merge master

Author: alexander-alderman-webb

.github/workflows/enforce-draft-pr.yml

@@ -0,0 +1,69 @@
+name: Enforce Draft PR
+
+on:
+  pull_request_target:
+    types: [opened, reopened]
+
+jobs:
+  enforce-draft:
+    name: Enforce Draft PR
+    runs-on: ubuntu-24.04
+    if: github.event.pull_request.draft == false
+    steps:
+      - name: Generate GitHub App token
+        id: app-token
+        uses: actions/create-github-app-token@f8d387b68d61c58ab83c6c016672934102569859  # v2
+        with:
+          app-id: ${{ vars.SDK_MAINTAINER_BOT_APP_ID }}
+          private-key: ${{ secrets.SDK_MAINTAINER_BOT_PRIVATE_KEY }}
+
+      - name: Convert PR to draft
+        env:
+          GH_TOKEN: ${{ steps.app-token.outputs.token }}
+          PR_URL: ${{ github.event.pull_request.html_url }}
+        run: |
+          gh pr ready "$PR_URL" --undo
+
+      - name: Label and comment
+        uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea  # v7.0.1
+        with:
+          github-token: ${{ steps.app-token.outputs.token }}
+          script: |
+            const pullRequest = context.payload.pull_request;
+            const repo = context.repo;
+
+            // Label the PR so maintainers can filter/track violations
+            await github.rest.issues.addLabels({
+              ...repo,
+              issue_number: pullRequest.number,
+              labels: ['converted-to-draft'],
+            });
+
+            // Check for existing bot comment to avoid duplicates on reopen
+            const comments = await github.rest.issues.listComments({
+              ...repo,
+              issue_number: pullRequest.number,
+            });
+            const botComment = comments.data.find(c =>
+              c.user.type === 'Bot' &&
+              c.body.includes('automatically converted to draft')
+            );
+            if (botComment) {
+              core.info('Bot comment already exists, skipping.');
+              return;
+            }
+
+            const contributingUrl = `https://github.com/${repo.owner}/${repo.repo}/blob/master/CONTRIBUTING.md`;
+
+            await github.rest.issues.createComment({
+              ...repo,
+              issue_number: pullRequest.number,
+              body: [
+                `This PR has been automatically converted to draft. All PRs must start as drafts per our [contributing guidelines](${contributingUrl}).`,
+                '',
+                '**Next steps:**',
+                '1. Ensure CI passes',
+                '2. Fill in the PR description completely',
+                '3. Mark as "Ready for review" when you\'re done'
+              ].join('\n')
+            });

CONTRIBUTING.md

@@ -10,6 +10,10 @@ Please search the [issue tracker](https://github.com/getsentry/sentry-python/iss
 
 ## Submitting Changes
 
+Before submitting a pull request, please check whether the issue you're planning to address, if there is one, is assigned to anyone. If you want to help with an issue that already has an assignee, comment on it first to coordinate with the person assigned whether the contribution would make sense before opening a pull request.
+
+To make a contribution:
+
 - Fork the `sentry-python` repo and prepare your changes.
 - Add tests for your changes to `tests/`.
 - Run tests and make sure all of them pass.
@@ -21,6 +25,18 @@ We will review your pull request as soon as possible. Thank you for contributing
 
 You are welcome to use whatever tools you prefer for making a contribution. However, any changes you propose have to be reviewed and tested by you, a human, first, before you submit a pull request with them for the Sentry team to review. If we feel like that didn't happen, we will close the PR outright. For example, we won't review visibly AI-generated PRs from an agent instructed to look for and "fix" open issues in the repo.
 
