mirror of
https://github.com/BerriAI/litellm.git
synced 2026-10-24 05:45:11 +00:00
feat(search): add Keenable as a search provider
This commit is contained in:
parent
c13746c951
commit
d2b55428fd
15 changed files with 555 additions and 30 deletions
3
litellm/llms/keenable/__init__.py
Normal file
3
litellm/llms/keenable/__init__.py
Normal file
|
|
@ -0,0 +1,3 @@
|
|||
from litellm.llms.keenable.search.transformation import KeenableSearchConfig
|
||||
|
||||
__all__ = ("KeenableSearchConfig",)
|
||||
3
litellm/llms/keenable/search/__init__.py
Normal file
3
litellm/llms/keenable/search/__init__.py
Normal file
|
|
@ -0,0 +1,3 @@
|
|||
from litellm.llms.keenable.search.transformation import KeenableSearchConfig
|
||||
|
||||
__all__ = ("KeenableSearchConfig",)
|
||||
270
litellm/llms/keenable/search/transformation.py
Normal file
270
litellm/llms/keenable/search/transformation.py
Normal file
|
|
@ -0,0 +1,270 @@
|
|||
"""
|
||||
Calls Keenable's /v1/search endpoint to search the web.
|
||||
|
||||
Keenable works without an API key: keyless requests go to /v1/search/public, which is rate
|
||||
limited per IP and identifies the calling app by the X-Keenable-Title header. With an API key
|
||||
the same body goes to /v1/search, which has higher limits.
|
||||
|
||||
Keenable API Reference: https://docs.keenable.ai/api-reference/search
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from collections.abc import Mapping
|
||||
from types import MappingProxyType
|
||||
from typing import TYPE_CHECKING, Final
|
||||
|
||||
import httpx
|
||||
from pydantic import ConfigDict, TypeAdapter, ValidationError
|
||||
|
||||
from litellm.llms.base_llm.chat.transformation import BaseLLMException
|
||||
from litellm.llms.base_llm.search.transformation import (
|
||||
BaseSearchConfig,
|
||||
SearchResponse,
|
||||
SearchResult,
|
||||
)
|
||||
from litellm.types.llms.base import LiteLLMBaseModel
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
|
||||
|
||||
_KEENABLE_DOCS_URL: Final = "https://docs.keenable.ai/api-reference/search"
|
||||
|
||||
# Sent on every request. The keyless endpoint rejects a request without it (400), and it
|
||||
# names the calling software only, never the user.
|
||||
_APP_TITLE: Final = "litellm"
|
||||
|
||||
|
||||
class _KeenableResult(LiteLLMBaseModel):
|
||||
"""One entry of Keenable's `results` array. Every field is optional so a single degraded
|
||||
result degrades to empty strings instead of failing the whole call."""
|
||||
|
||||
model_config = ConfigDict(extra="ignore", frozen=True)
|
||||
|
||||
title: str | None = None
|
||||
url: str | None = None
|
||||
# Usually empty. The page text is in `snippet`.
|
||||
description: str | None = None
|
||||
snippet: str | None = None
|
||||
published_at: str | None = None
|
||||
acquired_at: str | None = None
|
||||
|
||||
|
||||
class _KeenableSearchResponse(LiteLLMBaseModel):
|
||||
"""Keenable's /v1/search response envelope."""
|
||||
|
||||
model_config = ConfigDict(extra="ignore", frozen=True)
|
||||
|
||||
# Required: a search with no hits returns `[]`, so a null or absent `results` means the
|
||||
# body is not a search response and must not be reported as a successful empty search.
|
||||
results: tuple[_KeenableResult, ...]
|
||||
|
||||
|
||||
class _ErrorEnvelope(LiteLLMBaseModel):
|
||||
"""Keenable reports errors as `{"error": "Invalid parameter", "message": "..."}`."""
|
||||
|
||||
model_config = ConfigDict(extra="ignore", frozen=True)
|
||||
|
||||
error: str | None = None
|
||||
message: str | None = None
|
||||
|
||||
|
||||
_DomainListAdapter: Final = TypeAdapter(tuple[str, ...])
|
||||
|
||||
_NOTHING: Final[Mapping[str, object]] = MappingProxyType({})
|
||||
|
||||
|
||||
def _optional(key: str, value: object) -> Mapping[str, object]:
|
||||
"""A one-entry mapping to spread into a payload, or nothing when the value is absent."""
|
||||
return MappingProxyType({key: value}) if value is not None else _NOTHING
|
||||
|
||||
|
||||
class KeenableSearchConfig(BaseSearchConfig):
|
||||
KEENABLE_API_BASE = "https://api.keenable.ai/v1"
|
||||
|
||||
@staticmethod
|
||||
def ui_friendly_name() -> str:
|
||||
return "Keenable"
|
||||
|
||||
def _resolve_api_key(self, api_key: str | None, api_base: str | None) -> str | None:
|
||||
"""The caller's key, else KEENABLE_API_KEY, else None for a keyless call."""
