Files
qdrant/tests/openapi/test_query.py
Andrey Vasnetsov 8a7325ad5d Remove deprecated search endpoints from OpenAPI, deprecate them in gRPC (#9982)
* Remove deprecated search/recommend/discover endpoints from OpenAPI

Remove deprecated REST API endpoint definitions from the OpenAPI
generator. These endpoints were deprecated in v1.13.3 (`f4ced2567`,
#5907, 2025-01-30) in favor of the universal `/points/query` endpoint:

- POST /points/search
- POST /points/search/batch
- POST /points/search/groups
- POST /points/recommend
- POST /points/recommend/batch
- POST /points/recommend/groups
- POST /points/discover
- POST /points/discover/batch

Also removes the corresponding request types from the schema generator
and updates the expected API count in the consistency check.

Co-authored-by: Cursor <cursoragent@cursor.com>

* Migrate OpenAPI integration tests to /points/query

The deprecated /points/search, /points/recommend and /points/discover
endpoints (along with their /batch and /groups variants) were removed
from the OpenAPI spec, which caused validation failures in the Python
integration test harness.

This commit migrates the affected tests to the universal /points/query
endpoint:

- Delete tests dedicated to the deprecated endpoints:
  test_recommend.py, test_discover.py, test_multicollection_reco.py,
  test_recommendation_multivector.py
- Refactor remaining tests to call /points/query (and /query/batch,
  /query/groups), translating request bodies (vector -> query / using,
  positive/negative -> query.recommend, target/context -> query.discover)
  and unwrapping the new result.points response shape.
- Drop equivalence assertions against the now-removed legacy endpoints.

Co-authored-by: Cursor <cursoragent@cursor.com>

* Relax non-empty assertions in migrated recommend/discover tests

The previous migration added `len(...) > 0` assertions to tests that
previously only checked equivalence between the deprecated and new
API. These assertions are too strict because the parametrized
`query_filter` cases legitimately produce empty result sets.

Drop the `> 0` assertion and rely on `request_with_validation` to
verify the response is well-formed and HTTP OK.

Co-authored-by: Cursor <cursoragent@cursor.com>

* Migrate remaining OpenAPI tests off deprecated search endpoints

Tests added to dev after the original migration was written still call
/points/search and /points/recommend/groups through
`request_with_validation`, which resolves the endpoint against the
OpenAPI spec and therefore breaks once the endpoint is not in the spec:

- test_turbo4_storage.py, test_sparse_idf_corpus.py, test_validation.py:
  translate /points/search to /points/query (vector{name,vector} ->
  query + using, result -> result.points).
- test_group.py: drop the /points/recommend/groups half of the
  lookup_from validation test in favour of the query equivalent.

test_sparse_idf_corpus.py's test_query_api_supports_idf_corpus goes
away: with the helper on /points/query every test in the file now
exercises what it asserted.

Also record why test_recommend_group cannot assert on its groups: it
uses every point in the collection as a recommend example, so all of
them are excluded and the result is legitimately empty.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* Regenerate openapi.json without the deprecated search endpoints

Drops the 8 deprecated paths and the request schemas that only they
referenced: Search/Recommend/Discover request (+Batch, +Groups) types
and their exclusive dependencies (NamedVector, NamedSparseVector,
NamedVectorStruct, UsingVector, RecommendExample, ContextExamplePair).

Regenerated output is a strict subset of the previous spec, and every
remaining $ref still resolves.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* Deprecate the search/recommend/discover RPCs in gRPC

The REST counterparts have carried `deprecated: true` since v1.13.3 and
are now gone from the OpenAPI spec, while the gRPC RPCs never got any
deprecation annotation at all. Mark all 8 with `option deprecated = true`
so generated clients warn, and point each doc comment at its `Query`
replacement.

tonic puts `#[deprecated]` on the generated client methods only; the
server trait gets the doc comment alone, so our own `impl` is unaffected.
The RPCs keep serving traffic — this is annotation only.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* Restore the deleted recommend/discover suites on /points/query

The earlier migration deleted these four files outright, but the
query-side tests it left behind are all shallow smoke tests
(`len(result) > 0`, `"points" in result[0]`). The deleted ones carried
invariants with no query-API equivalent anywhere, so deleting them was a
real loss of coverage rather than de-duplication:

- test_recommend.py: default strategy equals average_vector; batch
  results identical to sequential singles across six request shapes;
  best_score with only negatives yields all-negative scores; best_score
  with a single positive orders identically to a nearest query; raw
  vectors as examples equal ids as examples.
- test_discover.py: context-only scores are all <= 0; target-only orders
  identically to a nearest query but scores differently; with a fixed
  context the integer part of the score is stable while the decimal part
  moves, and vice versa with a fixed target; batch equals singles;
  lookup_from by id equals by vector.
- test_multicollection_reco.py: cross-collection lookup_from, plus
  wrong-vector-size, unknown-collection and unknown-vector rejections.
- test_recommendation_multivector.py: the same recommend invariants over
  a max_sim multivector collection, which the query suite never covered.

