From e8eb4635408f6a106c7cf6afa59f5368f35f8bd6 Mon Sep 17 00:00:00 2001
From: "Jonathan D.A. Jewell" <6759885+hyperpolymath@users.noreply.github.com>
Date: Mon, 11 May 2026 02:37:36 +0200
Subject: [PATCH] feat: add extern type / extern fn parsing and codegen (closes
 #42)

extern was not a recognised keyword, causing a parse error at character 1 of
any file containing extern declarations.  Adds:

  - Token.EXTERN in lib/token.ml
  - ("extern", EXTERN) in the lexer keyword table (lib/lexer.ml)
  - TopExternType { et_name } and TopExternFn { ef_name; ef_params; ef_ret_ty }
    variants in the AST (lib/ast.ml)
  - extern_type_decl and extern_fn_decl parser rules in lib/parser.mly;
    both added to the top_level production
  - gen_decl cases in lib/codegen.ml: TopExternType is a no-op (opaque type
    for the type-checker); TopExternFn adds a Wasm import with module "env"
    and registers the name in func_indices so call sites resolve correctly

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 lib/ast.ml     |  8 ++++++++
 lib/codegen.ml | 22 ++++++++++++++++++++++
 lib/lexer.ml   |  1 +
 lib/parser.mly | 14 +++++++++++++-
 lib/token.ml   |  2 ++
 5 files changed, 46 insertions(+), 1 deletion(-)

diff --git a/lib/ast.ml b/lib/ast.ml
index 3ab2798..a25960c 100644
--- a/lib/ast.ml
+++ b/lib/ast.ml
@@ -367,6 +367,14 @@ type top_level =
       tc_ty : type_expr;
       tc_value : expr;
     }
+  | TopExternType of {
+      et_name : ident;
+    }
+  | TopExternFn of {
+      ef_name : ident;
+      ef_params : param list;
+      ef_ret_ty : type_expr option;
+    }
 [@@deriving show, eq]
 
 (** Complete program *)
diff --git a/lib/codegen.ml b/lib/codegen.ml
index 20bef89..9acdb1c 100644
--- a/lib/codegen.ml
+++ b/lib/codegen.ml
@@ -1871,6 +1871,28 @@ let gen_decl (ctx : context) (decl : top_level) : context result =
     (* These declarations don't generate code *)
     Ok ctx
 
+  | TopExternType _ ->
+    (* Opaque host type — no code generated; type is available to the type-checker *)
+    Ok ctx
+
+  | TopExternFn ef ->
+    (* Add a WebAssembly import for the extern fn declaration.
+       Module name defaults to "env" (conventional host environment namespace). *)
+    let param_types = List.map (fun _ -> I32) ef.ef_params in
+    let result_types = match ef.ef_ret_ty with
+      | None -> []
+      | Some _ -> [I32]
+    in
+    let func_type = { ft_params = param_types; ft_results = result_types } in
+    let type_idx = List.length ctx.types in
+    let ctx_with_type = { ctx with types = ctx.types @ [func_type] } in
+    let func_idx = import_func_count ctx_with_type in
+    let import_entry = { i_module = "env"; i_name = ef.ef_name.name;
+                         i_desc = ImportFunc type_idx } in
+    Ok { ctx_with_type with
+         imports = ctx_with_type.imports @ [import_entry];
+         func_indices = (ef.ef_name.name, func_idx) :: ctx_with_type.func_indices }
+
 (** Generate WASM module from AffineScript program *)
 let generate_module (prog : program) : wasm_module result =
   let ctx = create_context () in
diff --git a/lib/lexer.ml b/lib/lexer.ml
index 445f859..cc4f120 100644
--- a/lib/lexer.ml
+++ b/lib/lexer.ml
@@ -41,6 +41,7 @@ let () =
       ("use", USE);
       ("pub", PUB);
       ("as", AS);
+      ("extern", EXTERN);
       ("unsafe", UNSAFE);
       ("assume", ASSUME);
       ("transmute", TRANSMUTE);
diff --git a/lib/parser.mly b/lib/parser.mly
index cb140b1..16346c4 100644
--- a/lib/parser.mly
+++ b/lib/parser.mly
@@ -39,7 +39,7 @@ let mk_ident name startpos endpos =
 %token FN LET CONST MUT OWN REF TYPE STRUCT ENUM TRAIT IMPL
 %token EFFECT HANDLE RESUME MATCH IF ELSE WHILE FOR
 %token RETURN BREAK CONTINUE IN WHERE TOTAL MODULE USE
-%token PUB AS UNSAFE ASSUME TRANSMUTE FORGET TRY CATCH FINALLY
+%token PUB AS EXTERN UNSAFE ASSUME TRANSMUTE FORGET TRY CATCH FINALLY
 
 /* Built-in types */
 %token NAT INT_T BOOL FLOAT_T STRING_T CHAR_T TYPE_K ROW NEVER
@@ -121,12 +121,24 @@ top_level:
   | tr = trait_decl { TopTrait tr }
   | i = impl_block { TopImpl i }
   | c = const_decl { c }
+  | e = extern_type_decl { e }
+  | e = extern_fn_decl { e }
 
 const_decl:
   | vis = visibility? CONST name = ident COLON ty = type_expr EQ value = expr SEMICOLON
     { TopConst { tc_vis = Option.value vis ~default:Private;
                  tc_name = name; tc_ty = ty; tc_value = value } }
 
+extern_type_decl:
+  | EXTERN TYPE name = upper_ident SEMICOLON
+    { TopExternType { et_name = name } }
+
+extern_fn_decl:
+  | EXTERN FN name = ident LPAREN params = separated_list(COMMA, param) RPAREN SEMICOLON
+    { TopExternFn { ef_name = name; ef_params = params; ef_ret_ty = None } }
+  | EXTERN FN name = ident LPAREN params = separated_list(COMMA, param) RPAREN ARROW ret = type_expr SEMICOLON
+    { TopExternFn { ef_name = name; ef_params = params; ef_ret_ty = Some ret } }
+
 /* ========== Functions ========== */
 
 fn_decl:
diff --git a/lib/token.ml b/lib/token.ml
index d36a009..993e69b 100644
--- a/lib/token.ml
+++ b/lib/token.ml
@@ -47,6 +47,7 @@ type t =
   | USE
   | PUB
   | AS
+  | EXTERN
   | UNSAFE
   | ASSUME
   | SELF_KW     (** self receiver keyword *)
@@ -164,6 +165,7 @@ let to_string = function
   | USE -> "use"
   | PUB -> "pub"
   | AS -> "as"
+  | EXTERN -> "extern"
   | UNSAFE -> "unsafe"
   | ASSUME -> "assume"
   | SELF_KW -> "self"