Add allow_external flag & Fix Tests (#807)

# Motivation

<!-- Why is this change necessary? -->

# Content

<!-- Please include a summary of the change -->

# Testing

<!-- How was the change tested? -->

# Please check the following before marking your PR as ready for review

- [ ] I have added tests for my changes
- [ ] I have updated the documentation or added new documentation as
needed

---------

Co-authored-by: Edo Pujol <[email protected]>
Co-authored-by: kopekC <[email protected]>
Co-authored-by: tomcodgen <[email protected]>
Co-authored-by: tomcodgen <[email protected]>
Co-authored-by: tomcodegen <[email protected]>

Author: 5 people

docs/introduction/advanced-settings.mdx

@@ -309,6 +309,19 @@ Controls import path overrides during import resolution.
 
 Enables and disables resolution of imports from `sys.path`.
 
+<Warning>
+For this to properly work, you must also set `allow_external` to `True`.
+</Warning>
+
+## Flag: `allow_external`
+> **Default: `False`**
+
+Enables resolving imports, files, modules, and directories from outside of the repo path.
+
+<Warning>
+Turning this flag off may allow for bad actors to access files outside of the repo path! Use with caution!
+</Warning>
+
 ## Flag: `ts_dependency_manager`
 > **Default: `False`**
 

src/codegen/configs/models/codebase.py

@@ -21,6 +21,7 @@ def __init__(self, prefix: str = "CODEBASE", *args, **kwargs) -> None:
     import_resolution_paths: list[str] = Field(default_factory=lambda: [])
     import_resolution_overrides: dict[str, str] = Field(default_factory=lambda: {})
     py_resolve_syspath: bool = False
+    allow_external: bool = False
     ts_dependency_manager: bool = False
     ts_language_engine: bool = False
     v8_ts_engine: bool = False

src/codegen/sdk/codebase/codebase_context.py

@@ -381,7 +381,7 @@ def get_directory(self, directory_path: PathLike, create_on_missing: bool = Fals
         """
         # If not part of repo path, return None
         absolute_path = self.to_absolute(directory_path)
-        if not self.is_subdir(absolute_path):
+        if not self.is_subdir(absolute_path) and not self.config.allow_external:
             assert False, f"Directory {absolute_path} is not part of repo path {self.repo_path}"
             return None
 
@@ -611,7 +611,7 @@ def get_edges(self) -> list[tuple[NodeId, NodeId, EdgeType, Usage | None]]:
     def get_file(self, file_path: os.PathLike, ignore_case: bool = False) -> SourceFile | None:
         # If not part of repo path, return None
         absolute_path = self.to_absolute(file_path)
-        if not self.is_subdir(absolute_path):
+        if not self.is_subdir(absolute_path) and not self.config.allow_external:
             assert False, f"File {file_path} is not part of the repository path"
 
         # Check if file exists in graph

src/codegen/sdk/core/codebase.py

@@ -213,6 +213,13 @@ def __init__(
         self.ctx = CodebaseContext(projects, config=config, secrets=secrets, io=io, progress=progress)
         self.console = Console(record=True, soft_wrap=True)
 
+        # Assert config assertions
+        # External import resolution must be enabled if syspath is enabled
+        if self.ctx.config.py_resolve_syspath:
+            if not self.ctx.config.allow_external:
+                msg = "allow_external must be set to True when py_resolve_syspath is enabled"
+                raise ValueError(msg)
+
     @noapidoc
     def __str__(self) -> str:
         return f"<Codebase(name={self.name}, language={self.language}, path={self.repo_path})>"

tests/unit/codegen/sdk/python/import_resolution/test_import_resolution.py

@@ -277,6 +277,8 @@ def func():
 
         # Enable resolution via sys.path
         codebase.ctx.config.py_resolve_syspath = True
+        # Allow resolving files and modules outside of the repo path
+        codebase.ctx.config.allow_external = True
 
         # =====[ Imports cannot be found without sys.path being set ]=====
         assert len(consumer_file.imports) == 1
@@ -372,6 +374,8 @@ def func():
         # Ensure we don't have overrites and enable syspath resolution
         codebase.ctx.config.import_resolution_paths = []
         codebase.ctx.config.py_resolve_syspath = True
+        # Allow resolving files and modules outside of the repo path
+        codebase.ctx.config.allow_external = True
 
         # =====[ Import with sys.path set can be found ]=====
         assert len(consumer_file.imports) == 1
@@ -419,6 +423,8 @@ def func():
         # Ensure we don't have overrites and enable syspath resolution
         codebase.ctx.config.import_resolution_paths = []
         codebase.ctx.config.py_resolve_syspath = True
+        # Allow resolving files and modules outside of the repo path
+        codebase.ctx.config.allow_external = True
 
         # =====[ Default import works ]=====
         assert len(consumer_file.imports) == 1