Foundation for fine-grained incremental checking (#2838)

See #2838 for details.

Author: JukkaL

mypy/build.py

@@ -65,8 +65,9 @@ class BuildResult:
       errors:  List of error messages.
     """
 
-    def __init__(self, manager: 'BuildManager') -> None:
+    def __init__(self, manager: 'BuildManager', graph: Graph) -> None:
         self.manager = manager
+        self.graph = graph
         self.files = manager.modules
         self.types = manager.all_types
         self.errors = manager.errors.messages()
@@ -184,8 +185,8 @@ def build(sources: List[BuildSource],
                            )
 
     try:
-        dispatch(sources, manager)
-        return BuildResult(manager)
+        graph = dispatch(sources, manager)
+        return BuildResult(manager, graph)
     finally:
         manager.log("Build finished in %.3f seconds with %d modules, %d types, and %d errors" %
                     (time.time() - manager.start_time,
@@ -474,7 +475,7 @@ def parse_file(self, id: str, path: str, source: str, ignore_errors: bool) -> My
         return tree
 
     def module_not_found(self, path: str, line: int, id: str) -> None:
-        self.errors.set_file(path)
+        self.errors.set_file(path, id)
         stub_msg = "(Stub files are from https://github.com/python/typeshed)"
         if ((self.options.python_version[0] == 2 and moduleinfo.is_py2_std_lib_module(id)) or
                 (self.options.python_version[0] >= 3 and moduleinfo.is_py3_std_lib_module(id))):
@@ -1230,7 +1231,7 @@ def skipping_ancestor(self, id: str, path: str, ancestor_for: 'State') -> None:
         # so we'd need to cache the decision.
         manager = self.manager
         manager.errors.set_import_context([])
-        manager.errors.set_file(ancestor_for.xpath)
+        manager.errors.set_file(ancestor_for.xpath, ancestor_for.id)
         manager.errors.report(-1, -1, "Ancestor package '%s' ignored" % (id,),
                               severity='note', only_once=True)
         manager.errors.report(-1, -1,
@@ -1242,7 +1243,7 @@ def skipping_module(self, id: str, path: str) -> None:
         manager = self.manager
         save_import_context = manager.errors.import_context()
         manager.errors.set_import_context(self.caller_state.import_context)
-        manager.errors.set_file(self.caller_state.xpath)
+        manager.errors.set_file(self.caller_state.xpath, self.caller_state.id)
         line = self.caller_line
         manager.errors.report(line, 0,
                               "Import of '%s' ignored" % (id,),
@@ -1429,7 +1430,7 @@ def parse_file(self) -> None:
                 continue
             if id == '':
                 # Must be from a relative import.
-                manager.errors.set_file(self.xpath)
+                manager.errors.set_file(self.xpath, self.id)
                 manager.errors.report(line, 0,
                                       "No parent module -- cannot perform relative import",
                                       blocker=True)
@@ -1545,20 +1546,21 @@ def write_cache(self) -> None:
                 self.interface_hash = new_interface_hash
 
 
-def dispatch(sources: List[BuildSource], manager: BuildManager) -> None:
+def dispatch(sources: List[BuildSource], manager: BuildManager) -> Graph:
     manager.log("Mypy version %s" % __version__)
     graph = load_graph(sources, manager)
     if not graph:
         print("Nothing to do?!")
-        return
+        return graph
     manager.log("Loaded graph with %d nodes" % len(graph))
     if manager.options.dump_graph:
         dump_graph(graph)
-        return
+        return graph
     process_graph(graph, manager)
     if manager.options.warn_unused_ignores:
         # TODO: This could also be a per-module option.
         manager.errors.generate_unused_ignore_notes()
+    return graph
 
 
 class NodeInfo:
@@ -1633,7 +1635,7 @@ def load_graph(sources: List[BuildSource], manager: BuildManager) -> Graph:
         except ModuleNotFound:
             continue
         if st.id in graph:
-            manager.errors.set_file(st.xpath)
+            manager.errors.set_file(st.xpath, st.id)
             manager.errors.report(-1, -1, "Duplicate module named '%s'" % st.id)
             manager.errors.raise_error()
         graph[st.id] = st

