Merge branch 'master' into webb/litellm/close-spans

Author: alexander-alderman-webb

.github/workflows/validate-pr.yml

@@ -10,7 +10,7 @@ jobs:
     permissions:
       pull-requests: write
     steps:
-      - uses: getsentry/github-workflows/validate-pr@0b52fc6a867b744dcbdf5d25c18bc8d1c95710e1
+      - uses: getsentry/github-workflows/validate-pr@71588ddf95134f804e82c5970a8098588e2eaecd
         with:
           app-id: ${{ vars.SDK_MAINTAINER_BOT_APP_ID }}
           private-key: ${{ secrets.SDK_MAINTAINER_BOT_PRIVATE_KEY }}

scripts/populate_tox/config.py

@@ -18,6 +18,9 @@
         "deps": {
             "*": ["pytest-asyncio"],
             "<0.50": ["httpx<0.28.0"],
+            # tokenizers dropped Python 3.8 support, but didn't update package metadata.
+            # https://github.com/huggingface/tokenizers/commit/f4c9fd7f402fc794df8f1b547a95ee5305f9fe62
+            "py3.8": ["tokenizers<0.20.4"],
         },
         "python": ">=3.8",
     },

sentry_sdk/_werkzeug.py

@@ -38,6 +38,7 @@
     from typing import Dict
     from typing import Iterator
     from typing import Tuple
+    from typing import Optional
 
 
 #
@@ -62,35 +63,41 @@ def _get_headers(environ: "Dict[str, str]") -> "Iterator[Tuple[str, str]]":
             yield key.replace("_", "-").title(), value
 
 
-#
+def _strip_default_port(host: str, scheme: "Optional[str]") -> str:
+    """Strip the port from the host if it's the default for the scheme."""
+    if scheme == "http" and host.endswith(":80"):
+        return host[:-3]
+    if scheme == "https" and host.endswith(":443"):
+        return host[:-4]
+    return host
+
+
 # `get_host` comes from `werkzeug.wsgi.get_host`
 # https://github.com/pallets/werkzeug/blob/1.0.1/src/werkzeug/wsgi.py#L145
-#
+
+
 def get_host(environ: "Dict[str, str]", use_x_forwarded_for: bool = False) -> str:
     """
     Return the host for the given WSGI environment.
     """
+    scheme = environ.get("wsgi.url_scheme")
+    if use_x_forwarded_for:
+        scheme = environ.get("HTTP_X_FORWARDED_PROTO", scheme)
+
     if use_x_forwarded_for and "HTTP_X_FORWARDED_HOST" in environ:
-        rv = environ["HTTP_X_FORWARDED_HOST"]
-        if environ["wsgi.url_scheme"] == "http" and rv.endswith(":80"):
-            rv = rv[:-3]
-        elif environ["wsgi.url_scheme"] == "https" and rv.endswith(":443"):
-            rv = rv[:-4]
+        return _strip_default_port(environ["HTTP_X_FORWARDED_HOST"], scheme)
     elif environ.get("HTTP_HOST"):
-        rv = environ["HTTP_HOST"]
-        if environ["wsgi.url_scheme"] == "http" and rv.endswith(":80"):
-            rv = rv[:-3]
-        elif environ["wsgi.url_scheme"] == "https" and rv.endswith(":443"):
-            rv = rv[:-4]
+        return _strip_default_port(environ["HTTP_HOST"], scheme)
     elif environ.get("SERVER_NAME"):
+        # SERVER_NAME/SERVER_PORT describe the internal server, so use
+        # wsgi.url_scheme (not the forwarded scheme) for port decisions.
         rv = environ["SERVER_NAME"]
         if (environ["wsgi.url_scheme"], environ["SERVER_PORT"]) not in (
             ("https", "443"),
             ("http", "80"),
         ):
             rv += ":" + environ["SERVER_PORT"]
+        return rv
     else:
         # In spite of the WSGI spec, SERVER_NAME might not be present.
-        rv = "unknown"
-
-    return rv
+        return "unknown"

sentry_sdk/ai/consts.py

@@ -0,0 +1,6 @@
+import re
+
+# Matches data URLs with base64-encoded content, e.g. "data:image/png;base64,iVBORw0K..."
+DATA_URL_BASE64_REGEX = re.compile(
+    r"^data:(?:[a-zA-Z0-9][a-zA-Z0-9.+\-]*/[a-zA-Z0-9][a-zA-Z0-9.+\-]*)(?:;[a-zA-Z0-9\-]+=[^;,]*)*;base64,(?:[A-Za-z0-9+/\-_]+={0,2})$"
+)

sentry_sdk/ai/utils.py

@@ -4,6 +4,7 @@
 from typing import TYPE_CHECKING
 
 from sentry_sdk._types import BLOB_DATA_SUBSTITUTE
+from sentry_sdk.ai.consts import DATA_URL_BASE64_REGEX
 
 if TYPE_CHECKING:
     from typing import Any, Callable, Dict, List, Optional, Tuple
@@ -588,6 +589,20 @@ def _find_truncation_index(messages: "List[Dict[str, Any]]", max_bytes: int) ->
     return 0
 
 
+def _is_image_type_with_blob_content(item: "Dict[str, Any]") -> bool:
+    """
+    Some content blocks contain an image_url property with base64 content as its value.
+    This is used to identify those while not leading to unnecessary copying of data when the image URL does not contain base64 content.
+    """
+    if item.get("type") != "image_url":
+        return False
+
+    image_url = item.get("image_url", {}).get("url", "")
+    data_url_match = DATA_URL_BASE64_REGEX.match(image_url)
+
+    return bool(data_url_match)
+
+
 def redact_blob_message_parts(
     messages: "List[Dict[str, Any]]",
 ) -> "List[Dict[str, Any]]":
@@ -640,7 +655,9 @@ def redact_blob_message_parts(
         content = message.get("content")
         if isinstance(content, list):
             for item in content:
-                if isinstance(item, dict) and item.get("type") == "blob":
+                if isinstance(item, dict) and (
+                    item.get("type") == "blob" or _is_image_type_with_blob_content(item)
+                ):
                     has_blobs = True
                     break
         if has_blobs:
@@ -661,8 +678,11 @@ def redact_blob_message_parts(
         content = message.get("content")
         if isinstance(content, list):
             for item in content:
-                if isinstance(item, dict) and item.get("type") == "blob":
-                    item["content"] = BLOB_DATA_SUBSTITUTE
+                if isinstance(item, dict):
+                    if item.get("type") == "blob":
+                        item["content"] = BLOB_DATA_SUBSTITUTE
+                    elif _is_image_type_with_blob_content(item):
+                        item["image_url"]["url"] = BLOB_DATA_SUBSTITUTE
 
     return messages_copy
 

sentry_sdk/consts.py

@@ -636,12 +636,6 @@ class SPANDATA:
     Example: "rainy, 57°F"
     """
 
-    GEN_AI_TOOL_TYPE = "gen_ai.tool.type"
-    """
-    The type of tool being used.
-    Example: "function"
-    """
-
     GEN_AI_USAGE_INPUT_TOKENS = "gen_ai.usage.input_tokens"
     """
     The number of tokens in the input.