Cherry-pick: feat: better logger stream allocation (codegen-sh#1006)

# 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

Author: Zeeeepa

codegen-examples/uv.lock

@@ -4,6 +4,41 @@ icon: "clock"
 iconType: "solid"
 ---
 
+<Update label="v0.52.31" description="April 02, 2025">
+### [Updates dependency sentry-sdk.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.31)
+- Update dependency sentry-sdk to v2.25.1
+</Update>
+
+<Update label="v0.52.30" description="April 01, 2025">
+### [Improves debugging with git init failure logs.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.30)
+- Add logs for git initialization failure
+</Update>
+
+<Update label="v0.52.29" description="April 01, 2025">
+### [Fixes codebase initialization issue.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.29)
+- Fix issue with codebase initialization
+</Update>
+
+<Update label="v0.52.28" description="March 31, 2025">
+### [Updates OpenAI dependency.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.28)
+- Update OpenAI dependency to v1.70.0
+</Update>
+
+<Update label="v0.52.27" description="March 31, 2025">
+### [Upgrade sentry-sdk dependency.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.27)
+- Upgrade sentry-sdk dependency for better bug tracking
+</Update>
+
+<Update label="v0.52.26" description="March 27, 2025">
+### [Updates openai dependency to the latest version.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.26)
+- Update dependency openai to version 1.69.0
+</Update>
+
+<Update label="v0.52.25" description="March 26, 2025">
+### [remove unnecessary transaction in build_graph.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.25)
+- Remove build_graph transaction
+</Update>
+
 <Update label="v0.52.24" description="March 25, 2025">
 ### [Documentation update for `codegen create` command.](https://github.com/codegen-sh/codegen-sdk/releases/tag/v0.52.24)
 - Update documentation for `codegen create` command

docs/changelog/changelog.mdx

@@ -6,7 +6,7 @@ readme = "README.md"
 # renovate: datasource=python-version depName=python
 requires-python = ">=3.12, <3.14"
 dependencies = [
-  "openai==1.68.2",
+  "openai==1.70.0",
   "tiktoken<1.0.0,>=0.5.1",
   "tabulate>=0.9.0,<1.0.0",
   "codeowners<1.0.0,>=0.6.0",
@@ -35,7 +35,7 @@ dependencies = [
   "rustworkx>=0.15.1",
   "typing-extensions>=4.12.2",
   "termcolor>=2.4.0",
-  "sentry-sdk==2.24.1",
+  "sentry-sdk==2.25.1",
   "click>=8.1.7",
   "requests>=2.32.3",
   "lazy-object-proxy>=0.0.0",
@@ -127,7 +127,7 @@ dev-dependencies = [
   #    "scalene>=1.5.45",
   "filelock<4.0.0,>=3.15.4",
   "pytest>=8.3.3",
-  "pytest-cov>=6.0.0,<6.0.1",
+  "pytest-cov>=6.1.0,<6.1.1",
   "ruff>=0.6.8",
   "mypy[mypyc,faster-cache]>=1.13.0",
   "pre-commit>=4.0.1",

pyproject.toml

@@ -897,8 +897,7 @@ def create_from_repo(cls, repo_path: str, url: str, access_token: str | None = N
                     if local_head.hexsha == remote_head.hexsha:
                         return cls(repo_config=RepoConfig.from_repo_path(repo_path), bot_commit=False, access_token=access_token)
             except Exception:
-                # If any git operations fail, fallback to fresh clone
-                pass
+                logger.exception("Failed to initialize Git repository. Falling back to fresh clone.")
 
             # If we get here, repo exists but is not up to date or valid
             # Remove the existing directory to do a fresh clone
@@ -916,7 +915,6 @@ def create_from_repo(cls, repo_path: str, url: str, access_token: str | None = N
             # Initialize with the cloned repo
             git_cli = GitCLI(repo_path)
         except (GitCommandError, ValueError) as e:
-            logger.exception("Failed to initialize Git repository:")
-            logger.exception("Please authenticate with a valid token and ensure the repository is properly initialized.")
+            logger.exception("Failed to initialize Git repository")
             return None
         return cls(repo_config=RepoConfig.from_repo_path(repo_path), bot_commit=False, access_token=access_token)

src/codegen/git/repo_operator/repo_operator.py

@@ -25,7 +25,7 @@
 from codegen.configs.models.codebase import CodebaseConfig, PinkMode
 from codegen.configs.models.secrets import SecretsConfig
 from codegen.git.repo_operator.repo_operator import RepoOperator
-from codegen.git.schemas.enums import CheckoutResult, SetupOption
+from codegen.git.schemas.enums import CheckoutResult
 from codegen.git.schemas.repo_config import RepoConfig
 from codegen.git.utils.pr_review import CodegenPR
 from codegen.sdk._proxy import proxy_property
@@ -1337,7 +1337,6 @@ def from_repo(
         language: Literal["python", "typescript"] | ProgrammingLanguage | None = None,
         config: CodebaseConfig | None = None,
         secrets: SecretsConfig | None = None,
-        setup_option: SetupOption | None = None,
         full_history: bool = False,
     ) -> "Codebase":
         """Fetches a codebase from GitHub and returns a Codebase instance.
@@ -1382,6 +1381,11 @@ def from_repo(
             else:
                 # Ensure the operator can handle remote operations
                 repo_operator = RepoOperator.create_from_commit(repo_path=repo_path, commit=commit, url=repo_url, full_name=repo_full_name, access_token=access_token)
+
+            if repo_operator is None:
+                logger.error("Failed to clone repository")
+                return None
+
             logger.info("Clone completed successfully")
 
