (Var 1 "x")))) $
Abs "x" (Var 4 "S") (App (App (Var 1 "H") (Var 0 "x")) (App (Var 2 "HP") (Var 0 "x")))
theoremTest :: Test
theoremTest =
TestCase $
assertEqual
"theorem"
(Right theorem)
(pAll "fun (S : *) (P Q : S -> *) (HP : ∏ (x : S) . P x) (H : forall (x : S), P x -> Q x) (x : S) => H x (HP x)")
tests :: Test
tests =
TestList
[ TestLabel "ident1" ident1
, TestLabel "ident2" ident2
, TestLabel "double" doubleTest
, TestLabel "theorem" theoremTest
]

2
tests/Tests.hs
View file

@ -2,7 +2,6 @@ module Main where
import qualified CheckTests as C
import qualified ExprTests as E
import qualified ParserTests as P
import qualified System.Exit as Exit
import Test.HUnit
@ -10,7 +9,6 @@ tests :: Test
tests =
TestList
[ TestLabel "ExprTests" E.tests
, TestLabel "ParserTests" P.tests
, TestLabel "CheckTests" C.tests
]