Adding WITHATTRIBS option to vector set's vsim command.

.github/workflows/integration.yaml

@@ -27,8 +27,8 @@ env:
   CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
   # this speeds up coverage with Python 3.12: https://github.com/nedbat/coveragepy/issues/1665
   COVERAGE_CORE: sysmon
-  CURRENT_CLIENT_LIBS_TEST_STACK_IMAGE_TAG: '8.0.2'
-  CURRENT_REDIS_VERSION: '8.0.2'
+  CURRENT_CLIENT_LIBS_TEST_STACK_IMAGE_TAG: '8.2'
+  CURRENT_REDIS_VERSION: '8.2'
 
 jobs:
   dependency-audit:
@@ -74,7 +74,7 @@ jobs:
       max-parallel: 15
       fail-fast: false
       matrix:
-        redis-version: ['8.2', '${{ needs.redis_version.outputs.CURRENT }}', '7.4.4', '7.2.9']
+        redis-version: ['8.2.1-pre', '${{ needs.redis_version.outputs.CURRENT }}', '8.0.2' ,'7.4.4', '7.2.9']
         python-version: ['3.9', '3.13']
         parser-backend: ['plain']
         event-loop: ['asyncio']

docker-compose.yml

@@ -1,10 +1,10 @@
 ---
 # image tag 8.0-RC2-pre is the one matching the 8.0 GA release
 x-client-libs-stack-image: &client-libs-stack-image
-  image: "redislabs/client-libs-test:${CLIENT_LIBS_TEST_STACK_IMAGE_TAG:-8.0.2}"
+  image: "redislabs/client-libs-test:${CLIENT_LIBS_TEST_STACK_IMAGE_TAG:-8.2}"
 
 x-client-libs-image: &client-libs-image
-  image: "redislabs/client-libs-test:${CLIENT_LIBS_TEST_IMAGE_TAG:-8.0.2}"
+  image: "redislabs/client-libs-test:${CLIENT_LIBS_TEST_IMAGE_TAG:-8.2}"
 
 services:
 

redis/commands/vectorset/commands.py

@@ -1,6 +1,6 @@
 import json
 from enum import Enum
-from typing import Awaitable, Dict, List, Optional, Union
+from typing import Any, Awaitable, Dict, List, Optional, Union
 
 from redis.client import NEVER_DECODE
 from redis.commands.helpers import get_protocol_version
@@ -33,6 +33,7 @@ class CallbacksOptions(Enum):
 
     RAW = "RAW"
     WITHSCORES = "WITHSCORES"
+    WITHATTRIBS = "WITHATTRIBS"
     ALLOW_DECODING = "ALLOW_DECODING"
     RESP3 = "RESP3"
 
@@ -123,6 +124,7 @@ def vsim(
         key: KeyT,
         input: Union[List[float], bytes, str],
         with_scores: Optional[bool] = False,
+        with_attribs: Optional[bool] = False,
         count: Optional[int] = None,
         ef: Optional[Number] = None,
         filter: Optional[str] = None,
@@ -131,14 +133,35 @@ def vsim(
         no_thread: Optional[bool] = False,
         epsilon: Optional[Number] = None,
     ) -> Union[
-        Awaitable[Optional[List[Union[List[EncodableT], Dict[EncodableT, Number]]]]],
-        Optional[List[Union[List[EncodableT], Dict[EncodableT, Number]]]],
+        Awaitable[
+            Optional[
+                List[
+                    Union[
+                        List[EncodableT],
+                        Dict[EncodableT, Number],
+                        Dict[EncodableT, Dict[str, Any]],
+                    ]
+                ]
+            ]
+        ],
+        Optional[
+            List[
+                Union[
+                    List[EncodableT],
+                    Dict[EncodableT, Number],
+                    Dict[EncodableT, Dict[str, Any]],
+                ]
+            ]
+        ],
     ]:
         """
         Compare a vector or element ``input``  with the other vectors in a vector set ``key``.
 
