2024-10-31 07:03:33 +00:00
|
|
|
/*
|
|
|
|
This file is part of TON Blockchain Library.
|
|
|
|
|
|
|
|
TON Blockchain Library is free software: you can redistribute it and/or modify
|
|
|
|
it under the terms of the GNU Lesser General Public License as published by
|
|
|
|
the Free Software Foundation, either version 2 of the License, or
|
|
|
|
(at your option) any later version.
|
|
|
|
|
|
|
|
TON Blockchain Library is distributed in the hope that it will be useful,
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
GNU Lesser General Public License for more details.
|
|
|
|
|
|
|
|
You should have received a copy of the GNU Lesser General Public License
|
|
|
|
along with TON Blockchain Library. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
*/
|
|
|
|
#include "ast.h"
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
#ifdef TOLK_DEBUG
|
2024-10-31 07:03:33 +00:00
|
|
|
#include "ast-stringifier.h"
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
#endif
|
2024-10-31 07:03:33 +00:00
|
|
|
|
|
|
|
namespace tolk {
|
|
|
|
|
|
|
|
static_assert(sizeof(ASTNodeBase) == 12);
|
|
|
|
|
|
|
|
#ifdef TOLK_DEBUG
|
|
|
|
|
|
|
|
std::string ASTNodeBase::to_debug_string(bool colored) const {
|
|
|
|
ASTStringifier s(colored);
|
|
|
|
return s.to_string_with_children(this);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ASTNodeBase::debug_print() const {
|
|
|
|
std::cerr << to_debug_string(true) << std::endl;
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif // TOLK_DEBUG
|
|
|
|
|
|
|
|
UnexpectedASTNodeType::UnexpectedASTNodeType(AnyV v_unexpected, const char* place_where): v_unexpected(v_unexpected) {
|
|
|
|
message = "Unexpected ASTNodeType ";
|
|
|
|
#ifdef TOLK_DEBUG
|
|
|
|
message += ASTStringifier::ast_node_type_to_string(v_unexpected->type);
|
|
|
|
message += " ";
|
|
|
|
#endif
|
|
|
|
message += "in ";
|
|
|
|
message += place_where;
|
|
|
|
}
|
|
|
|
|
|
|
|
void ASTNodeBase::error(const std::string& err_msg) const {
|
|
|
|
throw ParseError(loc, err_msg);
|
|
|
|
}
|
|
|
|
|
2024-10-31 07:11:41 +00:00
|
|
|
AnnotationKind Vertex<ast_annotation>::parse_kind(std::string_view name) {
|
|
|
|
if (name == "@pure") {
|
|
|
|
return AnnotationKind::pure;
|
|
|
|
}
|
|
|
|
if (name == "@inline") {
|
|
|
|
return AnnotationKind::inline_simple;
|
|
|
|
}
|
|
|
|
if (name == "@inline_ref") {
|
|
|
|
return AnnotationKind::inline_ref;
|
|
|
|
}
|
|
|
|
if (name == "@method_id") {
|
|
|
|
return AnnotationKind::method_id;
|
|
|
|
}
|
|
|
|
if (name == "@deprecated") {
|
|
|
|
return AnnotationKind::deprecated;
|
|
|
|
}
|
|
|
|
return AnnotationKind::unknown;
|
|
|
|
}
|
|
|
|
|
|
|
|
int Vertex<ast_genericsT_list>::lookup_idx(std::string_view nameT) const {
|
2024-10-31 07:03:33 +00:00
|
|
|
for (size_t idx = 0; idx < children.size(); ++idx) {
|
2024-10-31 07:11:41 +00:00
|
|
|
if (children[idx] && children[idx]->as<ast_genericsT_item>()->nameT == nameT) {
|
2024-10-31 07:03:33 +00:00
|
|
|
return static_cast<int>(idx);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
2024-10-31 07:11:41 +00:00
|
|
|
int Vertex<ast_parameter_list>::lookup_idx(std::string_view param_name) const {
|
2024-10-31 07:03:33 +00:00
|
|
|
for (size_t idx = 0; idx < children.size(); ++idx) {
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
if (children[idx] && children[idx]->as<ast_parameter>()->param_name == param_name) {
|
2024-10-31 07:03:33 +00:00
|
|
|
return static_cast<int>(idx);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
2024-10-31 07:18:54 +00:00
|
|
|
int Vertex<ast_parameter_list>::get_mutate_params_count() const {
|
|
|
|
int n = 0;
|
|
|
|
for (AnyV param : children) {
|
|
|
|
if (param->as<ast_parameter>()->declared_as_mutate) {
|
|
|
|
n++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return n;
|
|
|
|
}
|
|
|
|
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
// ---------------------------------------------------------
|
|
|
|
// "assign" methods
|
|
|
|
//
|
|
|
|
// From the user's point of view, all AST vertices are constant, fields are public, but can't be modified.
|
|
|
|
// The only way to modify a field is to call "mutate()" and then use these "assign_*" methods.
|
|
|
|
// Therefore, there is a guarantee, that all AST mutations are done via these methods,
|
|
|
|
// easily searched by usages, and there is no another way to modify any other field.