|
||||
return self.resolve_server_api_key(
|
||||
caller_api_key=api_key,
|
||||
caller_api_base=api_base,
|
||||
key_env_vars=("KEENABLE_API_KEY",),
|
||||
base_env_var=None,
|
||||
default_api_base=self.KEENABLE_API_BASE,
|
||||
)
|
||||
|
||||
def validate_environment(
|
||||
self,
|
||||
headers: dict[str, str], # mutable-ok: BaseSearchConfig.validate_environment signature
|
||||
api_key: str | None = None,
|
||||
api_base: str | None = None,
|
||||
**kwargs: object, # kwargs-ok: BaseSearchConfig.validate_environment signature
|
||||
) -> dict[str, str]: # mutable-ok: the http handler passes this straight to httpx as headers
|
||||
"""
|
||||
Validate environment and return headers. No API key is required.
|
||||
|
||||
Returns a new dict rather than mutating ``headers``: the http handler calls this
|
||||
a second time after ``litellm/search/main.py`` already did, so it has to be idempotent.
|
||||
"""
|
||||
resolved_api_key: Final = self._resolve_api_key(api_key, api_base)
|
||||
return {
|
||||
**headers,
|
||||
"Content-Type": "application/json",
|
||||
"X-Keenable-Title": _APP_TITLE,
|
||||
**_optional("Authorization", f"Bearer {resolved_api_key}" if resolved_api_key else None),
|
||||
}
|
||||
|
||||
def get_complete_url(
|
||||
self,
|
||||
api_base: str | None,
|
||||
optional_params: dict[str, object], # mutable-ok: BaseSearchConfig.get_complete_url signature
|
||||
data: dict[str, object] | list[dict[str, object]] | None = None, # mutable-ok: base signature
|
||||
**kwargs: object, # kwargs-ok: BaseSearchConfig.get_complete_url signature
|
||||
) -> str:
|
||||
"""
|
||||
The keyed and keyless endpoints take the same body and differ only by path, so the key
|
||||
decides it: /search with a key, /search/public without. ``api_base`` is the version
|
||||
root (https://api.keenable.ai/v1); a base that already ends in either path is accepted.
|
||||
"""
|
||||
caller_api_key: Final = kwargs.get("api_key")
|
||||
resolved_api_key: Final = self._resolve_api_key(
|
||||
caller_api_key if isinstance(caller_api_key, str) else None, api_base
|
||||
)
|
||||
root: Final = (
|
||||
(api_base or self.KEENABLE_API_BASE).rstrip("/").removesuffix("/search/public").removesuffix("/search")
|
||||
)
|
||||
return f"{root}/search" if resolved_api_key else f"{root}/search/public"
|
||||
|
||||
def transform_search_request(
|
||||
self,
|
||||
query: str | list[str], # mutable-ok: BaseSearchConfig.transform_search_request signature
|
||||
optional_params: dict[str, object], # mutable-ok: base signature
|
||||
**kwargs: object, # kwargs-ok: BaseSearchConfig.transform_search_request signature
|
||||
) -> dict[str, object]: # mutable-ok: the http handler passes this straight to httpx as the JSON body
|
||||
"""
|
||||
Transform Search request to Keenable API format.
|
||||
|
||||
- query -> query (a list is joined with spaces; Keenable takes a single string)
|
||||
- max_results -> max_results (sent unclamped so Keenable's own 1-50 validation reports the error)
|
||||
- search_domain_filter -> `site` for a single domain, `site:` / `-site:` clauses in the query
|
||||
otherwise, following the Perplexity unified spec where a `-` prefix excludes a domain
|
||||
- country, max_tokens_per_page -> dropped (no Keenable equivalent)
|
||||
|
||||
Everything else is forwarded as-is, so the rest of Keenable's surface (date filters,
|
||||
snippet_max_length) stays reachable without LiteLLM tracking it.