             # Initialize and return codebase with proper context

src/codegen/sdk/core/codebase.py

@@ -1,41 +1,76 @@
 import logging
+import sys
 
 import colorlog
 
-
-def get_logger(name: str, level: int = logging.INFO) -> logging.Logger:
-    # Force configure the root logger with a NullHandler to prevent duplicate logs
-    logging.basicConfig(handlers=[logging.NullHandler()], force=True)
-
-    formatter = colorlog.ColoredFormatter(
-        "%(white)s%(asctime)s - %(name)s - %(log_color)s%(levelname)s%(reset)s%(white)s - %(message_log_color)s%(message)s",
-        log_colors={
+formatter = colorlog.ColoredFormatter(
+    "%(white)s%(asctime)s - %(name)s - %(log_color)s%(levelname)s%(reset)s%(white)s - %(message_log_color)s%(message)s",
+    log_colors={
+        "DEBUG": "white",
+        "INFO": "green",
+        "WARNING": "yellow",
+        "ERROR": "red",
+        "CRITICAL": "red,bg_white",
+    },
+    secondary_log_colors={
+        "message": {
             "DEBUG": "cyan",
-            "INFO": "green",
+            "INFO": "white",
             "WARNING": "yellow",
             "ERROR": "red",
             "CRITICAL": "red,bg_white",
-        },
-        secondary_log_colors={
-            "message": {
-                "DEBUG": "cyan",
-                "INFO": "blue",
-                "WARNING": "yellow",
-                "ERROR": "red",
-                "CRITICAL": "red,bg_white",
-            }
-        },
-    )
+        }
+    },
+)
+
+
+class StdOutFilter(logging.Filter):
+    def filter(self, record):
+        return record.levelno < logging.ERROR
+
+
+class StdErrFilter(logging.Filter):
+    def filter(self, record):
+        return record.levelno >= logging.ERROR
+
+
+# Create handlers
+stdout_handler = logging.StreamHandler(sys.stdout)  # Logs to stdout
+stdout_handler.setFormatter(formatter)
+stdout_handler.addFilter(StdOutFilter())
+
+stderr_handler = logging.StreamHandler(sys.stderr)  # Logs to stderr
+stderr_handler.setFormatter(formatter)
+stderr_handler.addFilter(StdErrFilter())
+
+
+def get_logger(name: str, level: int = logging.INFO) -> logging.Logger:
+    logger = _setup_logger(name, level)
+    _setup_exception_logging(logger)
+    return logger
+
+
+def _setup_logger(name: str, level: int = logging.INFO) -> logging.Logger:
+    # Force configure the root logger with a NullHandler to prevent duplicate logs
+    logging.basicConfig(handlers=[logging.NullHandler()], force=True)
     logger = logging.getLogger(name)
     if logger.hasHandlers():
         for h in logger.handlers:
             logger.removeHandler(h)
 
-    handler = colorlog.StreamHandler()
-    handler.setFormatter(formatter)
-    logger.addHandler(handler)
+    logger.addHandler(stdout_handler)
+    logger.addHandler(stderr_handler)
+
     # Ensure the logger propagates to the root logger
     logger.propagate = True
     # Set the level on the logger itself
     logger.setLevel(level)
     return logger
+
+
+def _setup_exception_logging(logger: logging.Logger) -> None:
+    def log_exception(exc_type, exc_value, exc_traceback):
+        logger.exception("Uncaught exception", exc_info=(exc_type, exc_value, exc_traceback))
+
+    # Set the log_exception function as the exception hook
+    sys.excepthook = log_exception

src/codegen/shared/logging/get_logger.py

@@ -2,12 +2,11 @@
 
 import pytest
 
-from codegen.git.schemas.enums import SetupOption
 from codegen.sdk.core.codebase import Codebase
 
 
 @pytest.fixture
 def codebase(tmpdir):
     os.chdir(tmpdir)
-    codebase = Codebase.from_repo(repo_full_name="codegen-sh/Kevin-s-Adventure-Game", tmp_dir=tmpdir, language="python", setup_option=SetupOption.PULL_OR_CLONE)
+    codebase = Codebase.from_repo(repo_full_name="codegen-sh/Kevin-s-Adventure-Game", tmp_dir=tmpdir, language="python")
     yield codebase

tests/integration/codegen/git/codebase/conftest.py

@@ -2,8 +2,10 @@
 
 import pytest
 
+from codegen.configs.models.secrets import SecretsConfig
 from codegen.sdk.codebase.codebase_context import CodebaseContext
 from codegen.sdk.codebase.factory.get_session import get_codebase_session
+from codegen.sdk.core.codebase import Codebase
 
 
 @pytest.fixture(autouse=True)
@@ -39,3 +41,9 @@ def test_codeowners_property(context_mock, codebase):
     assert len(codebase.codeowners) == 1
     assert callable(codebase.codeowners[0].files_source)
     assert codebase.codeowners[0].files_source() == codebase.files.return_value
+
+
+def test_from_codebase_non_existent_repo(context_mock, tmpdir):
+    with get_codebase_session(tmpdir=tmpdir, files={"src/main.py": "print('Hello, world!')"}, verify_output=False) as codebase:
+        codebase = Codebase.from_repo("some-org/non-existent-repo", tmp_dir=tmpdir, secrets=SecretsConfig(github_token="some-token"))
+        assert codebase is None