Rename CodebaseGraph to CodebaseContext (#405)

`CodebaseGraph` is misleading and has been a deprecated term since the
Codebase Initialization Refactor.

Author: EdwardJXLi

src/codegen/runner/diff/get_raw_diff.py

@@ -27,7 +27,7 @@ def patch_to_limited_diff_string(patch, codebase: Codebase, max_lines=10000):
 
     # Add flags that are not in the diff
     filenames = [patched_file.path for patched_file in patch]
-    flags_not_in_diff = list(filter(lambda flag: flag.symbol.filepath not in filenames, codebase.G.flags._flags))
+    flags_not_in_diff = list(filter(lambda flag: flag.symbol.filepath not in filenames, codebase.ctx.flags._flags))
 
     for flag in flags_not_in_diff:
         filename = flag.symbol.filepath
@@ -39,7 +39,7 @@ def patch_to_limited_diff_string(patch, codebase: Codebase, max_lines=10000):
         patch.append(patched_file)
 
     for patched_file in patch:
-        filtered_flags = filter(lambda flag: flag.symbol.filepath == patched_file.path, codebase.G.flags._flags)
+        filtered_flags = filter(lambda flag: flag.symbol.filepath == patched_file.path, codebase.ctx.flags._flags)
         sorted_flags = list(map(lambda flag: flag.symbol.start_point.row + 1, filtered_flags))
         sorted_flags.sort()
 

src/codegen/runner/sandbox/executor.py

@@ -36,8 +36,8 @@ async def find_flags(self, execute_func: Callable) -> list[CodeFlag]:
         """Runs the execute_func in find_mode to find flags"""
         self.codebase.set_find_mode(True)
         await self._execute_with_try_catch(execute_func, commit=False)
-        code_flags = self.codebase.G.flags._flags
-        logger.info(f"> Found {len(self.codebase.G.flags._flags)} CodeFlags")
+        code_flags = self.codebase.ctx.flags._flags
+        logger.info(f"> Found {len(self.codebase.ctx.flags._flags)} CodeFlags")
         return code_flags
 
     async def find_flag_groups(self, code_flags: list[CodeFlag], grouping_config: GroupingConfig) -> list[Group]:
@@ -77,7 +77,7 @@ async def execute_flag_groups(self, commit_msg: str, execute_func: Callable, fla
             run_results.append(run_result)
             head_branches.append(created_branch)
 
-        self.codebase.G.flags._flags.clear()
+        self.codebase.ctx.flags._flags.clear()
         return run_results, head_branches
 
     async def execute(self, execute_func: Callable, group: Group | None = None, session_options: SessionOptions = SessionOptions()) -> CodemodRunResult:
@@ -155,7 +155,7 @@ async def _get_structured_run_output(self, result: CodemodRunResult) -> CodemodR
                 "messageType": str(flag.message_type),
                 "messageRecipient": flag.message_recipient,
             }
-            for flag in self.codebase.G.flags._flags
+            for flag in self.codebase.ctx.flags._flags
         ]
         result.flags = flags
         if result.observation_meta is None:

src/codegen/runner/sandbox/middlewares.py

@@ -69,6 +69,6 @@ async def cleanup_after_codemod(self, is_exception: bool = False):
         if is_exception:
             # TODO: instead of committing transactions, we should just rollback
             logger.info("Committing pending transactions due to exception")
-            self.runner.codebase.G.commit_transactions(sync_graph=False)
+            self.runner.codebase.ctx.commit_transactions(sync_graph=False)
         self.runner.reset_runner()
         self.server_info.is_running_codemod = False

src/codegen/runner/sandbox/runner.py

@@ -112,5 +112,5 @@ async def create_branch(self, request: CreateBranchRequest) -> CreateBranchRespo
         response.results = run_results
         response.branches = branches
 
-        self.codebase.G.flags._flags.clear()
+        self.codebase.ctx.flags._flags.clear()
         return response

src/codegen/sdk/code_generation/doc_utils/generate_docs_json.py

@@ -8,7 +8,7 @@
 from codegen.sdk.core.placeholder.placeholder_type import TypePlaceholder
 
