import "@std/debug" import "@std/mem" import "@std/arraylist" import "@source/ast" import "@source/diagnostics" sema :: import "@source/sema" TypeKind :: enum { none, invalid, int, float, string, void, function } Type :: struct { kind TypeKind = .none function ast.NodeId = ast.NodeId(maxval!(u32)) diagnostic diagnostics.DiagnosticId = diagnostics.no_diagnostic } @hide:file Storage :: struct { types []mut Type allocator mem.Allocator } #! owns node-indexed types and retains its allocator; release with deinit. #! function IDs refer to the caller-owned AST, not checker-owned signatures. Result :: struct { storage Storage } @hide:file no_index :: maxval!(usize) @hide:file Role :: enum { none, expression, declaration, parameter, function, annotation, invalid, body } @hide:file Cell :: struct { role Role = .none parent usize = 0 size usize = 1 type Type = Type{} numeric bool = false anchored bool = false demand TypeKind = .none fixed bool = false first_edge usize = no_index } @hide:file Edge :: struct { dependent ast.NodeId next usize } @hide:file State :: struct { tree @ast.Tree resolution @sema.Resolution diagnostics @mut diagnostics.Store types []mut Type cells []mut Cell edges arraylist.ArrayList(Edge) } #! borrows the AST, resolution, and diagnostic store only during checking. #! temporary constraints are freed on every exit; Result owns only its type table. check proc( allocator mem.Allocator, tree @ast.Tree, resolution @sema.Resolution, diagnostic_store @mut diagnostics.Store, ) Result ! (mem.AllocError | diagnostics.CapacityError) { count :: ast.node_count(tree) types :: try mem.alloc(Type, allocator, count) errdefer mem.free(allocator, types) mem.set(types, Type{}) cells :: try mem.alloc(Cell, allocator, count) defer mem.free(allocator, cells) mem.set(cells, Cell{}) for (0..count) |i| cells[i].parent = i queue :: try mem.alloc(ast.NodeId, allocator, count) defer mem.free(allocator, queue) state := State{ tree = tree, resolution = resolution, diagnostics = diagnostic_store, types = types, cells = cells, edges = arraylist.init(allocator), } defer arraylist.deinit(&state.edges) for (0..ast.package_decl_count(tree)) |i| { id :: ast.package_decl(tree, i) match ast.decl_kind(tree, id) { .statement: try collect_declaration(&state, id) .function: try collect_function(&state, id) .invalid: collect_invalid(&state, id) } } # aliases share a type variable; explicit annotations and parameters are boundaries. for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role != .expression or ast.expr_kind(tree, id) != .identifier) continue binding :: sema.binding(resolution, id) if (binding.kind != .declaration) continue target :: binding.data.declaration try depend(&state, target, id) if (cells[usize(target)].fixed) cells[i].type = cells[usize(target)].type } for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role == .declaration and !cells[i].fixed) { join(&state, id, ast.statement_decl(tree, id).initializer) } if (cells[i].role != .expression or ast.expr_kind(tree, id) != .identifier) continue binding :: sema.binding(resolution, id) if (binding.kind == .declaration and !cells[usize(binding.data.declaration)].fixed) { join(&state, id, binding.data.declaration) } } # function aliases are known before any calls or bodies are checked. Honey's # return annotations are scalar, so a call cannot introduce a function type. for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role != .expression or ast.expr_kind(tree, id) != .call) continue callee :: ast.call_expr(tree, id).callee signature :: group_type(&state, callee) if (signature.function != ast.NodeId(maxval!(u32))) { result :: ast.function_decl(tree, signature.function).return_type seed(&state, id, types[usize(result)]) } else if (signature.kind == .invalid) { seed(&state, id, signature) } } for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role != .expression or ast.expr_kind(tree, id) != .call) continue callee :: ast.call_expr(tree, id).callee signature :: group_type(&state, callee) if (signature.function != ast.NodeId(maxval!