-        ``with_scores`` sets if the results should be returned with the
-                similarity scores of the elements in the result.
+        ``with_scores`` sets if returns, for each element, the JSON attribute associated
+        with the element or None when no attributes are present.
+
+        ``with_attribs`` sets if the results should be returned with the
+                attributes of the elements in the result.
 
         ``count`` sets the number of results to return.
 
@@ -177,6 +200,10 @@ def vsim(
             pieces.append("WITHSCORES")
             options[CallbacksOptions.WITHSCORES.value] = True
 
+        if with_attribs:
+            pieces.append("WITHATTRIBS")
+            options[CallbacksOptions.WITHATTRIBS.value] = True
+
         if count:
             pieces.extend(["COUNT", count])
 

redis/commands/vectorset/utils.py

@@ -1,3 +1,5 @@
+import json
+
 from redis._parsers.helpers import pairs_to_dict
 from redis.commands.vectorset.commands import CallbacksOptions
 
@@ -75,19 +77,40 @@ def parse_vsim_result(response, **options):
     structures depending on input options.
     Parsing VSIM result into:
     - List[List[str]]
-    - List[Dict[str, Number]]
+    - List[Dict[str, Number]] - when with_scores is used (without attributes)
+    - List[Dict[str, Mapping[str, Any]]] - when with_attribs is used (without scores)
+    - List[Dict[str, Union[Number, Mapping[str, Any]]]] - when with_scores and with_attribs are used
+
     """
     if response is None:
         return response
 
-    if options.get(CallbacksOptions.WITHSCORES.value):
+    withscores = bool(options.get(CallbacksOptions.WITHSCORES.value))
+    withattribs = bool(options.get(CallbacksOptions.WITHATTRIBS.value))
+
+    if withscores ^ withattribs:
         # Redis will return a list of list of pairs.
         # This list have to be transformed to dict
         result_dict = {}
         for key, value in pairs_to_dict(response).items():
-            value = float(value)
+            if withscores:
+                value = float(value)
+            else:
+                value = json.loads(value) if value else None
+
             result_dict[key] = value
         return result_dict
+    elif withscores and withattribs:
+        it = iter(response)
+        result_dict = {}
+        for elem, score, attribs in zip(it, it, it):
+            if attribs is not None:
+                attribs_dict = json.loads(attribs)
+            else:
+                attribs_dict = None
+
+            result_dict[elem] = {"score": float(score), "attributes": attribs_dict}
+        return result_dict
     else:
         # return the list of elements for each level
         # list of lists

tests/test_asyncio/test_vsets.py

@@ -262,6 +262,80 @@ async def test_vsim_with_scores(d_client):
     assert 0 <= vsim["elem1"] <= 1
 