 ATTRIBUTES_TO_IGNORE = [
-    "G",
+    "ctx",
     "node_id",
     "angular",
     "model_config",
@@ -19,7 +19,6 @@
     "node_type",
     "ts_node",
     "file_node_id",
-    "G",
     "statement_type",
     "assignment_types",
 ]

src/codegen/sdk/codebase/codebase_analysis.py

@@ -9,7 +9,7 @@
 
 
 def get_codebase_summary(codebase: Codebase) -> str:
-    node_summary = f"""Contains {len(codebase.G.get_nodes())} nodes
+    node_summary = f"""Contains {len(codebase.ctx.get_nodes())} nodes
 - {len(list(codebase.files))} files
 - {len(list(codebase.imports))} imports
 - {len(list(codebase.external_modules))} external_modules
@@ -19,10 +19,10 @@ def get_codebase_summary(codebase: Codebase) -> str:
 \t- {len(list(codebase.global_vars))} global_vars
 \t- {len(list(codebase.interfaces))} interfaces
 """
-    edge_summary = f"""Contains {len(codebase.G.edges)} edges
-- {len([x for x in codebase.G.edges if x[2].type == EdgeType.SYMBOL_USAGE])} symbol -> used symbol
-- {len([x for x in codebase.G.edges if x[2].type == EdgeType.IMPORT_SYMBOL_RESOLUTION])} import -> used symbol
-- {len([x for x in codebase.G.edges if x[2].type == EdgeType.EXPORT])} export -> exported symbol
+    edge_summary = f"""Contains {len(codebase.ctx.edges)} edges
+- {len([x for x in codebase.ctx.edges if x[2].type == EdgeType.SYMBOL_USAGE])} symbol -> used symbol
+- {len([x for x in codebase.ctx.edges if x[2].type == EdgeType.IMPORT_SYMBOL_RESOLUTION])} import -> used symbol
+- {len([x for x in codebase.ctx.edges if x[2].type == EdgeType.EXPORT])} export -> exported symbol
     """
 
     return f"{node_summary}\n{edge_summary}"

src/codegen/sdk/codebase/codebase_graph.py renamed to src/codegen/sdk/codebase/codebase_context.py

@@ -76,7 +76,7 @@ def get_node_classes(programming_language: ProgrammingLanguage) -> NodeClasses:
         raise ValueError(msg)
 
 
-class CodebaseGraph:
+class CodebaseContext:
     """MultiDiGraph Wrapper with TransactionManager"""
 
     # =====[ __init__ attributes ]=====

src/codegen/sdk/codebase/config_parser.py

@@ -4,21 +4,21 @@
 from codegen.sdk.enums import ProgrammingLanguage
 
 if TYPE_CHECKING:
-    from codegen.sdk.codebase.codebase_graph import CodebaseGraph
+    from codegen.sdk.codebase.codebase_context import CodebaseContext
 
 
 class ConfigParser(ABC):
     def __init__(self):
         pass
 
     @abstractmethod
-    def parse_configs(self, codebase_graph: "CodebaseGraph"): ...
+    def parse_configs(self, codebase_context: "CodebaseContext"): ...
 
 
-def get_config_parser_for_language(language: ProgrammingLanguage, codebase_graph: "CodebaseGraph") -> ConfigParser | None:
+def get_config_parser_for_language(language: ProgrammingLanguage, codebase_context: "CodebaseContext") -> ConfigParser | None:
     from codegen.sdk.typescript.config_parser import TSConfigParser
 
     if language == ProgrammingLanguage.TYPESCRIPT:
-        return TSConfigParser(codebase_graph)
+        return TSConfigParser(codebase_context)
 
