Skip to content

armgit / burr

Public Imported from GitHub
0

Build applications that make decisions (chatbots, agents, simulations, etc...). Monitor, trace, persist, and execute on your own infrastructure.

Adds logging of attributes to the tracing factory

elijahbenizzy2139509e0267
diff --git a/burr/visibility/tracing.py b/burr/visibility/tracing.py
index 99c1e175..7e6f6ee0 100644
--- a/burr/visibility/tracing.py
+++ b/burr/visibility/tracing.py
@@ -209,11 +209,17 @@ class ActionSpanTracer(AbstractContextManager, AbstractAsyncContextManager):
self._sync_hooks_exit(prior_execution_context)
return None
- def log_artifact(self, name: str, value: Any):
- raise NotImplementedError(
- "Logging artifacts is not yet implemented. See "
- "(and comment on) the following issue for more details: "
- "https://github.com/DAGWorks-Inc/burr/issues/46"
+ def log_attribute(self, key: str, value: Any):
+ self.log_attributes(**{key: value})
+
+ def log_attributes(self, **attributes):
+ self.lifecycle_adapters.call_all_lifecycle_hooks_sync(
+ "do_log_attributes",
+ attributes=attributes,
+ action=self.action,
+ action_sequence_id=self.action_sequence_id,
+ span=self.context_var.get(),
+ tags={}, # TODO -- add tags
)
async def __aenter__(self) -> "ActionSpanTracer":
@@ -236,7 +242,7 @@ class ActionSpanTracer(AbstractContextManager, AbstractAsyncContextManager):
return None
-class TracerFactory:
+class TracerFactory(ActionSpanTracer):
"""Represents a tracer factory to create tracer instances. User never instantiates this
directly. Rather, this gets injected by the application. This gives a new span.
@@ -270,13 +276,17 @@ class TracerFactory:
:param lifecycle_adapters: Lifecycle adapters
:param _context_var: Context var to use for the action span
"""
- self.action = action
- self.lifecycle_adapters = lifecycle_adapters
- self.context_var = _context_var
- self.top_level_span_count = 0
- self.action_sequence_id = sequence_id
- self.app_id = app_id
- self.partition_key = partition_key
+ super(TracerFactory, self).__init__(
+ action=action,
+ action_sequence_id=sequence_id,
+ span_name="root",
+ lifecycle_adapters=lifecycle_adapters,
+ app_id=app_id,
+ partition_key=partition_key,
+ span_dependencies=[],
+ top_level_span_count=0,
+ context_var=_context_var,
+ )
def __call__(
self, span_name: str, span_dependencies: Optional[List[str]] = None