Only test_recommend_missing_lookup_from_collection_with_raw_vector is
dropped as genuinely redundant — test_query.py's
test_query_missing_lookup_from_collection covers query, query/batch and
prefetch.

Two request-shape differences the translation had to absorb:

- Giving no examples at all is 422 (a RecommendInput validation rule),
  where the legacy API reported 400 from the query itself. A malformed
  example, such as an empty vector, is still 400.
- DiscoverInput requires the `context` key and accepts only an explicit
  null to mean "no context", so target-only discover must spell it out.
  The legacy API let it be omitted.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
2026-07-27 18:09:43 +02:00

642 lines
20 KiB
Python

from math import isclose
import pytest
import requests
from .helpers.collection_setup import basic_collection_setup, drop_collection
from .helpers.helpers import distribution_based_score_fusion, reciprocal_rank_fusion, request_with_validation, \
qdrant_host_headers
from .helpers.settings import QDRANT_HOST
@pytest.fixture(autouse=True, scope="module")
def setup(on_disk_vectors, collection_name):
basic_collection_setup(collection_name=collection_name, on_disk_vectors=on_disk_vectors)
response = request_with_validation(
api="/collections/{collection_name}/index",
method="PUT",
path_params={"collection_name": collection_name},
body={"field_name": "price", "field_schema": "float"},
)
assert response.ok
yield
drop_collection(collection_name=collection_name)
def test_query_validation(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"prefetch": [
{
"query": {
"recommend": {
"positive": [1]
},
}
}
]
},
)
assert not response.ok, response.text
assert response.json()["status"]["error"] == ("Bad request: A query is needed to merge the prefetches. Can't have prefetches without defining a query.")
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"score_threshold": 10,
},
)
assert not response.ok, response.text
assert response.json()["status"]["error"] == ("Bad request: A query is needed to use the score_threshold. Can't have score_threshold without defining a query.")
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"score_threshold": 10,
"query": {
"order_by": {
"key": "price",
}
}
},
)
assert not response.ok, response.text
assert response.json()["status"]["error"] == ("Bad request: Can't use score_threshold with an order_by query.")
# raw query to bypass local validation
response = requests.post(f"{QDRANT_HOST}/collections/{collection_name}/points/query",
headers=qdrant_host_headers(),
json={
"query": {
"recommend": {
"positive": [1]
},
},
"limit": 0,
},
)
assert not response.ok, response.text
assert response.json()["status"]["error"] == ("Validation error in JSON body: [internal.limit: value 0 invalid, must be 1 or larger]")
def root_and_rescored_query(collection_name, query, limit=None, with_payload=None):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": query,
"limit": limit,
"with_payload": with_payload,
},
)
assert response.ok
root_query_result = response.json()["result"]["points"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"prefetch": {
"limit": 1000,
},
"query": query,
"with_payload": with_payload,
},
)
assert response.ok
nested_query_result = response.json()["result"]["points"]
assert root_query_result == nested_query_result
return root_query_result
def test_basic_search(collection_name):
default_query_result = root_and_rescored_query(collection_name, [0.1, 0.2, 0.3, 0.4])
nearest_query_result = root_and_rescored_query(collection_name, {"nearest": [0.1, 0.2, 0.3, 0.4]})
assert default_query_result == nearest_query_result
# Test basic search with huge limit, it must not panic with allocation failure
# See: <https://github.com/qdrant/qdrant/issues/5483>
def test_basic_search_high_limit(collection_name):
default_query_result = root_and_rescored_query(
collection_name,
[0.1, 0.2, 0.3, 0.4],
limit=18446744073709551615, # u64::MAX
)
nearest_query_result = root_and_rescored_query(
collection_name,
{"nearest": [0.1, 0.2, 0.3, 0.4]},
limit=18446744073709551615, # u64::MAX
)
assert default_query_result == nearest_query_result
def test_basic_scroll(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/scroll",
method="POST",
path_params={"collection_name": collection_name},
body={},
)
assert response.ok
scroll_result = response.json()["result"]["points"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"with_payload": True,
},
)
assert response.ok
query_result = response.json()["result"]["points"]
for record, scored_point in zip(scroll_result, query_result):
assert record.get("id") == scored_point.get("id")
assert record.get("payload") == scored_point.get("payload")