|
||||
"""
|
||||
unified_params: Final = self.get_supported_perplexity_optional_params()
|
||||
include, exclude = _split_domains(optional_params.get("search_domain_filter"))
|
||||
site: Final = include[0] if len(include) == 1 else None
|
||||
|
||||
# Spread after the derived `site` so an explicitly supplied one wins.
|
||||
passthrough: Final = MappingProxyType(
|
||||
{param: value for param, value in optional_params.items() if param not in unified_params}
|
||||
)
|
||||
|
||||
return {
|
||||
**_optional("site", site),
|
||||
**passthrough,
|
||||
"query": _query_with_domain_clauses(
|
||||
" ".join(query) if isinstance(query, list) else query,
|
||||
include=() if site else include,
|
||||
exclude=exclude,
|
||||
),
|
||||
**_optional("max_results", optional_params.get("max_results")),
|
||||
}
|
||||
|
||||
def transform_search_response(
|
||||
self,
|
||||
raw_response: httpx.Response,
|
||||
logging_obj: LiteLLMLoggingObj,
|
||||
**kwargs: object, # kwargs-ok: BaseSearchConfig.transform_search_response signature
|
||||
) -> SearchResponse:
|
||||
"""
|
||||
Transform Keenable API response to LiteLLM unified SearchResponse format.
|
||||
|
||||
`snippet` carries the page text and `description` is usually empty, so the snippet comes
|
||||
first. `date` is the page's `published_at` and `last_updated` is `acquired_at`, when
|
||||
Keenable last fetched the page. A body that does not match the documented schema raises
|
||||
an attributed error rather than being reported as a successful empty search.
|
||||
"""
|
||||
try:
|
||||
parsed: Final = _KeenableSearchResponse.model_validate_json(raw_response.content)
|
||||
except ValidationError as e:
|
||||
raise self.get_error_class(
|
||||
error_message=f"response does not match the documented /v1/search schema: {e}",
|
||||
status_code=raw_response.status_code,
|
||||
headers=dict(raw_response.headers),
|
||||
)
|
||||
|
||||
return SearchResponse(
|
||||
results=[
|
||||
SearchResult(
|
||||
title=result.title or "",
|
||||
url=result.url or "",
|
||||
snippet=result.snippet or result.description or "",
|
||||
date=result.published_at,
|
||||
last_updated=result.acquired_at,
|
||||
)
|
||||
for result in parsed.results
|
||||
],
|
||||
object="search",
|
||||
)
|
||||
|
||||
def get_error_class(
|
||||
self,
|
||||
error_message: str,
|
||||
status_code: int,
|
||||
headers: dict[str, str], # mutable-ok: BaseSearchConfig.get_error_class signature
|
||||
) -> Exception:
|
||||
detail: Final = _unwrap_error_detail(error_message).rstrip(". ")
|
||||
# Keyless calls share one per-IP budget, which a proxy serving many users exhausts first.
|
||||
hint: Final = (
|
||||
" Without an API key, requests share a per-IP limit; set KEENABLE_API_KEY to raise it."
|
||||
if status_code == 429
|
||||
else ""
|
||||
)
|
||||
return BaseLLMException(
|
||||
status_code=status_code,
|
||||
message=f"Keenable Search: {detail}.{hint} See {_KEENABLE_DOCS_URL} for details.",
|
||||
headers=headers,
|
||||
)
|
||||
|
||||
|
||||
def _unwrap_error_detail(error_message: str) -> str:
|
||||
"""
|
||||
Surface the human-readable message inside Keenable's error envelope.
|
||||
|
||||
Falls back to the raw body for anything else (CDN HTML pages, plain text, other shapes).
|
||||
"""
|
||||
try:
|
||||
body: Final = _ErrorEnvelope.model_validate_json(error_message)
|
||||
except ValidationError:
|
||||
return error_message
|
||||
return body.message or body.error or error_message
|
||||
|
||||
|
||||
def _split_domains(search_domain_filter: object) -> tuple[tuple[str, ...], tuple[str, ...]]:
|
||||
"""
|
||||
Split the unified `search_domain_filter` into included and excluded domains.
|
||||
|
||||
Anything that is not a list of strings is ignored rather than raising, since it only
|
||||
ever narrows a search that is otherwise valid.
|
||||
"""
|
||||
try:
|
||||
domains: Final = _DomainListAdapter.validate_python(search_domain_filter)
|
||||
except ValidationError:
|
||||
return (), ()
|
||||
return (
|
||||
tuple(d for d in domains if d and not d.startswith("-")),
|
||||
tuple(d[1:] for d in domains if d.startswith("-") and len(d) > 1),
|
||||
)
|
||||
|
||||
|
||||
def _query_with_domain_clauses(query: str, include: tuple[str, ...], exclude: tuple[str, ...]) -> str:
|
||||
"""Append `(site:a OR site:b)` and `-site:c` clauses, which Keenable honors in the query."""
|
||||
included: Final = (f"({' OR '.join(f'site:{d}' for d in include)})",) if include else ()
|
||||
excluded: Final = tuple(f"-site:{d}" for d in exclude)
|
||||
return " ".join((query, *included, *excluded))
|
||||
|
|
@ -22865,6 +22865,14 @@
|
|||
"notes": "Nimble Search API pay-as-you-go list price: $5 per 1,000 searches, up to 100 results per search. Volume plans price differently."