     return None

src/codegen/sdk/codebase/factory/get_session.py

@@ -6,7 +6,7 @@
 
 from codegen.git.repo_operator.local_repo_operator import LocalRepoOperator
 from codegen.git.schemas.repo_config import BaseRepoConfig
-from codegen.sdk.codebase.codebase_graph import CodebaseGraph
+from codegen.sdk.codebase.codebase_context import CodebaseContext
 from codegen.sdk.codebase.config import CodebaseConfig, GSFeatureFlags, ProjectConfig, SessionOptions, TestFlags
 from codegen.sdk.codebase.factory.codebase_factory import CodebaseFactory
 from codegen.sdk.core.codebase import Codebase, PyCodebaseType, TSCodebaseType
@@ -108,12 +108,12 @@ def get_codebase_graph_session(
     files: dict[str, str] = {},
     sync_graph: bool = True,
     session_options: SessionOptions = SessionOptions(),
-) -> Generator[CodebaseGraph, None, None]:
+) -> Generator[CodebaseContext, None, None]:
     """Gives you a Codebase2 operating on the files you provided as a dict"""
     op = LocalRepoOperator.create_from_files(repo_path=tmpdir, files=files)
     config = CodebaseConfig(feature_flags=TestFlags)
     projects = [ProjectConfig(repo_operator=op, programming_language=programming_language)]
-    graph = CodebaseGraph(projects=projects, config=config)
+    graph = CodebaseContext(projects=projects, config=config)
     with graph.session(sync_graph=sync_graph, session_options=session_options):
         try:
             yield graph

src/codegen/sdk/codebase/node_classes/py_node_classes.py

@@ -31,10 +31,10 @@
 from codegen.sdk.python.statements.import_statement import PyImportStatement
 
 
-def parse_subscript(node: TSNode, file_node_id, G, parent):
+def parse_subscript(node: TSNode, file_node_id, ctx, parent):
     if (node.prev_named_sibling and node.prev_named_sibling.text.decode("utf-8") == "TypeAlias") or isinstance(parent, Type):
-        return PyGenericType(node, file_node_id, G, parent)
-    return SubscriptExpression(node, file_node_id, G, parent)
+        return PyGenericType(node, file_node_id, ctx, parent)
+    return SubscriptExpression(node, file_node_id, ctx, parent)
 
 
 PyExpressionMap = {

src/codegen/sdk/codebase/range_index.py

@@ -43,7 +43,7 @@ def nodes(self) -> list[Editable]:
     def children(self) -> dict[Editable, list[Editable]]:
         ret = defaultdict(list)
         for node in self.nodes:
-            # if node.G.config.feature_flags.debug:
+            # if node.ctx.config.feature_flags.debug:
             #     assert node.parent != node, node.__class__
             if node.parent != node:
                 ret[node.parent].append(node)

src/codegen/sdk/codebase/transactions.py

@@ -262,7 +262,7 @@ def __init__(
         priority: int = 0,
     ) -> None:
         super().__init__(0, 0, file.path, priority=priority, new_content=new_file_path)
-        self.new_file_path = file.G.to_absolute(new_file_path)
+        self.new_file_path = file.ctx.to_absolute(new_file_path)
         self.file = file
 
     def execute(self) -> None:

src/codegen/sdk/codebase/validation.py

@@ -30,14 +30,14 @@ class PostInitValidationStatus(StrEnum):
 
 def post_init_validation(codebase: CodebaseType) -> PostInitValidationStatus:
     """Post codebase._init_graph verifies that the built graph is valid."""
-    from codegen.sdk.codebase.codebase_graph import GLOBAL_FILE_IGNORE_LIST
+    from codegen.sdk.codebase.codebase_context import GLOBAL_FILE_IGNORE_LIST
 
     # Verify the graph has nodes
-    if len(codebase.G.nodes) == 0:
+    if len(codebase.ctx.nodes) == 0:
         return PostInitValidationStatus.NO_NODES
 
     # Verify the graph has the same number of files as there are in the repo
-    if len(codebase.files) != len(list(codebase.op.iter_files(codebase.G.projects[0].subdirectories, extensions=codebase.G.extensions, ignore_list=GLOBAL_FILE_IGNORE_LIST))):
+    if len(codebase.files) != len(list(codebase.op.iter_files(codebase.ctx.projects[0].subdirectories, extensions=codebase.ctx.extensions, ignore_list=GLOBAL_FILE_IGNORE_LIST))):
         return PostInitValidationStatus.MISSING_FILES
 