(u32))) continue if (signature.kind == .invalid) { seed(&state, id, signature) } else if (signature.kind != .none or cells[root(&state, usize(callee))].numeric) { seed(&state, id, invalid_type(try diagnostics.add(diagnostic_store, .not_callable, ast.node_span(tree, callee)))) } else { # without a function anchor this dependency cannot become callable; # share its eventual inference error instead of emitting a cascade. join(&state, id, callee) } } for (0..count) |i| { if (cells[i].role == .none or cells[i].role == .body) continue types[i] = cells[root(&state, i)].type } # reject inherently nonnumeric operations before connecting their operands. # their otherwise valid inputs must not join a failed constraint component. for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role != .expression) continue match ast.expr_kind(tree, id) { .unary: try require_numeric(&state, id, ast.unary_expr(tree, id).operand) .binary: { expression :: ast.binary_expr(tree, id) try require_numeric(&state, id, expression.left) try require_numeric(&state, id, expression.right) } else: {} } } propagate(&state, queue) # numeric edges are undirected constraints, not a traversal-time conversion. # rigid values supply demands without joining the component they constrain. for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role != .expression) continue match ast.expr_kind(tree, id) { .unary: try numeric_operand(&state, id, ast.unary_expr(tree, id).operand) .binary: { expression :: ast.binary_expr(tree, id) try numeric_operand(&state, id, expression.left) try numeric_operand(&state, id, expression.right) } else: {} } } for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role == .declaration) { declaration :: ast.statement_decl(tree, id) if (declaration.type) |annotation| { try demand(&state, declaration.initializer, types[usize(annotation)], .type_mismatch) } } if (cells[i].role != .expression or ast.expr_kind(tree, id) != .call) continue signature :: group_type(&state, ast.call_expr(tree, id).callee) if (signature.function == ast.NodeId(maxval!(u32))) continue parameter_count :: ast.function_param_count(tree, signature.function) for (0..ast.call_arg_count(tree, id)) |j| { if (j >= parameter_count) continue parameter :: ast.function_param(tree, signature.function, j) try demand(&state, ast.call_arg(tree, id, j), cells[usize(parameter)].type, .type_mismatch) } } # default only after the complete program has supplied its numeric demands. for (0..count) |i| { if (root(&state, i) != i) continue if (cells[i].type.kind == .invalid) continue if (cells[i].numeric and cells[i].anchored) { kind TypeKind :: if (cells[i].demand == .float) .float else .int cells[i].type = Type{ kind = kind } } } for (0..count) |i| { id :: ast.NodeId(i) if (cells[i].role != .expression or ast.expr_kind(tree, id) != .call) continue callee :: ast.call_expr(tree, id).callee signature :: group_type(&state, callee) if (signature.function != ast.NodeId(maxval!(u32)) or signature.kind == .none or signature.kind == .invalid) continue if (group_type(&state, id).kind != .invalid) { seed(&state, id, invalid_type(try diagnostics.add(diagnostic_store, .not_callable, ast.node_span(tree, callee)))) } } for (0..count) |i| { if (cells[i].role != .declaration or types[i].kind == .invalid) continue index :: root(&state, i) if (cells[index].type.kind != .none) continue diagnostic :: try diagnostics.add(diagnostic_store, .cannot_infer_type, ast.statement_decl(tree, ast.NodeId(i)).name) cells[index].type = invalid_type(diagnostic) } for (0..count) |i| { if (cells[i].role == .none or cells[i].role == .body or types[i].kind == .invalid) continue types[i] = cells[root(&state, i)].type } for (0..count) |i| try validate(&state, ast.NodeId(i)) # errors travel only from dependencies to users. A rejected use never poisons # an explicitly typed declaration, and body errors do not erase signatures. propagate(&state, queue) return Result{ storage = Storage{ types = types, allocator = allocator } } } #! returns a type without retaining a reference to checker temporaries. type_of proc(result @Result, id ast.NodeId) Type { debug.assert(usize(id) < result.storage.types.len) return result.storage.types[usize(id)] } #! releases type storage using the allocator retained by check. deinit proc(result @mut Result) void { mem.free(result.storage.allocator, result.storage.types) } @hide:file propagate proc(state @mut State, queue []mut ast.NodeId) void { next := usize(0) end := usize(0) for (0..