def test_basic_scroll_offset(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/scroll",
method="POST",
path_params={"collection_name": collection_name},
body={
"offset": 2, # skip first record, start at id 2
},
)
assert response.ok
scroll_result = response.json()["result"]["points"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"with_payload": True,
"offset": 1, # skip one record
},
)
assert response.ok
query_result = response.json()["result"]["points"]
for record, scored_point in zip(scroll_result, query_result):
assert record.get("id") == scored_point.get("id")
assert record.get("payload") == scored_point.get("payload")
def test_basic_recommend_avg(collection_name):
query_result = root_and_rescored_query(collection_name,
{
"recommend": {"positive": [1, 2, 3, 4], "negative": [3]}, # ids
}
)
assert len(query_result) > 0
def test_basic_recommend_best_score(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": {
"recommend": {
"positive": [1, 2, 3, 4], # ids
"negative": [3], # ids
"strategy": "best_score",
},
}
},
)
assert response.ok
query_result = response.json()["result"]["points"]
assert len(query_result) > 0
def test_basic_discover(collection_name):
query_result = root_and_rescored_query(collection_name,
{
"discover": {
"target": 2,
"context": [{"positive": 3, "negative": 4}],
}
},
)
assert len(query_result) > 0
def test_basic_context(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": {
"context": [{"positive": 2, "negative": 4}],
},
"limit": 100,
},
)
assert response.ok
query_result = response.json()["result"]["points"]
assert len(query_result) > 0
def test_basic_order_by(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/scroll",
method="POST",
path_params={"collection_name": collection_name},
body={
"order_by": "price",
},
)
assert response.ok, response.text
scroll_result = response.json()["result"]["points"]
query_result = root_and_rescored_query(collection_name, {"order_by": "price"}, with_payload=True)
for record, scored_point in zip(scroll_result, query_result):
assert record.get("id") == scored_point.get("id")
assert record.get("payload") == scored_point.get("payload")
def test_basic_random_query(collection_name):
ids_lists = set()
for _ in range(4):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": { "sample": "random" }
},
)
assert response.ok, response.text
points = response.json()["result"]["points"]
assert len(points) == 10
assert set(point["id"] for point in points) == set(range(1, 11))
ids = str([point["id"] for point in points])
ids_lists.add(ids)
# check the order of ids are different at least once
assert len(ids_lists) > 1
def test_basic_rrf(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": [0.1, 0.2, 0.3, 0.4],
"limit": 10,
},
)
assert response.ok
search_result_1 = response.json()["result"]["points"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": [0.5, 0.6, 0.7, 0.8],
"limit": 10,
},
)
assert response.ok
search_result_2 = response.json()["result"]["points"]
rrf_expected = reciprocal_rank_fusion([search_result_1, search_result_2], limit=10)
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"prefetch": [
{ "query": [0.1, 0.2, 0.3, 0.4] },
{ "query": [0.5, 0.6, 0.7, 0.8] },
],
"query": {"fusion": "rrf"},
},
)
assert response.ok, response.json()
rrf_result = response.json()["result"]["points"]
def get_id(x):
return x["id"]
# rrf order is not deterministic with same scores, so we need to sort by id
for expected, result in zip(sorted(rrf_expected, key=get_id), sorted(rrf_result, key=get_id)):
assert expected["id"] == result["id"]
assert expected.get("payload") == result.get("payload")
assert isclose(expected["score"], result["score"], rel_tol=1e-5)
def test_weighted_rrf(collection_name):
"""Test RRF with weights parameter."""
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": [0.1, 0.2, 0.3, 0.4],
"limit": 10,
},
)
assert response.ok
search_result_1 = response.json()["result"]["points"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": [0.5, 0.6, 0.7, 0.8],
"limit": 10,
},
)
assert response.ok
search_result_2 = response.json()["result"]["points"]
# Test with weights [3.0, 1.0] - first source has 3x weight
weights = [3.0, 1.0]
rrf_expected = reciprocal_rank_fusion([search_result_1, search_result_2], limit=10, weights=weights)
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"prefetch": [
{ "query": [0.1, 0.2, 0.3, 0.4] },
{ "query": [0.5, 0.6, 0.7, 0.8] },
],
"query": {
"rrf": {
"weights": weights
}
},
},
)
assert response.ok, response.json()
rrf_result = response.json()["result"]["points"]
def get_id(x):
return x["id"]
# rrf order is not deterministic with same scores, so we need to sort by id
for expected, result in zip(sorted(rrf_expected, key=get_id), sorted(rrf_result, key=get_id)):