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void ASTNodeExpressionBase::assign_inferred_type(TypePtr type) {
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
this->inferred_type = type;
|
|
|
|
}
|
|
|
|
|
|
|
|
void ASTNodeExpressionBase::assign_rvalue_true() {
|
|
|
|
this->is_rvalue = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
void ASTNodeExpressionBase::assign_lvalue_true() {
|
|
|
|
this->is_lvalue = true;
|
|
|
|
}
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void Vertex<ast_reference>::assign_sym(const Symbol* sym) {
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
this->sym = sym;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_function_call>::assign_fun_ref(const FunctionData* fun_ref) {
|
|
|
|
this->fun_maybe = fun_ref;
|
|
|
|
}
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void Vertex<ast_cast_as_operator>::assign_resolved_type(TypePtr cast_to_type) {
|
|
|
|
this->cast_to_type = cast_to_type;
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_global_var_declaration>::assign_var_ref(const GlobalVarData* var_ref) {
|
|
|
|
this->var_ref = var_ref;
|
|
|
|
}
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void Vertex<ast_global_var_declaration>::assign_resolved_type(TypePtr declared_type) {
|
|
|
|
this->declared_type = declared_type;
|
|
|
|
}
|
|
|
|
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
void Vertex<ast_constant_declaration>::assign_const_ref(const GlobalConstData* const_ref) {
|
|
|
|
this->const_ref = const_ref;
|
|
|
|
}
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void Vertex<ast_constant_declaration>::assign_resolved_type(TypePtr declared_type) {
|
|
|
|
this->declared_type = declared_type;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_instantiationT_item>::assign_resolved_type(TypePtr substituted_type) {
|
|
|
|
this->substituted_type = substituted_type;
|
|
|
|
}
|
|
|
|
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
void Vertex<ast_parameter>::assign_param_ref(const LocalVarData* param_ref) {
|
|
|
|
this->param_ref = param_ref;
|
|
|
|
}
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void Vertex<ast_parameter>::assign_resolved_type(TypePtr declared_type) {
|
|
|
|
this->declared_type = declared_type;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_set_assign>::assign_fun_ref(const FunctionData* fun_ref) {
|
|
|
|
this->fun_ref = fun_ref;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_unary_operator>::assign_fun_ref(const FunctionData* fun_ref) {
|
|
|
|
this->fun_ref = fun_ref;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_binary_operator>::assign_fun_ref(const FunctionData* fun_ref) {
|
|
|
|
this->fun_ref = fun_ref;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_dot_access>::assign_target(const DotTarget& target) {
|
|
|
|
this->target = target;
|
|
|
|
}
|
|
|
|
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
void Vertex<ast_function_declaration>::assign_fun_ref(const FunctionData* fun_ref) {
|
|
|
|
this->fun_ref = fun_ref;
|
|
|
|
}
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void Vertex<ast_function_declaration>::assign_resolved_type(TypePtr declared_return_type) {
|
|
|
|
this->declared_return_type = declared_return_type;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_local_var_lhs>::assign_var_ref(const LocalVarData* var_ref) {
|
|
|
|
this->var_ref = var_ref;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Vertex<ast_local_var_lhs>::assign_resolved_type(TypePtr declared_type) {
|
|
|
|
this->declared_type = declared_type;
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
}
|
|
|
|
|
[Tolk] Rewrite the type system from Hindley-Milner to static typing
FunC's (and Tolk's before this PR) type system is based on Hindley-Milner.
This is a common approach for functional languages, where
types are inferred from usage through unification.
As a result, type declarations are not necessary:
() f(a,b) { return a+b; } // a and b now int, since `+` (int, int)
While this approach works for now, problems arise with the introduction
of new types like bool, where `!x` must handle both int and bool.
It will also become incompatible with int32 and other strict integers.
This will clash with structure methods, struggle with proper generics,
and become entirely impractical for union types.
This PR completely rewrites the type system targeting the future.
1) type of any expression is inferred and never changed
2) this is available because dependent expressions already inferred
3) forall completely removed, generic functions introduced
(they work like template functions actually, instantiated while inferring)
4) instantiation `<...>` syntax, example: `t.tupleAt<int>(0)`
5) `as` keyword, for example `t.tupleAt(0) as int`
6) methods binding is done along with type inferring, not before
("before", as worked previously, was always a wrong approach)
2024-12-30 15:31:27 +00:00
|
|
|
void Vertex<ast_import_directive>::assign_src_file(const SrcFile* file) {
|
[Tolk] AST-based semantic analysis, get rid of Expr
This is a huge refactoring focusing on untangling compiler internals
(previously forked from FunC).
The goal is to convert AST directly to Op (a kind of IR representation),
doing all code analysis at AST level.
Noteable changes:
- AST-based semantic kernel includes: registering global symbols,
scope handling and resolving local/global identifiers,
lvalue/rvalue calc and check, implicit return detection,
mutability analysis, pure/impure validity checks,
simple constant folding
- values of `const` variables are calculated NOT based on CodeBlob,
but via a newly-introduced AST-based constant evaluator
- AST vertices are now inherited from expression/statement/other;
expression vertices have common properties (TypeExpr, lvalue/rvalue)
- symbol table is rewritten completely, SymDef/SymVal no longer exist,
lexer now doesn't need to register identifiers
- AST vertices have references to symbols, filled at different
stages of pipeline
- the remaining "FunC legacy part" is almost unchanged besides Expr
which was fully dropped; AST is converted to Ops (IR) directly
2024-12-16 18:19:45 +00:00
|
|
|
this->file = file;
|
2024-10-31 07:04:58 +00:00
|
|
|
}
|
|
|
|
|
2024-10-31 07:03:33 +00:00
|
|
|
} // namespace tolk
|