|
||||
}
|
||||
},
|
||||
"keenable/search": {
|
||||
"input_cost_per_query": 0.004,
|
||||
"litellm_provider": "keenable",
|
||||
"mode": "search",
|
||||
"metadata": {
|
||||
"notes": "Keenable pay-as-you-go list price: $4 per 1,000 requests, after a free allowance of 100,000 requests a month per organization. Keyless requests to /v1/search/public are free and rate limited per IP."
|
||||
}
|
||||
},
|
||||
"elevenlabs/scribe_v1": {
|
||||
"input_cost_per_second": 6.11e-05,
|
||||
"litellm_provider": "elevenlabs",
|
||||
|
|
|
|||
|
|
@ -4224,6 +4224,7 @@ class SearchProviders(str, Enum):
|
|||
DATAFORSEO = "dataforseo"
|
||||
FIRECRAWL = "firecrawl"
|
||||
FASTCRW = "fastcrw"
|
||||
KEENABLE = "keenable"
|
||||
SEARXNG = "searxng"
|
||||
LINKUP = "linkup"
|
||||
DUCKDUCKGO = "duckduckgo"
|
||||
|
|
|
|||
|
|
@ -9823,6 +9823,7 @@ class ProviderConfigManager:
|
|||
from litellm.llms.fastcrw.search.transformation import FastCRWSearchConfig
|
||||
from litellm.llms.firecrawl.search.transformation import FirecrawlSearchConfig
|
||||
from litellm.llms.google_pse.search.transformation import GooglePSESearchConfig
|
||||
from litellm.llms.keenable.search.transformation import KeenableSearchConfig
|
||||
from litellm.llms.linkup.search.transformation import LinkupSearchConfig
|
||||
from litellm.llms.nimble.search.transformation import NimbleSearchConfig
|
||||
from litellm.llms.parallel_ai.search.transformation import (
|
||||
|
|
@ -9846,6 +9847,7 @@ class ProviderConfigManager:
|
|||
SearchProviders.DATAFORSEO: DataForSEOSearchConfig,
|
||||
SearchProviders.FIRECRAWL: FirecrawlSearchConfig,
|
||||
SearchProviders.FASTCRW: FastCRWSearchConfig,
|
||||
SearchProviders.KEENABLE: KeenableSearchConfig,
|
||||
SearchProviders.SEARXNG: SearXNGSearchConfig,
|
||||
SearchProviders.LINKUP: LinkupSearchConfig,
|
||||
SearchProviders.DUCKDUCKGO: DuckDuckGoSearchConfig,
|
||||
|
|
|
|||
|
|
@ -22865,6 +22865,14 @@
|
|||
"notes": "Nimble Search API pay-as-you-go list price: $5 per 1,000 searches, up to 100 results per search. Volume plans price differently."
|
||||
}
|
||||
},
|
||||
"keenable/search": {
|
||||
"input_cost_per_query": 0.004,
|
||||
"litellm_provider": "keenable",
|
||||
"mode": "search",
|
||||
"metadata": {
|
||||
"notes": "Keenable pay-as-you-go list price: $4 per 1,000 requests, after a free allowance of 100,000 requests a month per organization. Keyless requests to /v1/search/public are free and rate limited per IP."
