feat: Upgrade linters for ruff and mypy (#330)

Author: nachiket-galileo

.pre-commit-config.yaml

@@ -60,4 +60,8 @@ repos:
   #           "./pyproject.toml",
   # ]
 
-exclude: "CHANGELOG.md"
+exclude: |
+  (?x)^(
+    CHANGELOG\.md|
+    src/galileo/resources/.*
+  )$

examples/langgraph/basic_langgraph.py

@@ -27,14 +27,14 @@ class State(TypedDict):
     messages: Annotated[list, add_messages]
 
 
-def node(state: State):
+def node(state: State) -> dict:
     messages = state["messages"]
     new_message = AIMessage("Hello!")
 
-    return {"messages": messages + [new_message], "extra_field": 10}
+    return {"messages": [*messages, new_message], "extra_field": 10}
 
 
-def node2(state: State):
+def node2(state: State) -> dict:
     return {"messages": state["messages"]}
 
 

examples/langgraph/with_openai.py

@@ -28,7 +28,7 @@ class State(TypedDict):
 llm = ChatOpenAI(model="gpt-4")
 
 
-def chatbot(state: State):
+def chatbot(state: State) -> dict:
     return {"messages": [llm.invoke(state["messages"])]}
 
 

pyproject.toml

@@ -84,13 +84,113 @@ fix = true
 src = ["src"]
 unsafe-fixes = true
 target-version = "py39"
+exclude = [
+    "src/galileo/resources",
+]
 
 [tool.ruff.format]
 skip-magic-trailing-comma = true
 
 [tool.ruff.lint]
-select = ["E4", "E7", "E9", "F", "I", "UP", "ASYNC"]
-ignore = []
+select = [
+    # Pyflakes (basic errors)
+    "F",
+    # Pycodestyle (style errors)
+    "E4", "E7", "E9", "W6",
+    # Import sorting
+    "I",
+    # Pyupgrade (Python version upgrades)
+    "UP",
+    # Async-related checks
+    "ASYNC",
+    # Flake8-bugbear (common bugs)
+    "B",
+    # Flake8-comprehensions (list/set/dict comprehension issues)
+    "C4",
+    # Flake8-pie (unnecessary code patterns)
+    "PIE",
+    # Flake8-simplify (code simplification)
+    "SIM",
+    # Pylint-like rules (selective)
+    "PLC", "PLE", "PLW",
+    # Ruff-specific rules
+    "RUF",
+    # Flake8-bandit (security issues)
+    "S",
+    # Flake8-blind-except (bare except clauses)
+    "BLE",
+    # Flake8-boolean-trap (boolean trap antipattern)
+    "FBT",
+    # Flake8-unused-arguments
+    "ARG",
+    # Flake8-pytest-style
+    "PT",
+    # Flake8-return (return statement issues)
+    "RET",
+    # Flake8-implicit-str-concat
+    "ISC",
+    # Type checking related (important for bug catching)
+    "ANN001", "ANN201", "ANN202", "ANN205", "ANN206",
+]
+ignore = [
+    # Ignore overly strict rules (aligned with popular frameworks)
+    "S101",     # Use of assert (needed for tests)
+    "PLR0913",  # Too many arguments
+    "PLR2004",  # Magic value used in comparison
+    "B008",     # Do not perform function calls in argument defaults
+    "FBT001",   # Boolean positional arg in function definition
+    "FBT002",   # Boolean default arg in function definition
+    "S603",     # subprocess call: check for execution of untrusted input
+    "S607",     # Starting a process with a partial executable path
+    "PLR0911",  # Too many return statements
+    "PLR0912",  # Too many branches
+    "PLR0915",  # Too many statements
+    "PLW2901",  # Redefined loop variable
+    # Type annotation related (balance strictness)
+    "ANN101",   # Missing type annotation for `self` (unnecessary)
+    "ANN102",   # Missing type annotation for `cls` (unnecessary)
+    "ANN002",   # Missing type annotation for `*args` (impractical to enforce)
+    "ANN003",   # Missing type annotation for `**kwargs` (impractical to enforce)
+    "ANN401",   # Dynamically typed expressions (Any) are disallowed
+    # Unused arguments (often legitimate in wrappers, callbacks, etc.)
+    "ARG001",   # Unused function argument
+    "ARG002",   # Unused method argument
+    "ARG005",   # Unused lambda argument (common in callbacks)
+    # Exception handling (often too strict)
+    "BLE001",   # Do not catch blind exception (sometimes necessary)
+    "B904",     # Use raise ... from err (not always needed)
+    # Style preferences (overly pedantic)
+    "SIM102",   # Use single if statement (readability preference)
+    "PLC0206",  # Extracting value from dictionary without calling .items()
+    "PLW0127",  # Self-assignment of variable (sometimes needed)
+    # Additional common ignores from popular frameworks
+    "S311",     # Standard pseudo-random generators (false positive for backoff/timing)
+    "S324",     # Insecure hash functions (sometimes needed for non-crypto)
+    "PLR0914",  # Too many local variables
+]
+
+[tool.ruff.lint.per-file-ignores]
+# Ignore most linting rules for test files (focus on functionality, not style)
+"tests/**/*.py" = [
+    # Type annotations (not critical for tests)
+    "ANN",      # All annotation rules
+    # Complexity (tests can be complex)
+    "PLR",      # All pylint refactor rules (complexity, etc.)
+    # Security (tests often need assertions, subprocess, etc.)
+    "S",        # All bandit security rules
+    # Style preferences (less important in tests)
+    "FBT",      # Boolean trap rules
+    "ARG",      # Unused argument rules
+    "RET",      # Return statement rules
+    "SIM",      # Simplify rules
+    "C4",       # Comprehension rules
+    "PIE",      # Unnecessary code patterns
+    "ISC",      # Implicit string concatenation
+    # Common test patterns
+    "B008",     # Function calls in argument defaults
+    "B017",     # Do not assert blind exception (needed for pytest.raises)
+    "PT",       # Pytest style rules (can be overly strict)
+]
 