(state.types.len)) |i| { if (state.types[i].kind != .invalid) continue queue[end] = ast.NodeId(i) end += 1 } while (next < end) { source :: queue[next] next += 1 edge := state.cells[usize(source)].first_edge while (edge != no_index) { dependent :: state.edges.items[edge].dependent if (state.types[usize(dependent)].kind != .invalid) { state.types[usize(dependent)] = invalid_type(state.types[usize(source)].diagnostic) queue[end] = dependent end += 1 } edge = state.edges.items[edge].next } } } @hide:file require_numeric proc(state @mut State, expression, operand ast.NodeId) void ! (mem.AllocError | diagnostics.CapacityError) { index :: root(state, usize(operand)) type :: state.cells[index].type if (type.kind == .invalid) { state.types[usize(expression)] = type return } if (type.kind == .none or state.cells[index].numeric or numeric_kind(type.kind)) return try fail(state, expression, .invalid_operand, ast.node_span(state.tree, expression)) } @hide:file invalid_type proc(diagnostic diagnostics.DiagnosticId) Type { return Type{ kind = .invalid, diagnostic = diagnostic } } @hide:file root proc(state @mut State, index usize) usize { current := index while (state.cells[current].parent != current) current = state.cells[current].parent next := index while (state.cells[next].parent != next) { parent :: state.cells[next].parent state.cells[next].parent = current next = parent } return current } @hide:file group_type proc(state @mut State, id ast.NodeId) Type { return state.cells[root(state, usize(id))].type } @hide:file seed proc(state @mut State, id ast.NodeId, type Type) void { index :: root(state, usize(id)) if (state.cells[index].type.kind != .invalid) state.cells[index].type = type } @hide:file join proc(state @mut State, left, right ast.NodeId) void { a := root(state, usize(left)) b := root(state, usize(right)) if (a == b) return if (state.cells[a].size < state.cells[b].size) { saved :: a a = b b = saved } state.cells[b].parent = a state.cells[a].size += state.cells[b].size if (state.cells[a].type.kind == .none or state.cells[b].type.kind == .invalid) { state.cells[a].type = state.cells[b].type } state.cells[a].numeric = state.cells[a].numeric or state.cells[b].numeric state.cells[a].anchored = state.cells[a].anchored or state.cells[b].anchored if (state.cells[a].demand == .none) state.cells[a].demand = state.cells[b].demand } @hide:file depend proc(state @mut State, source, dependent ast.NodeId) void ! mem.AllocError { next :: state.cells[usize(source)].first_edge try arraylist.append(&state.edges, Edge{ dependent = dependent, next = next }) state.cells[usize(source)].first_edge = state.edges.items.len - 1 } @hide:file collect_invalid proc(state @mut State, id ast.NodeId) void { state.cells[usize(id)].role = .invalid state.cells[usize(id)].type = invalid_type(ast.invalid(state.tree, id).diagnostic) } @hide:file check_annotation proc(state @mut State, id ast.NodeId, allow_void bool) Type ! (mem.AllocError | diagnostics.CapacityError) { if (state.cells[usize(id)].role == .annotation) return state.types[usize(id)] state.cells[usize(id)].role = .annotation binding :: sema.binding(state.resolution, id) type := Type{} match binding.kind { .invalid: type = invalid_type(binding.data.diagnostic) .builtin: { type.kind = match binding.data.builtin { .int: .int .float: .float .string: .string .void: .void } if (type.kind == .void and !allow_void) { type = invalid_type(try diagnostics.add(state.diagnostics, .invalid_void_type, ast.node_span(state.tree, id))) } } else: type = invalid_type(try diagnostics.add(state.diagnostics, .expected_type, ast.node_span(state.tree, id))) } state.types[usize(id)] = type state.cells[usize(id)].type = type return type } @hide:file collect_declaration proc(state @mut State, id ast.NodeId) void ! (mem.AllocError | diagnostics.CapacityError) { declaration :: ast.statement_decl(state.tree, id) state.cells[usize(id)].role = .declaration if (declaration.type) |type| { state.cells[usize(id)].fixed = true state.cells[usize(id)].type = try check_annotation(state, type, false) try depend(state, type, id) } binding :: sema.binding(state.resolution, id) if (binding.kind == .invalid) { state.cells[usize(id)].fixed = true state.cells[usize(id)].type = invalid_type(binding.data.diagnostic) } try collect_expression(state, declaration.initializer) try depend(state, declaration.initializer, id) } @hide:file collect_function proc(state @mut State, id ast.NodeId) void ! (mem.AllocError | diagnostics.CapacityError) { tree :: state.tree function :: ast.function_decl(tree, id) state.cells[usize(id)].role = .function state.cells[usize(id)].fixed = true state.cells[usize(id)].type = Type{ kind = .function, function = id } _ = try check_annotation(state, function.return_type, true) try depend(state, function.return_type, id) for (0..ast.function_param_count(tree, id)) |i| { parameter :: ast.function_param(tree, id, i) state.cells[usize(parameter)].role = .parameter state.cells[usize(parameter)].fixed = true state.cells[usize(parameter)].type = try check_annotation(state, ast.param(tree, parameter).type, false) binding :: sema.binding(state.resolution, parameter) if (binding.kind == .invalid) state.cells[usize(parameter)].type = invalid_type(binding.data.diagnostic) try depend(state, ast.param(tree, parameter).type, parameter) try depend(state, parameter, id) } binding :: sema.binding(state.resolution, id) if (binding.kind == .invalid) { state.cells[usize(id)].type = invalid_type(binding.data.diagnostic) state.cells[usize(id)].type.function = id } state.cells[usize(function.body)].role = .body for (0..ast.block_statement_count(tree, function.body)) |i| { statement :: ast.block_statement(tree, function.body, i) match ast.stmt_kind(tree, statement) { .declaration: try collect_declaration(state, statement) .expression: try collect_expression(state, ast.statement_expr(tree, statement)) .invalid: collect_invalid(state, statement) } source :: if (ast.stmt_kind(tree, statement) == .expression) ast.statement_expr(tree, statement) else statement try depend(state, source, function.body) } } @hide:file collect_expression proc(state @mut State, id ast.NodeId) void ! (mem.AllocError | diagnostics.CapacityError) { if (state.cells[usize(id)].role != .none) return tree :: state.tree state.cells[usize(id)].role = .expression match ast.expr_kind(tree, id) { .literal_int: { state.cells[usize(id)].numeric = true state.cells[usize(id)].anchored = true } .literal_float: state.cells[usize(id)].type = Type{ kind = .float } .literal_string: state.cells[usize(id)].type = Type{ kind = .string } .identifier: { binding :: sema.binding(state.resolution, id) match binding.kind { .invalid: state.cells[usize(id)].type = invalid_type(binding.data.diagnostic) .builtin: state.cells[usize(id)].type = invalid_type(try diagnostics.add(state.diagnostics, .type_not_value, ast.node_span(tree, id))) else: {} } } .unary: { state.cells[usize(id)].numeric = true operand :: ast.unary_expr(tree, id).operand try collect_expression(state, operand) try depend(state, operand, id) } .binary: { state.cells[usize(id)].numeric = true expression :: ast.binary_expr(tree, id) try collect_expression(state, expression.left) try collect_expression(state, expression.right) try depend(state, expression.left, id) try depend(state, expression.right, id) } .call: { callee :: ast.call_expr(tree, id).callee try collect_expression(state, callee) try depend(state, callee, id) for (0..ast.call_arg_count(tree, id)) |i| { argument :: ast.call_arg(tree, id, i) try collect_expression(state, argument) try depend(state, argument, id) } } .invalid: collect_invalid(state, id) } } @hide:file numeric_kind proc(kind TypeKind) bool { return kind == .int or kind == .float } @hide:file impose proc(state @mut State, id ast.NodeId, kind TypeKind, code diagnostics.Code, span ast.Span) void ! (mem.AllocError | diagnostics.CapacityError) { index :: root(state, usize(id)) if (state.cells[index].type.kind == .invalid) return state.cells[index].anchored = true if (state.cells[index].demand == .none) { state.cells[index].demand = kind } else if (state.cells[index].demand != kind) { state.cells[index].type = invalid_type(try diagnostics.add(state.diagnostics, code, span)) } } @hide:file numeric_operand proc(state @mut State, expression, operand ast.NodeId) void ! (mem.AllocError | diagnostics.CapacityError) { index :: root(state, usize(operand)) type :: state.cells[index].type if (state.types[usize(expression)].kind == .invalid or state.types[usize(operand)].kind == .invalid) return if (group_type(state, expression).kind == .invalid) return if (state.cells[index].numeric or type.kind == .none) { expression_demand :: state.cells[root(state, usize(expression))].demand operand_demand :: state.cells[index].demand join(state, expression, operand) if (expression_demand != .none) { try impose(state, expression, expression_demand, .invalid_operand, ast.node_span(state.tree, expression)) } if (operand_demand != .none) { try impose(state, expression, operand_demand, .invalid_operand, ast.node_span(state.tree, expression)) } } else if (numeric_kind(type.kind)) { try impose(state, expression, type.kind, .invalid_operand, ast.node_span(state.tree, expression)) } } @hide:file demand proc(state @mut State, id ast.NodeId, expected Type, code diagnostics.Code) void ! (mem.AllocError | diagnostics.CapacityError) { if (!numeric_kind(expected.kind) or state.types[usize(id)].kind == .invalid) return index :: root(state, usize(id)) if (!state.cells[index].numeric and state.cells[index].type.kind != .none) return state.cells[index].numeric = true try impose(state, id, expected.kind, code, ast.node_span(state.tree, id)) } @hide:file