|
||||
}
|
||||
},
|
||||
"elevenlabs/scribe_v1": {
|
||||
"input_cost_per_second": 6.11e-05,
|
||||
"litellm_provider": "elevenlabs",
|
||||
|
|
|
|||
|
|
@ -2619,6 +2619,13 @@
|
|||
"search": true
|
||||
}
|
||||
},
|
||||
"keenable": {
|
||||
"display_name": "Keenable (`keenable`)",
|
||||
"url": "https://docs.keenable.ai/api-reference/search",
|
||||
"endpoints": {
|
||||
"search": true
|
||||
}
|
||||
},
|
||||
"triton": {
|
||||
"display_name": "Triton (`triton`)",
|
||||
"url": "https://docs.litellm.ai/docs/providers/triton-inference-server",
|
||||
|
|
|
|||
|
|
@ -13,6 +13,7 @@ SEARCH_PROVIDERS = [
|
|||
"brave",
|
||||
"firecrawl",
|
||||
"fastcrw",
|
||||
"keenable",
|
||||
"searxng",
|
||||
"linkup",
|
||||
"duckduckgo",
|
||||
|
|
|
|||
|
|
@ -25,6 +25,7 @@ from litellm.llms.brave.search.transformation import BraveSearchConfig
|
|||
from litellm.llms.dataforseo.search.transformation import DataForSEOSearchConfig
|
||||
from litellm.llms.exa_ai.search.transformation import ExaAISearchConfig
|
||||
from litellm.llms.fastcrw.search.transformation import FastCRWSearchConfig
|
||||
from litellm.llms.keenable.search.transformation import KeenableSearchConfig
|
||||
from litellm.llms.firecrawl.search.transformation import FirecrawlSearchConfig
|
||||
from litellm.llms.google_pse.search.transformation import GooglePSESearchConfig
|
||||
from litellm.llms.linkup.search.transformation import LinkupSearchConfig
|
||||
|
|
@ -100,6 +101,7 @@ PROVIDERS: Tuple[ProviderSpec, ...] = (
|
|||
),
|
||||
(TinyfishSearchConfig, {"TINYFISH_API_KEY": "srv"}, "caller-key", {}),
|
||||
(FastCRWSearchConfig, {"CRW_API_KEY": "srv"}, "caller-key", {}),
|
||||
(KeenableSearchConfig, {"KEENABLE_API_KEY": "srv"}, "caller-key", {}),
|
||||
(NimbleSearchConfig, {"NIMBLE_API_KEY": "srv"}, "caller-key", {}),
|
||||
(BingGroundingSearchConfig, {"BING_GROUNDING_TOKEN": "srv"}, "caller-key", {}),
|
||||
)
|
||||
|
|
@ -107,9 +109,7 @@ PROVIDERS: Tuple[ProviderSpec, ...] = (
|
|||
_IDS = tuple(spec[0].__name__ for spec in PROVIDERS)
|
||||
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
"config_cls, server_env, caller_key, extra_env", PROVIDERS, ids=_IDS
|
||||
)
|
||||
@pytest.mark.parametrize("config_cls, server_env, caller_key, extra_env", PROVIDERS, ids=_IDS)
|
||||
def test_server_secret_refused_for_caller_api_base(
|
||||
config_cls: Type[BaseSearchConfig],
|
||||
server_env: Dict[str, str],
|
||||
|
|
@ -124,9 +124,7 @@ def test_server_secret_refused_for_caller_api_base(
|
|||
config_cls().validate_environment(headers={}, api_base=ATTACKER_BASE)
|
||||
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
"config_cls, server_env, caller_key, extra_env", PROVIDERS, ids=_IDS
|
||||
)
|
||||
@pytest.mark.parametrize("config_cls, server_env, caller_key, extra_env", PROVIDERS, ids=_IDS)
|
||||
def test_caller_supplied_key_is_honored_for_custom_api_base(
|
||||
config_cls: Type[BaseSearchConfig],
|
||||
server_env: Dict[str, str],
|
||||
|
|
@ -139,14 +137,10 @@ def test_caller_supplied_key_is_honored_for_custom_api_base(
|
|||
|
||||
# An explicit caller key is the caller's own credential, so pointing it at
|
||||
# the caller's own host must be allowed.
|
||||
config_cls().validate_environment(
|
||||
headers={}, api_key=caller_key, api_base=ATTACKER_BASE
|
||||
)
|
||||
config_cls().validate_environment(headers={}, api_key=caller_key, api_base=ATTACKER_BASE)
|
||||
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
"config_cls, server_env, caller_key, extra_env", PROVIDERS, ids=_IDS
|
||||
)
|
||||
@pytest.mark.parametrize("config_cls, server_env, caller_key, extra_env", PROVIDERS, ids=_IDS)
|
||||
def test_server_secret_used_without_caller_api_base(
|
||||
config_cls: Type[BaseSearchConfig],
|
||||
server_env: Dict[str, str],
|
||||
|
|
@ -167,9 +161,7 @@ def test_keyless_provider_allows_caller_api_base(
|
|||
) -> None:
|
||||
monkeypatch.delenv("SEARXNG_API_KEY", raising=False)
|
||||
|
||||
headers = SearXNGSearchConfig().validate_environment(
|
||||
headers={}, api_base="https://my-searxng.internal"
|
||||
)
|
||||
headers = SearXNGSearchConfig().validate_environment(headers={}, api_base="https://my-searxng.internal")