mypy/checker.py

@@ -65,13 +65,16 @@
 LAST_PASS = 1  # Pass numbers start at 0
 
 
-# A node which is postponed to be type checked during the next pass.
+# A node which is postponed to be processed during the next pass.
+# This is used for both batch mode and fine-grained incremental mode.
 DeferredNode = NamedTuple(
     'DeferredNode',
     [
-        ('node', FuncItem),
+        # In batch mode only FuncDef and LambdaExpr are supported
+        ('node', Union[FuncDef, LambdaExpr, MypyFile]),
         ('context_type_name', Optional[str]),  # Name of the surrounding class (for error messages)
-        ('active_class', Optional[Type]),  # And its type (for selftype handling)
+        ('active_typeinfo', Optional[TypeInfo]),  # And its TypeInfo (for semantic analysis
+                                                  # self type handling)
     ])
 
 
@@ -167,7 +170,7 @@ def check_first_pass(self) -> None:
 
         Deferred functions will be processed by check_second_pass().
         """
-        self.errors.set_file(self.path)
+        self.errors.set_file(self.path, self.tree.fullname())
         with self.enter_partial_types():
             with self.binder.top_frame_context():
                 for d in self.tree.defs:
@@ -187,38 +190,57 @@ def check_first_pass(self) -> None:
                 self.fail(messages.ALL_MUST_BE_SEQ_STR.format(str_seq_s, all_s),
                           all_.node)
 
-    def check_second_pass(self) -> bool:
+    def check_second_pass(self, todo: List[DeferredNode] = None) -> bool:
         """Run second or following pass of type checking.
 
         This goes through deferred nodes, returning True if there were any.
         """
-        if not self.deferred_nodes:
+        if not todo and not self.deferred_nodes:
             return False
-        self.errors.set_file(self.path)
+        self.errors.set_file(self.path, self.tree.fullname())
         self.pass_num += 1
-        todo = self.deferred_nodes
+        if not todo:
+            todo = self.deferred_nodes
+        else:
+            assert not self.deferred_nodes
         self.deferred_nodes = []
-        done = set()  # type: Set[FuncItem]
-        for node, type_name, active_class in todo:
+        done = set()  # type: Set[Union[FuncDef, LambdaExpr, MypyFile]]
+        for node, type_name, active_typeinfo in todo:
             if node in done:
                 continue
             # This is useful for debugging:
             # print("XXX in pass %d, class %s, function %s" %
             #       (self.pass_num, type_name, node.fullname() or node.name()))
             done.add(node)
             with self.errors.enter_type(type_name) if type_name else nothing():
-                with self.scope.push_class(active_class) if active_class else nothing():
-                    if isinstance(node, Statement):
-                        self.accept(node)
-                    elif isinstance(node, Expression):
-                        self.expr_checker.accept(node)
-                    else:
-                        assert False
+                with self.scope.push_class(active_typeinfo) if active_typeinfo else nothing():
+                    self.check_partial(node)
         return True
 
+    def check_partial(self, node: Union[FuncDef, LambdaExpr, MypyFile]) -> None:
+        if isinstance(node, MypyFile):
+            self.check_top_level(node)
+        elif isinstance(node, LambdaExpr):
+            self.expr_checker.accept(node)
+        else:
+            self.accept(node)
+
+    def check_top_level(self, node: MypyFile) -> None:
+        """Check only the top-level of a module, skipping function definitions."""
+        with self.enter_partial_types():
+            with self.binder.top_frame_context():
+                for d in node.defs:
+                    # TODO: Type check class bodies.
+                    if not isinstance(d, (FuncDef, ClassDef)):
+                        d.accept(self)
+
+        assert not self.current_node_deferred
+        # TODO: Handle __all__
+
     def handle_cannot_determine_type(self, name: str, context: Context) -> None:
         node = self.scope.top_function()
-        if self.pass_num < LAST_PASS and node is not None:
+        if (self.pass_num < LAST_PASS and node is not None
+                and isinstance(node, (FuncDef, LambdaExpr))):
             # Don't report an error yet. Just defer.
             if self.errors.type_name:
                 type_name = self.errors.type_name[-1]
@@ -635,7 +657,7 @@ def is_implicit_any(t: Type) -> bool:
                 for i in range(len(typ.arg_types)):
                     arg_type = typ.arg_types[i]
 
-                    ref_type = self.scope.active_class()
+                    ref_type = self.scope.active_self_type()  # type: Optional[Type]
                     if (isinstance(defn, FuncDef) and ref_type is not None and i == 0
                             and not defn.is_static
                             and typ.arg_kinds[0] not in [nodes.ARG_STAR, nodes.ARG_STAR2]):
@@ -946,7 +968,7 @@ def check_method_override_for_base_with_name(
             # The name of the method is defined in the base class.
 
