Merge pull request #995 from WardBrian/rewrite-typechecker

Refactor typechecker

Author: WardBrian

README.md

@@ -28,12 +28,44 @@ The entrypoint for the compiler is in `src/stanc/stanc.ml` which sequences the v
 ### Distinct stanc Phases
 
 The phases of stanc are summarized in the following information flowchart and list.
-
+<!---
+digraph G {
+    rankdir=TB;
+    ranksep=.25;
+    bgcolor=white;
+    size=5;
+    node [shape="box"];
+
+    origin[style=invis];
+    stanc[label="stanc/stanc.ml"];
+    lexer[label="frontend/lexer.mll"];
+    parser[label="frontend/parser.mly"];
+    type[label="frontend/Typechecker.ml"];
+    lower[label="frontend/Ast_to_Mir.ml"];
+    transform[label="*_backend/Transform_Mir.ml"];
+    optimize[label="analysis_and_optimization/Optimize.ml"];
+    codegen[label="*_backend/*_code_gen.ml"];
+    output[shape="oval" label=".hpp file"]
+
+
+    origin -> stanc[label=" .stan file path"];
+    stanc -> lexer[label=" string"];
+    lexer -> parser[label=" tokens"];
+    parser -> type[label=" untyped AST"];
+    type -> lower[label=" typed AST"];
+    lower -> transform[label=" MIR"];
+    transform -> optimize[label=" transformed MIR"];
+    transform -> codegen[label="                                  "];
+    optimize -> codegen[headlabel="optimized MIR      "];
+    codegen -> output[label=" C++ code"];
+
+}
+--->
 ![stanc3 information flow](docs/img/information-flow.png)
 
 1. [Lex](src/frontend/lexer.mll) the Stan language into tokens.
 1. [Parse](src/frontend/parser.mly) Stan language into AST that represents the syntax quite closely and aides in development of pretty-printers and linters. `stanc --debug-ast` to print this out.
-1. Typecheck & add type information [Semantic_check.ml](src/frontend/Semantic_check.ml).  `stanc --debug-decorated-ast`
+1. Typecheck & add type information [Typechecker.ml](src/frontend/Typechecker.ml).  `stanc --debug-decorated-ast`
 1. [Lower](src/frontend/Ast_to_Mir.ml) into [Middle Intermediate Representation](src/middle/Mir.ml) (AST -> MIR) `stanc --debug-mir` (or `--debug-mir-pretty`)
 1. Analyze & optimize (MIR -> MIR)
 1. Backend MIR transform (MIR -> MIR) [Transform_Mir.ml](src/stan_math_backend/Transform_Mir.ml)  `stanc --debug-transformed-mir`

docs/img/information-flow.png

@@ -2,7 +2,6 @@
 module Foldable = Foldable
 
 (* Other signatures *)
-module Validation = Validation
 module Pretty = Pretty
 
 (* 'Two-level type' signatures and functors *)

src/common/Common.ml

@@ -4,7 +4,6 @@ module Foldable : module type of Foldable
 
 (** Other signatures *)
 
-module Validation : module type of Validation
 module Pretty : module type of Pretty
 
 (** 'Two-level type' signatures and functors *)

src/common/Common.mli

@@ -878,5 +878,5 @@ let trans_prog filename (p : Ast.typed_program) : Program.Typed.t =
   ; generate_quantities
   ; transform_inits
   ; output_vars
-  ; prog_name= normalize_prog_name !Semantic_check.model_name
+  ; prog_name= normalize_prog_name !Typechecker.model_name
   ; prog_path= filename }

src/common/Validation.ml

@@ -1,7 +1,7 @@
 open Core_kernel
 open Ast
 
-type t = Middle.Warnings.t
+type t = Warnings.t
 
 val find_udf_log_suffix :
   typed_statement -> (string * Middle.UnsizedType.t) option

src/common/Validation.mli

@@ -0,0 +1,48 @@
+open Core_kernel
+open Middle
+
+type originblock =
+  | MathLibrary
+  | Functions
+  | Data
+  | TData
+  | Param
+  | TParam
+  | Model
+  | GQuant
+[@@deriving sexp]
+
+type varinfo = {origin: originblock; global: bool; readonly: bool}
+[@@deriving sexp]
+
+type info =
+  { type_: UnsizedType.t
+  ; kind:
+      [ `Variable of varinfo
+      | `UserDeclared of Location_span.t
+      | `StanMath
+      | `UserDefined ] }
+[@@deriving sexp]
+
+type t = info list String.Map.t
+
+let create () =
+  let functions =
+    Hashtbl.to_alist Stan_math_signatures.stan_math_signatures
+    |> List.map ~f:(fun (key, values) ->
+           ( key
+           , List.map values ~f:(fun (rt, args, mem) ->
+                 let type_ =
+                   UnsizedType.UFun
+                     (args, rt, Fun_kind.suffix_from_name key, mem)
+                 in
+                 {type_; kind= `StanMath} ) ) )
+    |> String.Map.of_alist_exn
+  in
+  functions
+
+let add env key type_ kind = Map.add_multi env ~key ~data:{type_; kind}
+let set_raw env key data = Map.set env ~key ~data
+let find env key = Map.find_multi env key
+let mem env key = Map.mem env key
+let iter env f = Map.iter env ~f

src/frontend/Ast_to_Mir.ml

@@ -0,0 +1,53 @@
+(** Type environments used during typechecking. Maps from strings to function or variable information *)
+
+open Middle
+
+(** Origin blocks, to keep track of where variables are declared *)
+type originblock =
+  | MathLibrary
+  | Functions
+  | Data
+  | TData
+  | Param
+  | TParam
+  | Model
+  | GQuant
+[@@deriving sexp]
+
+(** Information available for each variable *)
+type varinfo = {origin: originblock; global: bool; readonly: bool}
+[@@deriving sexp]
+
+type info =
+  { type_: UnsizedType.t
+  ; kind:
+      [ `Variable of varinfo
+      | `UserDeclared of Location_span.t
+      | `StanMath
+      | `UserDefined ] }
+[@@deriving sexp]
+
+type t
+
+val create : unit -> t
+(** Return a new type environment which contains the Stan math library functions
+*)
+
+val find : t -> string -> info list
+
+val add :
+     t
+  -> string
+  -> Middle.UnsizedType.t
+  -> [ `UserDeclared of Location_span.t
+     | `StanMath
+     | `UserDefined
+     | `Variable of varinfo ]
+  -> t
+(** Add a new item to the type environment. Does not overwrite existing, but shadows *)
+
+val set_raw : t -> string -> info list -> t
+(** Overwrite the existing items bound to a name *)
+
+val mem : t -> string -> bool
+val iter : t -> (info list -> unit) -> unit