|
||||
|
||||
assert "Authorization" not in headers
|
||||
|
||||
|
|
@ -182,9 +174,7 @@ def test_operator_env_base_override_is_trusted(
|
|||
|
||||
# Mirrors the second validate_environment call in the search handler, which
|
||||
# receives the already-resolved operator base as api_base.
|
||||
headers = SerperSearchConfig().validate_environment(
|
||||
headers={}, api_base="https://serper.internal.corp/search"
|
||||
)
|
||||
headers = SerperSearchConfig().validate_environment(headers={}, api_base="https://serper.internal.corp/search")
|
||||
|
||||
assert headers["X-API-KEY"] == "srv"
|
||||
|
||||
|
|
@ -200,9 +190,7 @@ class TestResolveServerApiKey:
|
|||
)
|
||||
assert result == "mine"
|
||||
|
||||
def test_returns_none_when_no_server_secret(
|
||||
self, monkeypatch: pytest.MonkeyPatch
|
||||
) -> None:
|
||||
def test_returns_none_when_no_server_secret(self, monkeypatch: pytest.MonkeyPatch) -> None:
|
||||
monkeypatch.delenv("SEARXNG_API_KEY", raising=False)
|
||||
result = BaseSearchConfig().resolve_server_api_key(
|
||||
caller_api_key=None,
|
||||
|
|
@ -228,14 +216,10 @@ class TestResolveServerApiKey:
|
|||
|
||||
class TestIsTrustedSearchApiBase:
|
||||
def test_matches_default_host(self) -> None:
|
||||
assert _is_trusted_search_api_base(
|
||||
"https://google.serper.dev/search", "https://google.serper.dev", None
|
||||
)
|
||||
assert _is_trusted_search_api_base("https://google.serper.dev/search", "https://google.serper.dev", None)
|
||||
|
||||
def test_foreign_host_untrusted(self) -> None:
|
||||
assert not _is_trusted_search_api_base(
|
||||
ATTACKER_BASE, "https://google.serper.dev", None
|
||||
)
|
||||
assert not _is_trusted_search_api_base(ATTACKER_BASE, "https://google.serper.dev", None)
|
||||
|
||||
def test_env_override_host_trusted(self, monkeypatch: pytest.MonkeyPatch) -> None:
|
||||
monkeypatch.setenv("SERPER_API_BASE", "https://serper.internal.corp")
|
||||
|
|
@ -248,9 +232,7 @@ class TestIsTrustedSearchApiBase:
|
|||
def test_schemeless_candidate_untrusted(self) -> None:
|
||||
# Without a scheme urlsplit puts the value in the path, leaving an empty
|
||||
# netloc; an unparseable host must never be treated as trusted.
|
||||
assert not _is_trusted_search_api_base(
|
||||
"attacker.example.com", "https://google.serper.dev", None
|
||||
)
|
||||
assert not _is_trusted_search_api_base("attacker.example.com", "https://google.serper.dev", None)
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
|
|
|
|||
0
tests/unit/llms/keenable/__init__.py
Normal file
0
tests/unit/llms/keenable/__init__.py
Normal file
0
tests/unit/llms/keenable/search/__init__.py
Normal file
0
tests/unit/llms/keenable/search/__init__.py
Normal file
|
|
@ -0,0 +1,238 @@
|
|||
import json
|
||||
from unittest.mock import Mock, patch
|
||||
|
||||
import pytest
|
||||
|
||||
import litellm
|
||||
from litellm.llms.base_llm.chat.transformation import BaseLLMException
|
||||
from litellm.llms.keenable.search.transformation import KeenableSearchConfig
|
||||
|
||||
DEFAULT_ROOT = "https://api.keenable.ai/v1"
|
||||
|
||||
|
||||
@pytest.fixture(autouse=True)
|
||||
def _no_server_key(monkeypatch: pytest.MonkeyPatch):
|
||||
monkeypatch.delenv("KEENABLE_API_KEY", raising=False)
|
||||
|
||||
|
||||
def _config() -> KeenableSearchConfig:
|
||||
return KeenableSearchConfig()
|
||||
|
||||
|
||||
def _resp(payload, status_code: int = 200):
|
||||
r = Mock()
|
||||
r.status_code = status_code
|
||||
r.headers = {}
|
||||
r.content = (payload if isinstance(payload, str) else json.dumps(payload)).encode()
|
||||
return r
|
||||
|
||||
|
||||
def _result(**overrides):
|
||||
base = {
|
||||
"title": "Test Title",
|
||||
"url": "https://example.com",
|
||||
"description": "",
|
||||
"snippet": "Test page text",
|
||||
"published_at": "2026-01-15T10:30:00Z",
|
||||
"acquired_at": "2026-01-16T08:12:34Z",
|
||||
}
|
||||
return {**base, **overrides}
|
||||
|
||||
|
||||
def test_ui_friendly_name():
|
||||
assert _config().ui_friendly_name() == "Keenable"
|
||||
|
||||
|
||||
def test_keyless_request_names_the_app_and_targets_the_public_endpoint():
|
||||
headers = _config().validate_environment({})
|
||||
|
||||
assert "Authorization" not in headers
|
||||
assert headers["X-Keenable-Title"] == "litellm"
|
||||
assert headers["Content-Type"] == "application/json"
|
||||
assert _config().get_complete_url(None, {}) == f"{DEFAULT_ROOT}/search/public"
|
||||
|
||||
|
||||
def test_caller_key_authenticates_and_targets_the_keyed_endpoint():
|
||||
headers = _config().validate_environment({}, api_key="caller-key")
|
||||
|
||||
assert headers["Authorization"] == "Bearer caller-key"
|
||||
assert headers["X-Keenable-Title"] == "litellm"
|
||||
assert _config().get_complete_url(None, {}, api_key="caller-key") == f"{DEFAULT_ROOT}/search"
|
||||
|
||||
|
||||
def test_server_key_authenticates_and_targets_the_keyed_endpoint(monkeypatch: pytest.MonkeyPatch):
|
||||
monkeypatch.setenv("KEENABLE_API_KEY", "env-key")
|
||||
|
||||
assert _config().validate_environment({})["Authorization"] == "Bearer env-key"
|
||||
assert _config().get_complete_url(None, {}) == f"{DEFAULT_ROOT}/search"
|
||||
|
||||
|
||||
def test_validate_environment_does_not_mutate_and_is_idempotent():
|
||||
"""The http handler re-runs validate_environment after search/main.py already did."""
