Merge branch 'master' into dependabot/github_actions/getsentry/codecov-action-afa60378e8e4939ddb77575a51da92885c61ca86

Author: ericapisani

.github/workflows/ai-integration-test.yml

@@ -34,7 +34,7 @@ jobs:
           token: ${{ secrets.GITHUB_TOKEN }}
 
       - name: Run Python SDK Tests
-        uses: getsentry/testing-ai-sdk-integrations@6b1f51ec8af03e19087df452b426aa7e46d2b20a
+        uses: getsentry/testing-ai-sdk-integrations@1dd9ee2a2d821f41473fc90809a758e8394c689c
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:

.github/workflows/ci.yml

@@ -54,7 +54,7 @@ jobs:
           # This will also trigger "make dist" that creates the Python packages
           make aws-lambda-layer
       - name: Upload Python Packages
-        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7
+        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7
         with:
           name: artifact-build_lambda_layer
           path: |
@@ -76,7 +76,7 @@ jobs:
           make apidocs
           cd docs/_build && zip -r gh-pages ./
 
-      - uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7
+      - uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7
         with:
           name: artifact-docs
           path: |

.github/workflows/release.yml

@@ -22,7 +22,7 @@ jobs:
     steps:
     - name: Get auth token
       id: token
-      uses: actions/create-github-app-token@f8d387b68d61c58ab83c6c016672934102569859 # v2
+      uses: actions/create-github-app-token@1b10c78c7865c340bc4f6099eb2f838309f1e8c3 # v2
       with:
         app-id: ${{ vars.SENTRY_RELEASE_BOT_CLIENT_ID }}
         private-key: ${{ secrets.SENTRY_RELEASE_BOT_PRIVATE_KEY }}

scripts/populate_tox/package_dependencies.jsonl

@@ -91,13 +91,21 @@ def add(self, span: "StreamedSpan") -> None:
     @staticmethod
     def _estimate_size(item: "StreamedSpan") -> int:
         # Rough estimate of serialized span size that's quick to compute.
-        # 210 is the rough size of the payload without attributes, and we
-        # estimate additional 70 bytes on top of that per attribute.
-        return 210 + 70 * len(item._attributes)
+        # 210 is the rough size of the payload without attributes, and then we
+        # estimate the attributes separately.
+        estimate = 210
+        for value in item._attributes.values():
+            estimate += 50
+
+            if isinstance(value, str):
+                estimate += len(value)
+            else:
+                estimate += len(str(value))
+
+        return estimate
 
     @staticmethod
     def _to_transport_format(item: "StreamedSpan") -> "Any":
-        # TODO[span-first]
         res: "dict[str, Any]" = {
             "trace_id": item.trace_id,
             "span_id": item.span_id,
@@ -126,7 +134,7 @@ def _flush(self) -> None:
                 return
 
             envelopes = []
-            for trace_id, spans in self._span_buffer.items():
+            for spans in self._span_buffer.values():
                 if spans:
                     dsc = spans[0]._dynamic_sampling_context()
 

scripts/populate_tox/releases.jsonl

@@ -105,3 +105,33 @@ def _get_request_data(asgi_scope: "Any") -> "Dict[str, Any]":
         request_data["env"] = {"REMOTE_ADDR": _get_ip(asgi_scope)}
 
     return request_data
+
+
+def _get_request_attributes(asgi_scope: "Any") -> "dict[str, Any]":
+    """
+    Return attributes related to the HTTP request from the ASGI scope.
+    """
+    attributes: "dict[str, Any]" = {}
+
+    ty = asgi_scope["type"]
+    if ty in ("http", "websocket"):
+        if asgi_scope.get("method"):
+            attributes["http.request.method"] = asgi_scope["method"].upper()
+
+        headers = _filter_headers(_get_headers(asgi_scope), use_annotated_value=False)
+        for header, value in headers.items():
+            attributes[f"http.request.header.{header.lower()}"] = value
+
+        query = _get_query(asgi_scope)
+        if query:
+            attributes["http.query"] = query
+
+        attributes["url.full"] = _get_url(
+            asgi_scope, "http" if ty == "http" else "ws", headers.get("host")
+        )
+
+    client = asgi_scope.get("client")
+    if client and should_send_default_pii():
+        attributes["client.address"] = _get_ip(asgi_scope)
+
+    return attributes

sentry_sdk/_span_batcher.py

@@ -3,6 +3,7 @@
 from copy import deepcopy
 
 import sentry_sdk
+from sentry_sdk._types import SENSITIVE_DATA_SUBSTITUTE
 from sentry_sdk.scope import should_send_default_pii
 from sentry_sdk.utils import AnnotatedValue, logger
 
@@ -211,16 +212,19 @@ def _is_json_content_type(ct: "Optional[str]") -> bool:
 
 def _filter_headers(
     headers: "Mapping[str, str]",
+    use_annotated_value: bool = True,
 ) -> "Mapping[str, Union[AnnotatedValue, str]]":
     if should_send_default_pii():
         return headers
 