 [tool.ruff.lint.isort]
 known-first-party = ["galileo_core"]
@@ -108,12 +208,41 @@ wrap-descriptions = 120
 
 [tool.mypy]
 mypy_path = ["src"]
+# Type checking strictness (balanced for production use)
 disallow_untyped_defs = true
-disable_error_code = ["import-untyped"]
+disallow_incomplete_defs = true
+disallow_untyped_decorators = false  # Often problematic with third-party decorators
+disallow_untyped_calls = false       # Set to false to avoid issues with external libraries
+# Error detection
+warn_redundant_casts = true
+warn_unused_ignores = true
+warn_return_any = false     # Too noisy with external APIs
+warn_unreachable = true
+# Import handling
 ignore_missing_imports = true
-no_implicit_optional = false
 follow_imports = "skip"
+# Optional handling
+no_implicit_optional = true  # Good practice for modern Python
+strict_optional = true
+# Error codes (balanced approach)
+disable_error_code = [
+    "import-untyped",
+    "misc",  # Disable misc warnings (like untyped decorators we can't control)
+]
+enable_error_code = ["truthy-bool", "redundant-expr", "unused-awaitable"]
+# Plugins
 plugins = ["pydantic.mypy"]
+# Additional strictness (keep what catches real bugs)
+check_untyped_defs = true
+strict_equality = true
+extra_checks = true
+# Performance and compatibility
+show_error_codes = true
+pretty = true
+show_column_numbers = true
+# Incremental mode for better performance on individual files
+incremental = true
+sqlite_cache = true
 
 # Release.
 

scripts/convert-md-to-mdx.py

@@ -7,7 +7,7 @@
 from pathlib import Path
 
 
-def process_markdown_files():
+def process_markdown_files() -> None:
     """
     Process all .md files in .generated_docs/reference folder and subdirectories:
     - Remove sidebar_label: line

scripts/lint.sh

@@ -1,4 +1,6 @@
 #!/bin/sh -ex
 
-pre-commit run ruff --all-files
-pre-commit run mypy --all-files
+echo "🔍 Running lint checks (excluding auto-generated files)..."
+poetry run pre-commit run ruff --all-files
+poetry run pre-commit run mypy --all-files
+echo "✅ All lint checks completed successfully!"

src/galileo/config.py

@@ -1,6 +1,6 @@
 # mypy: disable-error-code=syntax
 # We need to ignore syntax errors until https://github.com/python/mypy/issues/17535 is resolved.
-from typing import Any, Optional
+from typing import Any, ClassVar, Optional
 
 from pydantic_core import Url
 
@@ -12,18 +12,14 @@ class GalileoPythonConfig(GalileoConfig):
     config_filename: str = "galileo-python-config.json"
     console_url: Url = "https://app.galileo.ai"
 
-    def reset(self) -> None:
-        global _galileo_config
-        _galileo_config = None
+    _instance: ClassVar[Optional["GalileoPythonConfig"]] = None
 
+    def reset(self) -> None:
+        GalileoPythonConfig._instance = None
         super().reset()
 