             # Construct the type of the overriding method.
-            typ = bind_self(self.function_type(defn), self.scope.active_class())
+            typ = bind_self(self.function_type(defn), self.scope.active_self_type())
             # Map the overridden method type to subtype context so that
             # it can be checked for compatibility.
             original_type = base_attr.type
@@ -959,7 +981,7 @@ def check_method_override_for_base_with_name(
                     assert False, str(base_attr.node)
             if isinstance(original_type, FunctionLike):
                 original = map_type_from_supertype(
-                    bind_self(original_type, self.scope.active_class()),
+                    bind_self(original_type, self.scope.active_self_type()),
                     defn.info, base)
                 # Check that the types are compatible.
                 # TODO overloaded signatures
@@ -1051,7 +1073,7 @@ def visit_class_def(self, defn: ClassDef) -> None:
             old_binder = self.binder
             self.binder = ConditionalTypeBinder()
             with self.binder.top_frame_context():
-                with self.scope.push_class(fill_typevars(defn.info)):
+                with self.scope.push_class(defn.info):
                     self.accept(defn.defs)
             self.binder = old_binder
             if not defn.has_incompatible_baseclass:
@@ -1317,8 +1339,8 @@ def check_compatibility_super(self, lvalue: NameExpr, lvalue_type: Type, rvalue:
                     # Class-level function objects and classmethods become bound
                     # methods: the former to the instance, the latter to the
                     # class
-                    base_type = bind_self(base_type, self.scope.active_class())
-                    compare_type = bind_self(compare_type, self.scope.active_class())
+                    base_type = bind_self(base_type, self.scope.active_self_type())
+                    compare_type = bind_self(compare_type, self.scope.active_self_type())
 
                 # If we are a static method, ensure to also tell the
                 # lvalue it now contains a static method
@@ -1347,7 +1369,8 @@ def lvalue_type_from_base(self, expr_node: Var,
 
             if base_type:
                 if not has_no_typevars(base_type):
-                    instance = cast(Instance, self.scope.active_class())
+                    # TODO: Handle TupleType, don't cast
+                    instance = cast(Instance, self.scope.active_self_type())
                     itype = map_instance_to_supertype(instance, base)
                     base_type = expand_type_by_instance(base_type, itype)
 
@@ -2996,7 +3019,7 @@ def is_node_static(node: Node) -> Optional[bool]:
 
 class Scope:
     # We keep two stacks combined, to maintain the relative order
-    stack = None  # type: List[Union[Type, FuncItem, MypyFile]]
+    stack = None  # type: List[Union[TypeInfo, FuncItem, MypyFile]]
 
     def __init__(self, module: MypyFile) -> None:
         self.stack = [module]
@@ -3007,20 +3030,26 @@ def top_function(self) -> Optional[FuncItem]:
                 return e
         return None
 
-    def active_class(self) -> Optional[Type]:
-        if isinstance(self.stack[-1], Type):
+    def active_class(self) -> Optional[TypeInfo]:
+        if isinstance(self.stack[-1], TypeInfo):
             return self.stack[-1]
         return None
 
+    def active_self_type(self) -> Optional[Union[Instance, TupleType]]:
+        info = self.active_class()
+        if info:
+            return fill_typevars(info)
+        return None
+
     @contextmanager
     def push_function(self, item: FuncItem) -> Iterator[None]:
         self.stack.append(item)
         yield
         self.stack.pop()
 
     @contextmanager
-    def push_class(self, t: Type) -> Iterator[None]:
-        self.stack.append(t)
+    def push_class(self, info: TypeInfo) -> Iterator[None]:
+        self.stack.append(info)
         yield
         self.stack.pop()