|
||||
config = _config()
|
||||
caller_headers = {"X-Custom": "keep-me"}
|
||||
|
||||
once = config.validate_environment(caller_headers, api_key="k")
|
||||
twice = config.validate_environment(once, api_key="k")
|
||||
|
||||
assert caller_headers == {"X-Custom": "keep-me"}
|
||||
assert once == twice
|
||||
assert once["X-Custom"] == "keep-me"
|
||||
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
"api_base",
|
||||
[
|
||||
"https://self-hosted.local/v1",
|
||||
"https://self-hosted.local/v1/",
|
||||
"https://self-hosted.local/v1/search",
|
||||
"https://self-hosted.local/v1/search/public",
|
||||
"https://self-hosted.local/v1/search/public/",
|
||||
],
|
||||
)
|
||||
@pytest.mark.parametrize(
|
||||
"api_key, path",
|
||||
[(None, "/search/public"), ("caller-key", "/search")],
|
||||
)
|
||||
def test_get_complete_url_is_idempotent_and_the_key_picks_the_path(api_base: str, api_key, path: str):
|
||||
# search/main.py builds the URL, then the http handler calls get_complete_url again with
|
||||
# that URL as api_base, so feeding the result back in must not change it.
|
||||
url = _config().get_complete_url(api_base, {}, api_key=api_key)
|
||||
|
||||
assert url == f"https://self-hosted.local/v1{path}"
|
||||
assert _config().get_complete_url(url, {}, api_key=api_key) == url
|
||||
|
||||
|
||||
def test_request_maps_query_and_max_results_and_forwards_provider_params():
|
||||
body = _config().transform_search_request(
|
||||
["latest", "rust release"],
|
||||
{
|
||||
"max_results": 7,
|
||||
"country": "US",
|
||||
"max_tokens_per_page": 512,
|
||||
"published_after": "7d",
|
||||
"snippet_max_length": 500,
|
||||
},
|
||||
)
|
||||
|
||||
assert body == {
|
||||
"query": "latest rust release",
|
||||
"max_results": 7,
|
||||
"published_after": "7d",
|
||||
"snippet_max_length": 500,
|
||||
}
|
||||
|
||||
|
||||
def test_single_domain_filter_uses_the_site_field():
|
||||
body = _config().transform_search_request("asyncio", {"search_domain_filter": ["docs.python.org"]})
|
||||
|
||||
assert body == {"query": "asyncio", "site": "docs.python.org"}
|
||||
|
||||
|
||||
def test_several_domains_and_exclusions_become_query_clauses():
|
||||
body = _config().transform_search_request(
|
||||
"asyncio",
|
||||
{"search_domain_filter": ["realpython.com", "docs.python.org", "-medium.com", "-"]},
|
||||
)
|
||||
|
||||
assert body == {"query": "asyncio (site:realpython.com OR site:docs.python.org) -site:medium.com"}
|
||||
|
||||
|
||||
def test_single_domain_with_exclusion_combines_site_and_clause():
|
||||
body = _config().transform_search_request("asyncio", {"search_domain_filter": ["docs.python.org", "-medium.com"]})
|
||||
|
||||
assert body == {"query": "asyncio -site:medium.com", "site": "docs.python.org"}
|
||||
|
||||
|
||||
def test_explicit_site_wins_over_the_derived_one():
|
||||
body = _config().transform_search_request(
|
||||
"asyncio", {"search_domain_filter": ["docs.python.org"], "site": "peps.python.org"}
|
||||
)
|
||||
|
||||
assert body["site"] == "peps.python.org"
|
||||
|
||||
|
||||
def test_malformed_domain_filter_is_ignored():
|
||||
body = _config().transform_search_request("asyncio", {"search_domain_filter": "docs.python.org"})
|
||||
|
||||
assert body == {"query": "asyncio"}
|
||||
|
||||
|
||||
def test_response_maps_snippet_and_dates():
|
||||
response = _config().transform_search_response(
|
||||
_resp({"query": "q", "results": [_result(), _result(snippet="", description="Only a description")]}),
|
||||
logging_obj=Mock(),
|
||||
)
|
||||
|
||||