+    substitute: "Union[AnnotatedValue, str]"
+    if use_annotated_value:
+        substitute = AnnotatedValue.removed_because_over_size_limit()
+    else:
+        substitute = SENSITIVE_DATA_SUBSTITUTE
+
     return {
-        k: (
-            v
-            if k.upper().replace("-", "_") not in SENSITIVE_HEADERS
-            else AnnotatedValue.removed_because_over_size_limit()
-        )
+        k: (v if k.upper().replace("-", "_") not in SENSITIVE_HEADERS else substitute)
         for k, v in headers.items()
     }
 

sentry_sdk/integrations/_asgi_common.py

@@ -15,6 +15,7 @@
 from sentry_sdk.consts import OP
 from sentry_sdk.integrations._asgi_common import (
     _get_headers,
+    _get_request_attributes,
     _get_request_data,
     _get_url,
 )
@@ -23,7 +24,11 @@
     nullcontext,
 )
 from sentry_sdk.sessions import track_session
-from sentry_sdk.traces import StreamedSpan
+from sentry_sdk.traces import (
+    StreamedSpan,
+    SegmentSource,
+    SOURCE_FOR_STYLE as SEGMENT_SOURCE_FOR_STYLE,
+)
 from sentry_sdk.tracing import (
     SOURCE_FOR_STYLE,
     Transaction,
@@ -40,6 +45,7 @@
     _get_installed_modules,
     reraise,
     capture_internal_exceptions,
+    qualname_from_function,
 )
 
 from typing import TYPE_CHECKING
@@ -235,7 +241,7 @@ async def _run_app(
                                 transaction_source, "value", transaction_source
                             ),
                             "sentry.origin": self.span_origin,
-                            "asgi.type": ty,
+                            "network.protocol.name": ty,
                         }
 
                         if ty in ("http", "websocket"):
@@ -302,6 +308,12 @@ async def _run_app(
                         )
 
                     with span_ctx as span:
+                        if isinstance(span, StreamedSpan):
+                            for attribute, value in _get_request_attributes(
+                                scope
+                            ).items():
+                                span.set_attribute(attribute, value)
+
                         try:
 
                             async def _sentry_wrapped_send(
@@ -336,6 +348,7 @@ async def _sentry_wrapped_send(
                                 return await self.app(
                                     scope, receive, _sentry_wrapped_send
                                 )
+
                         except Exception as exc:
                             suppress_chained_exceptions = (
                                 sentry_sdk.get_client()
@@ -350,6 +363,28 @@ async def _sentry_wrapped_send(
                             with capture_internal_exceptions():
                                 self._capture_request_exception(exc)
                             reraise(*exc_info)
+
+                        finally:
+                            if isinstance(span, StreamedSpan):
+                                already_set = (
+                                    span is not None
+                                    and span.name != _DEFAULT_TRANSACTION_NAME
+                                    and span.get_attributes().get("sentry.span.source")
+                                    in [
+                                        SegmentSource.COMPONENT.value,
+                                        SegmentSource.ROUTE.value,
+                                        SegmentSource.CUSTOM.value,
+                                    ]
+                                )
+                                with capture_internal_exceptions():
+                                    if not already_set:
+                                        name, source = (
+                                            self._get_segment_name_and_source(
+                                                self.transaction_style, scope
+                                            )
+                                        )
+                                        span.name = name
+                                        span.set_attribute("sentry.span.source", source)
         finally:
             _asgi_middleware_applied.set(False)
 
@@ -424,3 +459,40 @@ def _get_transaction_name_and_source(
             return name, source
 
         return name, source
+
+    def _get_segment_name_and_source(
+        self: "SentryAsgiMiddleware", segment_style: str, asgi_scope: "Any"
+    ) -> "Tuple[str, str]":
+        name = None
+        source = SEGMENT_SOURCE_FOR_STYLE[segment_style].value
+        ty = asgi_scope.get("type")
+
+        if segment_style == "endpoint":
+            endpoint = asgi_scope.get("endpoint")
+            # Webframeworks like Starlette mutate the ASGI env once routing is
+            # done, which is sometime after the request has started. If we have
+            # an endpoint, overwrite our generic transaction name.
+            if endpoint:
+                name = qualname_from_function(endpoint) or ""
+            else:
+                name = _get_url(asgi_scope, "http" if ty == "http" else "ws", host=None)
+                source = SegmentSource.URL.value
+
+        elif segment_style == "url":
+            # FastAPI includes the route object in the scope to let Sentry extract the
+            # path from it for the transaction name
+            route = asgi_scope.get("route")
+            if route:
+                path = getattr(route, "path", None)
+                if path is not None:
+                    name = path
+            else:
+                name = _get_url(asgi_scope, "http" if ty == "http" else "ws", host=None)
+                source = SegmentSource.URL.value
+
+        if name is None:
+            name = _DEFAULT_TRANSACTION_NAME
+            source = SegmentSource.ROUTE.value
+            return name, source
+
+        return name, source

sentry_sdk/integrations/_wsgi_common.py

@@ -1,22 +1,20 @@
-import sys
 import inspect
+import sys
 from functools import wraps
+from typing import TYPE_CHECKING
 
 import sentry_sdk
 from sentry_sdk.ai.monitoring import record_token_usage
 from sentry_sdk.ai.utils import set_data_normalized
 from sentry_sdk.consts import OP, SPANDATA
 from sentry_sdk.integrations import DidNotEnable, Integration
 from sentry_sdk.scope import should_send_default_pii
-from sentry_sdk.tracing_utils import set_span_errored
 from sentry_sdk.utils import (
     capture_internal_exceptions,
     event_from_exception,
     reraise,
 )
 
-from typing import TYPE_CHECKING
-
 if TYPE_CHECKING:
     from typing import Any, Callable, Iterable
 
@@ -53,8 +51,6 @@ def setup_once() -> None:
 
 
 def _capture_exception(exc: "Any") -> None:
-    set_span_errored()
-
     event, hint = event_from_exception(
         exc,
         client_options=sentry_sdk.get_client().options,
@@ -131,7 +127,7 @@ def new_huggingface_task(*args: "Any", **kwargs: "Any") -> "Any":
             exc_info = sys.exc_info()
             with capture_internal_exceptions():
                 _capture_exception(e)
-                span.__exit__(None, None, None)
+                span.__exit__(*exc_info)
             reraise(*exc_info)
 
         # Output attributes