+## Pull Requests
+
+All PRs must be created as **drafts**. Non-draft PRs will be automatically converted to draft. Mark your PR as "Ready for review" once:
+
+- CI passes
+- The PR description is complete (what, why, and links to relevant issues)
+- You've personally reviewed your own changes
+
+A PR should do one thing well. Don't mix functional changes with unrelated refactors or cleanup. Smaller, focused PRs are easier to review, reason about, and revert if needed.
+
+For the full set of PR standards, see the [code submission standard](https://develop.sentry.dev/sdk/getting-started/standards/code-submission/#pull-requests).
+
 ## Development Environment
 
 ### Set up Python

sentry_sdk/integrations/asyncpg.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 import contextlib
+import re
 from typing import Any, TypeVar, Callable, Awaitable, Iterator
 
 import sentry_sdk
@@ -55,6 +56,10 @@ def setup_once() -> None:
 T = TypeVar("T")
 
 
+def _normalize_query(query: str) -> str:
+    return re.sub(r"\s+", " ", query).strip()
+
+
 def _wrap_execute(f: "Callable[..., Awaitable[T]]") -> "Callable[..., Awaitable[T]]":
     async def _inner(*args: "Any", **kwargs: "Any") -> "T":
         if sentry_sdk.get_client().get_integration(AsyncPGIntegration) is None:
@@ -67,7 +72,7 @@ async def _inner(*args: "Any", **kwargs: "Any") -> "T":
         if len(args) > 2:
             return await f(*args, **kwargs)
 
-        query = args[1]
+        query = _normalize_query(args[1])
         with record_sql_queries(
             cursor=None,
             query=query,
@@ -103,6 +108,7 @@ def _record(
 
     param_style = "pyformat" if params_list else None
 
+    query = _normalize_query(query)
     with record_sql_queries(
         cursor=cursor,
         query=query,

sentry_sdk/integrations/openai.py

@@ -443,43 +443,9 @@ def _set_embeddings_input_data(
         "input"
     )
 
-    tools = kwargs.get("tools")
-    if tools is not None and _is_given(tools) and len(tools) > 0:
-        set_data_normalized(
-            span, SPANDATA.GEN_AI_REQUEST_AVAILABLE_TOOLS, safe_serialize(tools)
-        )
-
     model = kwargs.get("model")
-    if model is not None and _is_given(model):
-        set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_MODEL, model)
-
-    stream = kwargs.get("stream")
-    if stream is not None and _is_given(stream):
-        set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_STREAMING, stream)
-
-    max_tokens = kwargs.get("max_tokens")
-    if max_tokens is not None and _is_given(max_tokens):
-        set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_MAX_TOKENS, max_tokens)
-
-    presence_penalty = kwargs.get("presence_penalty")
-    if presence_penalty is not None and _is_given(presence_penalty):
-        set_data_normalized(
-            span, SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY, presence_penalty
-        )
-
-    frequency_penalty = kwargs.get("frequency_penalty")
-    if frequency_penalty is not None and _is_given(frequency_penalty):
-        set_data_normalized(
-            span, SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY, frequency_penalty
-        )
-
-    temperature = kwargs.get("temperature")
-    if temperature is not None and _is_given(temperature):
-        set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TEMPERATURE, temperature)
-
-    top_p = kwargs.get("top_p")
-    if top_p is not None and _is_given(top_p):
-        set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TOP_P, top_p)
+    if model is not None:
+        span.set_data(SPANDATA.GEN_AI_REQUEST_MODEL, model)
 
     if (
         not should_send_default_pii()

tests/integrations/asyncpg/test_asyncpg.py

@@ -463,10 +463,7 @@ async def test_connection_pool(sentry_init, capture_events) -> None:
         {
             "category": "query",
             "data": {},
-            "message": "SELECT pg_advisory_unlock_all();\n"
-            "CLOSE ALL;\n"
-            "UNLISTEN *;\n"
-            "RESET ALL;",
+            "message": "SELECT pg_advisory_unlock_all(); CLOSE ALL; UNLISTEN *; RESET ALL;",
             "type": "default",
         },
         {
@@ -478,10 +475,7 @@ async def test_connection_pool(sentry_init, capture_events) -> None:
         {
             "category": "query",
             "data": {},
-            "message": "SELECT pg_advisory_unlock_all();\n"
-            "CLOSE ALL;\n"
-            "UNLISTEN *;\n"
-            "RESET ALL;",
+            "message": "SELECT pg_advisory_unlock_all(); CLOSE ALL; UNLISTEN *; RESET ALL;",
             "type": "default",
         },
     ]
@@ -786,3 +780,79 @@ async def test_span_origin(sentry_init, capture_events):
 