assert expected["id"] == result["id"]
assert expected.get("payload") == result.get("payload")
assert isclose(expected["score"], result["score"], rel_tol=1e-5)
# Test with custom k parameter as well
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"prefetch": [
{ "query": [0.1, 0.2, 0.3, 0.4] },
{ "query": [0.5, 0.6, 0.7, 0.8] },
],
"query": {
"rrf": {
"k": 60,
"weights": [80, 20]
}
},
},
)
assert response.ok, response.json()
# Just verify it returns results without error
assert len(response.json()["result"]["points"]) > 0
def test_basic_dbsf(collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": [0.1, 0.2, 0.3, 0.4],
"limit": 10,
},
)
assert response.ok
search_result_1 = response.json()["result"]["points"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": [0.5, 0.6, 0.7, 0.8],
"limit": 10,
},
)
assert response.ok
search_result_2 = response.json()["result"]["points"]
dbsf_expected = distribution_based_score_fusion([search_result_1, search_result_2], limit=10)
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"prefetch": [
{ "query": [0.1, 0.2, 0.3, 0.4] },
{ "query": [0.5, 0.6, 0.7, 0.8] },
],
"query": {"fusion": "dbsf"},
},
)
assert response.ok, response.json()
dbsf_result = response.json()["result"]["points"]
for point, expected in zip(dbsf_result, dbsf_expected):
assert point["id"] == expected["id"]
assert point.get("payload") == expected.get("payload")
assert isclose(point["score"], expected["score"], rel_tol=1e-5)
def test_nearest_with_mmr(collection_name):
# Regular nearest neighbor search
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": {
"nearest": [0.35, 0.08, 0.11, 0.47]
},
},
)
assert response.ok, response.json()
search_result = response.json()["result"]["points"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": {
"nearest": [0.35, 0.08, 0.11, 0.47],
"mmr": {
"diversity": 0.5,
"candidates_limit": 100
}
},
},
)
assert response.ok, response.json()
mmr_result = response.json()["result"]["points"]
# Assert that both results are in different order
search_ids = [point["id"] for point in search_result]
mmr_ids = [point["id"] for point in mmr_result]
assert search_ids != mmr_ids, "MMR should produce different ordering than regular search"
# We didn't request vectors nor payloads
for point in mmr_result:
assert point.get("payload") is None
assert point.get("vector") is None
# Run request with default parameters
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": {
"nearest": [0.35, 0.08, 0.11, 0.47],
"mmr": {}
},
},
)
assert response.ok, response.json()
mmr_default_result = response.json()["result"]["points"]
# Assert that both results are equal
assert mmr_result == mmr_default_result, "MMR with explicit vs implicit defaults should produce the same output"
@pytest.mark.parametrize("body", [
{
"prefetch": [
{ "query": [0.1, 0.2, 0.3, 0.4] },
{ "query": [0.5, 0.6, 0.7, 0.8] },
],
"query": {"fusion": "rrf"},
},
{ "query": [0.1, 0.2, 0.3, 0.4] }
])
def test_score_threshold(body, collection_name):
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
**body
},
)
assert response.ok, response.json()
points = response.json()["result"]["points"]
assert len(points) == 8
score_threshold = points[3]["score"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"score_threshold": score_threshold,
**body
},
)
assert response.ok, response.json()
points = response.json()["result"]["points"]
assert len(points) < 8
for point in points:
assert point["score"] >= score_threshold
def test_query_missing_lookup_from_collection(collection_name):
missing_collection = "missing_lookup_from_collection"
lookup_from = {"collection": missing_collection, "vector": "default"}
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"query": [0.1, 0.2, 0.3, 0.4],
"limit": 3,
"lookup_from": lookup_from,
},
)
assert response.status_code == 404, response.text
assert missing_collection in response.json()["status"]["error"]
response = request_with_validation(
api="/collections/{collection_name}/points/query/batch",
method="POST",
path_params={"collection_name": collection_name},
body={
"searches": [
{
"query": [0.1, 0.2, 0.3, 0.4],
"limit": 3,
"lookup_from": lookup_from,
}
]
},
)
assert response.status_code == 404, response.text
assert missing_collection in response.json()["status"]["error"]
response = request_with_validation(
api="/collections/{collection_name}/points/query",
method="POST",
path_params={"collection_name": collection_name},
body={
"prefetch": [
{
"query": [0.1, 0.2, 0.3, 0.4],
"limit": 3,
"lookup_from": lookup_from,
}
],
"query": {"fusion": "rrf"},
"limit": 3,
},
)
assert response.status_code == 404, response.text
assert missing_collection in response.json()["status"]["error"]