Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,7 @@

import hashlib
import json
import re
import sys
from typing import TYPE_CHECKING, Any, ClassVar, Literal, Self

Expand All @@ -24,6 +25,7 @@
from redisvl.query import AggregateHybridQuery, FilterQuery, TextQuery
from redisvl.query.filter import FilterExpression
from redisvl.redis.utils import array_to_buffer
from redisvl.utils.token_escaper import TokenEscaper
from redisvl.utils.vectorize import BaseVectorizer

from ._feature_usage import FeatureIndex
Expand All @@ -43,6 +45,14 @@
_TAG_DEFAULTS = {"separator": ",", "case_sensitive": False, "withsuffixtrie": False}
_TEXT_DEFAULTS = {"weight": 1.0, "no_stem": False}

# RediSearch reads several punctuation marks as query syntax — `|` unions terms, `?` is a fuzzy
# marker, backticks quote exact phrases — so a user message like "ps aux | grep python" or
# "Why ?" reaches the engine as a malformed query and raises a syntax error that fails the whole
# agent run. The search text is plain prose, not a query: escape the syntax characters before it
# reaches TextQuery / AggregateHybridQuery. Unlike the default TokenEscaper set (built for tag
# filter values), spaces stay unescaped so the terms keep tokenizing normally.
_SEARCH_TEXT_ESCAPER = TokenEscaper(re.compile(r"[,.<>{}\[\]\\\"':;!@#$%^&*()\-+=~/|?`]"))


def _scope_token_field(name: str) -> str:
if name not in _SCOPE_FIELDS:
Expand Down Expand Up @@ -405,7 +415,7 @@ async def _redis_search(
await self._ensure_index()
self._validate_filters()

q = (text or "").strip()
q = _SEARCH_TEXT_ESCAPER.escape((text or "").strip())
if not q:
raise IntegrationInvalidRequestException("text_search() requires non-empty text")
num_results = max(int(num_results or 10), 1)
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -95,16 +95,17 @@ def __init__(
Args:
source_id: Unique identifier for this provider instance.
redis_url: Redis connection URL (e.g., "redis://localhost:6379").
Mutually exclusive with credential_provider and redis_client.
Ignored when redis_client is supplied.
credential_provider: Redis credential provider for Azure AD authentication.
Requires host parameter. Mutually exclusive with redis_url and redis_client.
Requires host parameter. Ignored when redis_client is supplied.
host: Redis host name. Required when using credential_provider.
port: Redis port number. Defaults to 6380 (Azure Redis SSL port).
ssl: Enable SSL/TLS connection. Defaults to True.
username: Redis username.
redis_client: Borrowed standalone async Redis client configured with
``decode_responses=True``. The caller owns its lifetime. Mutually exclusive
with redis_url and credential_provider.
``decode_responses=True``. The caller owns its lifetime and takes
precedence over redis_url, credential_provider, and connection
defaults, mirroring the Redis vector-store constructors.
key_prefix: Base prefix for Redis keys. Scoped mode appends independently encoded
tenant, application, agent, provider source, and session segments.
Defaults to 'chat_messages'.
Expand All @@ -130,8 +131,8 @@ def __init__(
Raises:
TypeError: If redis_client is not a standalone async Redis client.
ValueError: If no Redis connection source is provided.
ValueError: If more than one Redis connection source is provided.
ValueError: If credential_provider is used without host parameter.
ValueError: If more than one non-client Redis connection source is provided.
ValueError: If credential_provider is used without host parameter and no client is provided.
ValueError: If redis_client does not use decoded string responses.
ValueError: If max_messages is negative.
ValueError: If key_format or its scoped identifiers are invalid.
Expand All @@ -148,9 +149,9 @@ def __init__(
connection_sources = sum(source is not None for source in (redis_client, redis_url, credential_provider))
if connection_sources == 0:
raise ValueError("Either redis_client, redis_url, or credential_provider must be provided")
if connection_sources > 1:
if connection_sources > 1 and redis_client is None:
raise ValueError("redis_client, redis_url, and credential_provider are mutually exclusive")
if credential_provider is not None and host is None:
if credential_provider is not None and host is None and redis_client is None:
raise ValueError("host is required when using credential_provider")
if redis_client is not None:
if not isinstance(redis_client, redis.Redis):
Expand Down
62 changes: 62 additions & 0 deletions python/packages/redis/tests/test_context_provider_edges.py
Original file line number Diff line number Diff line change
Expand Up @@ -353,6 +353,68 @@ async def test_redis_search_combines_explicit_filter_expression(
mock_index.query.assert_awaited_once()