fail proc(state @mut State, id ast.NodeId, code diagnostics.Code, span ast.Span) void ! (mem.AllocError | diagnostics.CapacityError) { if (state.types[usize(id)].kind == .invalid) return state.types[usize(id)] = invalid_type(try diagnostics.add(state.diagnostics, code, span)) } @hide:file compatible proc(state @mut State, value ast.NodeId, expected Type) void ! (mem.AllocError | diagnostics.CapacityError) { actual :: state.types[usize(value)] if (actual.kind == .invalid or expected.kind == .invalid) return if (actual.kind != expected.kind) try fail(state, value, .type_mismatch, ast.node_span(state.tree, value)) } @hide:file validate proc(state @mut State, id ast.NodeId) void ! (mem.AllocError | diagnostics.CapacityError) { tree :: state.tree match state.cells[usize(id)].role { .declaration: { declaration :: ast.statement_decl(tree, id) if (state.types[usize(declaration.initializer)].kind == .void) { try fail(state, declaration.initializer, .type_mismatch, ast.node_span(tree, declaration.initializer)) } if (declaration.type) |type| try compatible(state, declaration.initializer, state.types[usize(type)]) } .function: { function :: ast.function_decl(tree, id) result :: state.types[usize(function.return_type)] if (result.kind == .void or result.kind == .invalid or state.types[usize(id)].kind == .invalid) return if (syntactically_valid_body(tree, function.body)) { diagnostic :: try diagnostics.add(state.diagnostics, .missing_return, ast.node_span(tree, function.body)) if (state.types[usize(function.body)].kind != .invalid) { state.types[usize(function.body)] = invalid_type(diagnostic) } } } .expression: { match ast.expr_kind(tree, id) { .unary: { operand :: state.types[usize(ast.unary_expr(tree, id).operand)] if (operand.kind != .invalid and !numeric_kind(operand.kind)) { try fail(state, id, .invalid_operand, ast.node_span(tree, id)) } } .binary: { expression :: ast.binary_expr(tree, id) left :: state.types[usize(expression.left)] right :: state.types[usize(expression.right)] if (left.kind == .invalid or right.kind == .invalid) return if (!numeric_kind(left.kind) or !numeric_kind(right.kind) or left.kind != right.kind) { try fail(state, id, .invalid_operand, expression.span) } else if (expression.operator == .divide and left.kind == .int) { try fail(state, id, .integer_division, expression.span) } } .call: try validate_call(state, id) else: {} } } else: {} } } @hide:file validate_call proc(state @mut State, id ast.NodeId) void ! (mem.AllocError | diagnostics.CapacityError) { tree :: state.tree expression :: ast.call_expr(tree, id) callee :: state.types[usize(expression.callee)] signature :: group_type(state, expression.callee) if (signature.function == ast.NodeId(maxval!(u32))) { if (callee.kind != .invalid) { try fail(state, id, .not_callable, ast.node_span(tree, expression.callee)) } return } parameter_count :: ast.function_param_count(tree, signature.function) if (parameter_count != ast.call_arg_count(tree, id)) { diagnostic :: try diagnostics.add(state.diagnostics, .argument_count_mismatch, expression.span) if (callee.kind != .invalid) state.types[usize(id)] = invalid_type(diagnostic) } for (0..ast.call_arg_count(tree, id)) |i| { if (i >= parameter_count) continue parameter :: ast.function_param(tree, signature.function, i) try compatible(state, ast.call_arg(tree, id, i), state.types[usize(parameter)]) } } @hide:file syntactically_valid_expression proc(tree @ast.Tree, id ast.NodeId) bool { match ast.expr_kind(tree, id) { .invalid: return false .unary: return syntactically_valid_expression(tree, ast.unary_expr(tree, id).operand) .binary: { expression :: ast.binary_expr(tree, id) return syntactically_valid_expression(tree, expression.left) and syntactically_valid_expression(tree, expression.right) } .call: { if (!syntactically_valid_expression(tree, ast.call_expr(tree, id).callee)) return false for (0..ast.call_arg_count(tree, id)) |i| { if (!syntactically_valid_expression(tree, ast.call_arg(tree, id, i))) return false } } else: {} } return true } @hide:file syntactically_valid_body proc(tree @ast.Tree, id ast.NodeId) bool { for (0..ast.block_statement_count(tree, id)) |i| { statement :: ast.block_statement(tree, id, i) match ast.stmt_kind(tree, statement) { .invalid: return false .expression: { if (!syntactically_valid_expression(tree, ast.statement_expr(tree, statement))) return false } .declaration: { if (!syntactically_valid_expression(tree, ast.statement_decl(tree, statement).initializer)) return false } } } return true }