 
+@skip_if_server_version_lt("7.9.0")
+async def test_vsim_with_attribs_attribs_set(d_client):
+    elements_count = 5
+    vector_dim = 10
+    attrs_dict = {"key1": "value1", "key2": "value2"}
+    for i in range(elements_count):
+        float_array = [random.uniform(0, 5) for x in range(vector_dim)]
+        await d_client.vset().vadd(
+            "myset",
+            float_array,
+            f"elem{i}",
+            numlinks=64,
+            attributes=attrs_dict if i % 2 == 0 else None,
+        )
+
+    vsim = await d_client.vset().vsim("myset", input="elem1", with_attribs=True)
+    assert len(vsim) == 5
+    assert isinstance(vsim, dict)
+    assert vsim["elem1"] is None
+    assert vsim["elem2"] == attrs_dict
+
+
+@skip_if_server_version_lt("7.9.0")
+async def test_vsim_with_scores_and_attribs_attribs_set(d_client):
+    elements_count = 5
+    vector_dim = 10
+    attrs_dict = {"key1": "value1", "key2": "value2"}
+    for i in range(elements_count):
+        float_array = [random.uniform(0, 5) for x in range(vector_dim)]
+        await d_client.vset().vadd(
+            "myset",
+            float_array,
+            f"elem{i}",
+            numlinks=64,
+            attributes=attrs_dict if i % 2 == 0 else None,
+        )
+
+    vsim = await d_client.vset().vsim(
+        "myset", input="elem1", with_scores=True, with_attribs=True
+    )
+    assert len(vsim) == 5
+    assert isinstance(vsim, dict)
+    assert isinstance(vsim["elem1"], dict)
+    assert "score" in vsim["elem1"]
+    assert "attributes" in vsim["elem1"]
+    assert isinstance(vsim["elem1"]["score"], float)
+    assert vsim["elem1"]["attributes"] is None
+
+    assert isinstance(vsim["elem2"], dict)
+    assert "score" in vsim["elem2"]
+    assert "attributes" in vsim["elem2"]
+    assert isinstance(vsim["elem2"]["score"], float)
+    assert vsim["elem2"]["attributes"] == attrs_dict
+
+
+@skip_if_server_version_lt("7.9.0")
+async def test_vsim_with_attribs_attribs_not_set(d_client):
+    elements_count = 20
+    vector_dim = 50
+    for i in range(elements_count):
+        float_array = [random.uniform(0, 10) for x in range(vector_dim)]
+        await d_client.vset().vadd(
+            "myset",
+            float_array,
+            f"elem{i}",
+            numlinks=64,
+        )
+
+    vsim = await d_client.vset().vsim("myset", input="elem1", with_attribs=True)
+    assert len(vsim) == 10
+    assert isinstance(vsim, dict)
+    assert vsim["elem1"] is None
+
+
 @skip_if_server_version_lt("7.9.0")
 async def test_vsim_with_different_vector_input_types(d_client):
     elements_count = 10
@@ -789,9 +863,12 @@ async def test_vrandmember(d_client):
 async def test_vset_commands_without_decoding_responces(client):
     # test vadd
     elements = ["elem1", "elem2", "elem3"]
+    attrs_dict = {"key1": "value1", "key2": "value2"}
     for elem in elements:
         float_array = [random.uniform(0.5, 10) for x in range(0, 8)]
-        resp = await client.vset().vadd("myset", float_array, element=elem)
+        resp = await client.vset().vadd(
+            "myset", float_array, element=elem, attributes=attrs_dict
+        )
         assert resp == 1
 
     # test vemb
@@ -820,6 +897,28 @@ async def test_vset_commands_without_decoding_responces(client):
     assert isinstance(vsim_with_scores, dict)
     assert isinstance(vsim_with_scores[b"elem1"], float)
 
+    # test vsim with attributes
+    vsim_with_attribs = await client.vset().vsim(
+        "myset", input="elem1", with_attribs=True
+    )
+    assert len(vsim_with_attribs) == 3
+    assert isinstance(vsim_with_attribs, dict)
+    assert isinstance(vsim_with_attribs[b"elem1"], dict)
+    assert vsim_with_attribs[b"elem1"] == attrs_dict
+
+    # test vsim with score and attributes
+    vsim_with_scores_and_attribs = await client.vset().vsim(
+        "myset", input="elem1", with_scores=True, with_attribs=True
+    )
+    assert len(vsim_with_scores_and_attribs) == 3
+    assert isinstance(vsim_with_scores_and_attribs, dict)
+    assert isinstance(vsim_with_scores_and_attribs[b"elem1"], dict)
+    assert "score" in vsim_with_scores_and_attribs[b"elem1"]
+    assert "attributes" in vsim_with_scores_and_attribs[b"elem1"]
+    assert isinstance(vsim_with_scores_and_attribs[b"elem1"]["score"], float)
+    assert isinstance(vsim_with_scores_and_attribs[b"elem1"]["attributes"], dict)
+    assert vsim_with_scores_and_attribs[b"elem1"]["attributes"] == attrs_dict
+
     # test vlinks - no scores
     element_links_all_layers = await client.vset().vlinks("myset", "elem1")
     assert len(element_links_all_layers) >= 1