@pytest.mark.parametrize(
("message", "broken_shape", "escaped"),
[
("Why ?", "| ?)", r"\?"),
("ps aux | grep python", "| |)", r"\|"),
("Is it A | B?", "| |)", r"\|"),
("Run ` ls ` please", "| ` |", r"\`"),
("| name | age |", "| |)", r"\|"),
],
)
async def test_redis_search_escapes_query_syntax_in_user_text(
message: str,
broken_shape: str,
escaped: str,
mock_index: AsyncMock,
patch_index: MagicMock, # noqa: ARG001
) -> None:
"""User prose reaches RediSearch as escaped terms, not as query syntax (#10002).

A bare `|`, `?` or backtick in a message previously turned the generated query
string into malformed grammar (`@content:(ps | aux | | | grep | python)`), and
the syntax error failed the whole agent run.
"""
provider = RedisContextProvider(source_id="ctx", user_id="user-1")

with patch.object(provider, "_ensure_index", AsyncMock()):
await provider._redis_search(text=message)

query = mock_index.query.await_args.args[0]
query_string = query.query_string()
assert broken_shape not in query_string
assert escaped in query_string


async def test_redis_search_escapes_text_for_hybrid_query(
mock_index: AsyncMock,
patch_index: MagicMock, # noqa: ARG001
) -> None:
"""The hybrid vector path gets the same escaped text."""
mock_vectorizer = MagicMock(spec=BaseVectorizer)
mock_vectorizer.dims = 128
mock_vectorizer.dtype = "float32"
mock_vectorizer.aembed = AsyncMock(return_value=[0.1] * 128)

provider = RedisContextProvider(
source_id="ctx",
user_id="u1",
redis_vectorizer=mock_vectorizer,
vector_field_name="embedding",
)

with (
patch.object(provider, "_ensure_index", AsyncMock()),
patch("agent_framework_redis._context_provider.AggregateHybridQuery") as hybrid_query,
):
hybrid_query.return_value = MagicMock()
await provider._redis_search(text="ps aux | grep python")

assert hybrid_query.call_args.kwargs["text"] == r"ps aux \| grep python"
mock_index.query.assert_awaited_once()


async def test_search_all_collects_paginated_batches(
mock_index: AsyncMock,
patch_index: MagicMock, # noqa: ARG001
Expand Down
30 changes: 21 additions & 9 deletions python/packages/redis/tests/test_providers.py
Original file line number Diff line number Diff line change
Expand Up @@ -417,19 +417,31 @@ def test_borrowed_client_is_used_directly(self):
"connection_kwargs",
[
{"redis_url": "redis://localhost:6379"},
{"credential_provider": MagicMock(), "host": "myhost"},
{"credential_provider": None, "host": "myhost"},
],
)
def test_borrowed_client_rejects_other_connection_sources(self, connection_kwargs: dict[str, Any]):
def test_borrowed_client_takes_precedence_over_other_connection_sources(self, connection_kwargs: dict[str, Any]):
"""An injected client wins; URL / credential settings are never validated (#9966).

The vector-store constructors already behave this way: a working
caller-owned client must not be rejected because a stray URL or
credential provider came along with it.
"""
kwargs = dict(connection_kwargs)
if kwargs.get("credential_provider") is None:
kwargs["credential_provider"] = MagicMock()

borrowed = Redis(decode_responses=True)

with pytest.raises(ValueError, match="mutually exclusive"):
RedisHistoryProvider(
"mem",
redis_client=borrowed,
application_id="test-app",
**connection_kwargs,
)
provider = RedisHistoryProvider(
"mem",
redis_client=borrowed,
application_id="test-app",
**kwargs,
)

assert provider._redis_client is borrowed
assert provider._owns_client is False

def test_borrowed_client_must_be_async_redis(self):
with pytest.raises(TypeError, match=r"redis\.asyncio\.Redis"):
Expand Down
Loading