     # Verify import resolution
@@ -64,14 +64,14 @@ def post_reset_validation(init_nodes, nodes, init_edges, edges, repo_name: str,
 
 def post_sync_validation(codebase: CodebaseType) -> bool:
     """Post codebase.sync, checks that the codebase graph is in a valid state (i.e. not corrupted by codebase.sync)"""
-    if len(codebase.G.all_syncs) > 0 or len(codebase.G.pending_syncs) > 0 or len(codebase.G.transaction_manager.to_commit()) > 0:
+    if len(codebase.ctx.all_syncs) > 0 or len(codebase.ctx.pending_syncs) > 0 or len(codebase.ctx.transaction_manager.to_commit()) > 0:
         msg = "Can only be called on a reset codebase"
         raise NotImplementedError(msg)
-    if not codebase.G.config.feature_flags.track_graph:
+    if not codebase.ctx.config.feature_flags.track_graph:
         msg = "Can only be called with track_graph=true"
         raise NotImplementedError(msg)
-    return len(dict.fromkeys(codebase.G.old_graph.nodes())) == len(dict.fromkeys(codebase.G.nodes)) and len(dict.fromkeys(codebase.G.old_graph.weighted_edge_list())) == len(
-        dict.fromkeys(codebase.G.edges)
+    return len(dict.fromkeys(codebase.ctx.old_graph.nodes())) == len(dict.fromkeys(codebase.ctx.nodes)) and len(dict.fromkeys(codebase.ctx.old_graph.weighted_edge_list())) == len(
+        dict.fromkeys(codebase.ctx.edges)
     )
 
 
@@ -140,7 +140,7 @@ def get_nodes_error(init_nodes, nodes):
         from codegen.sdk.core.external_module import ExternalModule
 
         if isinstance(node, ExternalModule):
-            message += "External Module persisted with following dependencies: " + str(list((node.G.get_node(source), edge) for source, _, edge in node.G.in_edges(node.node_id)))
+            message += "External Module persisted with following dependencies: " + str(list((node.ctx.get_node(source), edge) for source, _, edge in node.ctx.in_edges(node.node_id)))
     return message
 
 

src/codegen/sdk/core/assignment.py

@@ -27,7 +27,7 @@
 
     from tree_sitter import Node as TSNode
 
-    from codegen.sdk.codebase.codebase_graph import CodebaseGraph
+    from codegen.sdk.codebase.codebase_context import CodebaseContext
     from codegen.sdk.codebase.resolution_stack import ResolutionStack
     from codegen.sdk.core.expressions.type import Type
     from codegen.sdk.core.interfaces.editable import Editable
@@ -58,37 +58,37 @@ class Assignment(Symbol[Parent, ...], Typeable[Parent, ...], HasValue, Generic[P
     _left: Expression[Self]
     symbol_type = SymbolType.GlobalVar
 
-    def __init__(self, ts_node: TSNode, file_node_id: NodeId, G: CodebaseGraph, parent: Parent, left: TSNode, value: TSNode, name_node: TSNode, type: Type | None = None) -> None:
+    def __init__(self, ts_node: TSNode, file_node_id: NodeId, ctx: CodebaseContext, parent: Parent, left: TSNode, value: TSNode, name_node: TSNode, type: Type | None = None) -> None:
         self._unique_node = name_node  # HACK: This prevents deduplication of Assignments
-        super().__init__(ts_node, file_node_id, G, parent=parent, name_node=name_node, name_node_type=Name)
+        super().__init__(ts_node, file_node_id, ctx, parent=parent, name_node=name_node, name_node_type=Name)
         self._left = self._parse_expression(left, default=Name)
         self._value_node = self._parse_expression(value)
         self.type = type
         if self.type is None:
             self._init_type()
 