     @classmethod
     def get(cls, **kwargs: Any) -> "GalileoPythonConfig":
-        global _galileo_config
-        _galileo_config = cls._get(_galileo_config, **kwargs)  # type: ignore[arg-type]
-        assert _galileo_config is not None, "Failed to initialize GalileoPythonConfig"
-        return _galileo_config
-
-
-_galileo_config: Optional[GalileoPythonConfig] = None
+        cls._instance = cls._get(cls._instance, **kwargs)
+        assert cls._instance is not None, "Failed to initialize GalileoPythonConfig"
+        return cls._instance

src/galileo/constants/__init__.py

@@ -3,4 +3,4 @@
 
 DEFAULT_API_URL = "https://api.galileo.ai/"
 
-__all__ = ("DEFAULT_PROJECT_NAME", "DEFAULT_LOG_STREAM_NAME", "DEFAULT_API_URL")
+__all__ = ("DEFAULT_API_URL", "DEFAULT_LOG_STREAM_NAME", "DEFAULT_PROJECT_NAME")

src/galileo/constants/routes.py

@@ -5,7 +5,7 @@ class Routes(str, Enum):
     healthcheck = "healthcheck"
     login = "login"
     api_key_login = "login/api_key"
-    get_token = "get-token"
+    get_token = "get-token"  # noqa: S105  # This is a URL path, not a password
 
     projects = "projects"
     all_projects = "projects/all"

src/galileo/datasets.py

@@ -137,10 +137,9 @@ def add_rows(self, row_data: list[dict[str, Any]]) -> "Dataset":
         return self
 
     def get_version_history(self) -> Optional[Union[HTTPValidationError, ListDatasetVersionResponse]]:
-        list_dataset = query_dataset_versions_datasets_dataset_id_versions_query_post.sync(
+        return query_dataset_versions_datasets_dataset_id_versions_query_post.sync(
             dataset_id=self.dataset.id, client=self.config.api_client, body=ListDatasetVersionParams()
         )
-        return list_dataset
 
     def load_version(self, version_index: int) -> DatasetContent:
         return get_dataset_version_content_datasets_dataset_id_versions_version_index_content_get.sync(
@@ -557,7 +556,7 @@ def create_dataset(name: str, content: DatasetType) -> Dataset:
 
 
 def get_dataset_version_history(
-    *, dataset_name: str = None, dataset_id: str = None
+    *, dataset_name: Optional[str] = None, dataset_id: Optional[str] = None
 ) -> Optional[Union[HTTPValidationError, ListDatasetVersionResponse]]:
     """
     Retrieves a dataset version history by dataset name or dataset id.
@@ -582,17 +581,16 @@ def get_dataset_version_history(
         if dataset is None:
             raise ValueError(f"Dataset '{dataset_name}' not found")
         return dataset.get_version_history()
-    elif dataset_id is not None:
+    if dataset_id is not None:
         dataset = Datasets().get(id=dataset_id)
         if dataset is None:
             raise ValueError(f"Dataset '{dataset_id}' not found")
         return dataset.get_version_history()
-    else:
-        raise ValueError("Either dataset_name or dataset_id must be provided.")
+    raise ValueError("Either dataset_name or dataset_id must be provided.")
 
 
 def get_dataset_version(
-    *, version_index: int, dataset_name: str = None, dataset_id: str = None
+    *, version_index: int, dataset_name: Optional[str] = None, dataset_id: Optional[str] = None
 ) -> Optional[DatasetContent]:
     """
     Retrieves a dataset version by dataset name or dataset id.
@@ -618,13 +616,12 @@ def get_dataset_version(
             raise ValueError(f"Dataset '{dataset_name}' not found")
         return dataset.load_version(version_index)
 
-    elif dataset_id is not None:
+    if dataset_id is not None:
         dataset = Datasets().get(id=dataset_id)
         if dataset is None:
             raise ValueError(f"Dataset '{dataset_id}' not found")
         return dataset.load_version(version_index)
-    else:
-        raise ValueError("Either dataset_name or dataset_id must be provided.")
+    raise ValueError("Either dataset_name or dataset_id must be provided.")
 
 
 def extend_dataset(
@@ -706,6 +703,6 @@ def convert_dataset_row_to_record(dataset_row: DatasetRow) -> "DatasetRecord":
     return DatasetRecord(
         id=dataset_row.row_id,
         input=values_dict["input"],
-        output=values_dict["output"] if "output" in values_dict else None,
-        metadata=values_dict["metadata"] if "metadata" in values_dict else None,
+        output=values_dict.get("output", None),
+        metadata=values_dict.get("metadata", None),
     )