From cfb3941cf318e3943552af93fcd3794b60c24be5 Mon Sep 17 00:00:00 2001
From: stroblme <stroblme@posteo.de>
Date: Sat, 29 Aug 2026 13:49:25 +0200
Subject: [PATCH] Give a node's fluksio module the logger the SDK exports
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Inside a worker `import fluksio` is the reporter, which had emit and the
artifact calls but no logger — so `fluksio.logger.info(...)`, written
against the SDK's top-level export, died with AttributeError after the
training it was reporting on had already succeeded. Its records go to the
same capture a print does; the handler resolves sys.stderr per record
because a call runs under redirect_stderr.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019Hra4ndWMCLU5F3KjUuVAc
---
 backend/tests/flow/test_workers.py   | 18 ++++++++++++++++++
 worker/fluksio_worker/worker_main.py | 24 ++++++++++++++++++++++++
 2 files changed, 42 insertions(+)

diff --git a/backend/tests/flow/test_workers.py b/backend/tests/flow/test_workers.py
index 4d8c9a0..a2be684 100644
--- a/backend/tests/flow/test_workers.py
+++ b/backend/tests/flow/test_workers.py
@@ -69,6 +69,24 @@ def test_a_node_returns_its_value_and_what_it_printed(pool, capsys):
     assert "seen 21" in capsys.readouterr().out
 
 
+def test_a_node_can_log_through_the_module_it_imports(pool, capsys):
+    """The SDK exports `logger`; inside a worker `fluksio` is the reporter.
+
+    Without it, `fluksio.logger.info(...)` died with AttributeError — after
+    the training it was reporting on had already succeeded.
+    """
+    result = run(
+        pool,
+        "import fluksio\n\n\n"
+        "def process(value):\n"
+        "    fluksio.logger.info('tuned %s', value)\n"
+        "    return {'out': value}\n",
+        value=7,
+    )
+    assert result == {"out": 7}
+    assert "tuned 7" in capsys.readouterr().out
+
+
 def test_a_failure_keeps_its_class_and_points_at_the_node(pool):
     with pytest.raises(Exception) as caught:
         run(pool, "def process():\n    raise ValueError('bad input')\n")
diff --git a/worker/fluksio_worker/worker_main.py b/worker/fluksio_worker/worker_main.py
index 9d7c604..5ba0880 100644
--- a/worker/fluksio_worker/worker_main.py
+++ b/worker/fluksio_worker/worker_main.py
@@ -45,6 +45,7 @@ import hashlib
 import inspect
 import io
 import json
+import logging
 import tempfile
 import time
 import traceback
@@ -150,6 +151,18 @@ class _Reporter(ModuleType):
         return _Declared()
 
 
+class _LogHandler(logging.Handler):
+    """Writes to whatever ``sys.stderr`` is at the time of the record.
+
+    A call runs under ``redirect_stderr`` into the buffer that becomes the
+    node's logs, so a handler holding the stream it was built with would write
+    past every one of them.
+    """
+
+    def emit(self, record: logging.LogRecord) -> None:
+        sys.stderr.write(self.format(record) + "\n")
+
+
 class _Declared:
     """Stands in for a declaration whose work was done before the run.
 
@@ -294,6 +307,17 @@ def _install_reporter() -> None:
     """Put ``fluksio`` on the import path of every node this worker runs."""
     module = _Reporter("fluksio")
     module.__doc__ = "Report metrics and progress from inside a node."
+    # The SDK exports a logger at top level, so a node written against it says
+    # `fluksio.logger.info(...)` — which used to die with AttributeError, after
+    # whatever it was reporting on had already succeeded. Its records land in
+    # the same capture as a print.
+    logger = logging.getLogger("fluksio")
+    logger.setLevel(logging.INFO)
+    logger.addHandler(_LogHandler())
+    # Nothing above it should see these twice: a node calling `basicConfig` for
+    # its own logging installs a root handler on the same stream.
+    logger.propagate = False
+    module.logger = logger
     sys.modules["fluksio"] = module