     for span in event["spans"]:
         assert span["origin"] == "auto.db.asyncpg"
+
+
+@pytest.mark.asyncio
+async def test_multiline_query_description_normalized(sentry_init, capture_events):
+    sentry_init(
+        integrations=[AsyncPGIntegration()],
+        traces_sample_rate=1.0,
+    )
+    events = capture_events()
+
+    with start_transaction(name="test_transaction"):
+        conn: Connection = await connect(PG_CONNECTION_URI)
+        await conn.execute(
+            """
+            SELECT
+                id,
+                name
+            FROM
+                users
+            WHERE
+                name = 'Alice'
+            """
+        )
+        await conn.close()
+
+    (event,) = events
+
+    spans = [
+        s
+        for s in event["spans"]
+        if s["op"] == "db" and "SELECT" in s.get("description", "")
+    ]
+    assert len(spans) == 1
+    assert spans[0]["description"] == "SELECT id, name FROM users WHERE name = 'Alice'"
+
+
+@pytest.mark.asyncio
+async def test_before_send_transaction_sees_normalized_description(
+    sentry_init, capture_events
+):
+    def before_send_transaction(event, hint):
+        for span in event.get("spans", []):
+            desc = span.get("description", "")
+            if "SELECT id, name FROM users" in desc:
+                span["description"] = "filtered"
+        return event
+
+    sentry_init(
+        integrations=[AsyncPGIntegration()],
+        traces_sample_rate=1.0,
+        before_send_transaction=before_send_transaction,
+    )
+    events = capture_events()
+
+    with start_transaction(name="test_transaction"):
+        conn: Connection = await connect(PG_CONNECTION_URI)
+        await conn.execute(
+            """
+            SELECT
+                id,
+                name
+            FROM
+                users
+            """
+        )
+        await conn.close()
+
+    (event,) = events
+    spans = [
+        s
+        for s in event["spans"]
+        if s["op"] == "db" and "filtered" in s.get("description", "")
+    ]
+
+    assert len(spans) == 1
+    assert spans[0]["description"] == "filtered"

tests/integrations/openai/test_openai.py

@@ -1145,6 +1145,7 @@ def test_embeddings_create_no_pii(
     span = tx["spans"][0]
     assert span["op"] == "gen_ai.embeddings"
     assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
+    assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "text-embedding-3-large"
 
     assert SPANDATA.GEN_AI_EMBEDDINGS_INPUT not in span["data"]
 
@@ -1226,6 +1227,7 @@ def test_embeddings_create(sentry_init, capture_events, input, request):
     span = tx["spans"][0]
     assert span["op"] == "gen_ai.embeddings"
     assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
+    assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "text-embedding-3-large"
 
     param_id = request.node.callspec.id
     if param_id == "string":
@@ -1298,6 +1300,7 @@ async def test_embeddings_create_async_no_pii(
     span = tx["spans"][0]
     assert span["op"] == "gen_ai.embeddings"
     assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
+    assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "text-embedding-3-large"
 
     assert SPANDATA.GEN_AI_EMBEDDINGS_INPUT not in span["data"]
 
@@ -1382,6 +1385,7 @@ async def test_embeddings_create_async(sentry_init, capture_events, input, reque
     span = tx["spans"][0]
     assert span["op"] == "gen_ai.embeddings"
     assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
+    assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "text-embedding-3-large"
 
     param_id = request.node.callspec.id
     if param_id == "string":