     @classmethod
-    def _from_left_and_right_nodes(cls, ts_node: TSNode, file_node_id: NodeId, G: CodebaseGraph, parent: Parent, left_node: TSNode, right_node: TSNode) -> list[Assignment]:
-        left = G.parser.parse_expression(left_node, file_node_id, G, parent)
-        value = G.parser.parse_expression(right_node, file_node_id, G, parent)
+    def _from_left_and_right_nodes(cls, ts_node: TSNode, file_node_id: NodeId, ctx: CodebaseContext, parent: Parent, left_node: TSNode, right_node: TSNode) -> list[Assignment]:
+        left = ctx.parser.parse_expression(left_node, file_node_id, ctx, parent)
+        value = ctx.parser.parse_expression(right_node, file_node_id, ctx, parent)
 
         if isinstance(left, Collection | Dict):
             assignments = []
             for var in left.symbols:
                 # Make a deep copy of the value expression for each child
-                value = G.parser.parse_expression(right_node, file_node_id, G, parent)
-                assignments.extend(cls._from_value_expression(ts_node, file_node_id, G, parent, left, value, var.ts_node))
+                value = ctx.parser.parse_expression(right_node, file_node_id, ctx, parent)
+                assignments.extend(cls._from_value_expression(ts_node, file_node_id, ctx, parent, left, value, var.ts_node))
             return sort_editables(assignments)
-        return cls._from_value_expression(ts_node, file_node_id, G, parent, left, value, left_node)
+        return cls._from_value_expression(ts_node, file_node_id, ctx, parent, left, value, left_node)
 
     @classmethod
     def _from_value_expression(
-        cls, ts_node: TSNode, file_node_id: NodeId, G: CodebaseGraph, parent: Parent, left: Expression[Self], value: Expression[Self] | list[Expression], name_node: TSNode
+        cls, ts_node: TSNode, file_node_id: NodeId, ctx: CodebaseContext, parent: Parent, left: Expression[Self], value: Expression[Self] | list[Expression], name_node: TSNode
     ) -> list[Assignment]:
-        assignments = [cls(ts_node, file_node_id, G, parent, left, value, name_node)]
+        assignments = [cls(ts_node, file_node_id, ctx, parent, left, value, name_node)]
         if value and isinstance(value, MultiExpression) and isinstance(value.expressions[0], Assignment):
             for expr in value.expressions:
-                assignments.extend(cls._from_value_expression(expr.ts_node, file_node_id, G, parent, expr.left, expr.value, expr.get_name().ts_node))
+                assignments.extend(cls._from_value_expression(expr.ts_node, file_node_id, ctx, parent, expr.left, expr.value, expr.get_name().ts_node))
         return sort_editables(assignments)
 
     @noapidoc

src/codegen/sdk/core/autocommit/decorators.py

@@ -47,7 +47,7 @@ def wrapper(wrapped: Callable[P, T], instance: "Editable", args, kwargs) -> T:
             instance = args[0]
         if instance.removed:
             logger.warning("Editing a removed node")
-        autocommit = instance.G._autocommit
+        autocommit = instance.ctx._autocommit
         logger.debug("Writing node %r,%r", instance, wrapped)
         with autocommit.write_state(instance, commit=commit):
             return wrapped(*args, **kwargs)
@@ -69,9 +69,9 @@ def remover(
     if instance is None:
         instance = args[0]
     logger.debug("Removing node %r, %r", instance, wrapped)
-    with instance.G._autocommit.write_state(instance):
+    with instance.ctx._autocommit.write_state(instance):
         ret = wrapped(*args, **kwargs)
-    # instance.G._autocommit.set_pending(instance, REMOVED)
+    # instance.ctx._autocommit.set_pending(instance, REMOVED)
     instance.removed = True
     return ret
 
@@ -86,7 +86,7 @@ def repr_func(
     """Indicates the method is use in debugging/logs."""
     if instance is None:
         instance = args[0]
-    autocommit = instance.G._autocommit
+    autocommit = instance.ctx._autocommit
     old_state = autocommit.enter_state(AutoCommitState.Special)
     try:
         ret = wrapped(*args, **kwargs)
@@ -108,8 +108,8 @@ def mover(
     """
     if instance is None:
         instance = args[0]
-    with instance.G._autocommit.write_state(instance, move=True):
+    with instance.ctx._autocommit.write_state(instance, move=True):
         file_node_id, node_id = wrapped(*args, **kwargs)
-    instance.G._autocommit.set_pending(instance, node_id, file_node_id)
+    instance.ctx._autocommit.set_pending(instance, node_id, file_node_id)
     instance.removed = False
     return None