first, second = response.results
|
||||
assert (first.title, first.url, first.snippet) == ("Test Title", "https://example.com", "Test page text")
|
||||
assert first.date == "2026-01-15T10:30:00Z"
|
||||
assert first.last_updated == "2026-01-16T08:12:34Z"
|
||||
assert second.snippet == "Only a description"
|
||||
|
||||
|
||||
def test_degraded_result_maps_to_empty_strings():
|
||||
response = _config().transform_search_response(_resp({"results": [{}]}), logging_obj=Mock())
|
||||
|
||||
only = response.results[0]
|
||||
assert (only.title, only.url, only.snippet, only.date, only.last_updated) == ("", "", "", None, None)
|
||||
|
||||
|
||||
def test_empty_result_list_is_a_successful_search():
|
||||
assert _config().transform_search_response(_resp({"results": []}), logging_obj=Mock()).results == []
|
||||
|
||||
|
||||
@pytest.mark.parametrize("payload", [{}, {"results": None}, "<html>bad gateway</html>"])
|
||||
def test_body_that_is_not_a_search_response_raises(payload):
|
||||
with pytest.raises(BaseLLMException, match="Keenable Search: response does not match"):
|
||||
_config().transform_search_response(_resp(payload, status_code=200), logging_obj=Mock())
|
||||
|
||||
|
||||
def test_error_class_surfaces_the_api_message():
|
||||
error = _config().get_error_class(
|
||||
'{"error": "Invalid parameter", "message": "\\"max_results\\": must be an integer between 1 and 50"}',
|
||||
400,
|
||||
{},
|
||||
)
|
||||
|
||||
assert error.status_code == 400
|
||||
assert error.message.startswith('Keenable Search: "max_results": must be an integer between 1 and 50.')
|
||||
assert "KEENABLE_API_KEY" not in error.message
|
||||
|
||||
|
||||
def test_rate_limit_error_points_at_the_api_key():
|
||||
error = _config().get_error_class(
|
||||
'{"error": "Rate limit exceeded", "message": "Public API hourly limit reached"}', 429, {}
|
||||
)
|
||||
|
||||
assert error.status_code == 429
|
||||
assert "Public API hourly limit reached" in error.message
|
||||
assert "set KEENABLE_API_KEY" in error.message
|
||||
|
||||
|
||||
def test_non_json_error_body_is_kept():
|
||||
assert "upstream timeout" in _config().get_error_class("upstream timeout", 502, {}).message
|
||||
|
||||
|
||||
@pytest.mark.parametrize(
|
||||
"api_key, path, auth",
|
||||
[(None, "/search/public", None), ("caller-key", "/search", "Bearer caller-key")],
|
||||
)
|
||||
def test_litellm_search_end_to_end(api_key, path: str, auth):
|
||||
with patch(
|
||||
"litellm.llms.custom_httpx.http_handler.HTTPHandler.post",
|
||||
return_value=_resp({"query": "q", "results": [_result()]}),
|
||||
) as mock_post:
|
||||
response = litellm.search(
|
||||
query="rust release",
|
||||
search_provider="keenable",
|
||||
api_key=api_key,
|
||||
max_results=3,
|
||||
)
|
||||
|
||||
call = mock_post.call_args.kwargs
|
||||
assert call["url"] == f"{DEFAULT_ROOT}{path}"
|
||||
assert call["headers"]["X-Keenable-Title"] == "litellm"
|
||||
assert call["headers"].get("Authorization") == auth
|
||||
assert call["json"] == {"query": "rust release", "max_results": 3}
|
||||
assert response.results[0].snippet == "Test page text"
|
||||
BIN
ui/litellm-dashboard/public/assets/logos/keenable.png
Normal file
BIN
ui/litellm-dashboard/public/assets/logos/keenable.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 6.3 KiB |
|
|
@ -31,6 +31,7 @@ import bingLogo from "../../../../../public/assets/logos/bing.png";
|
|||
import dataforseoLogo from "../../../../../public/assets/logos/dataforseo.png";
|
||||
import exaAiLogo from "../../../../../public/assets/logos/exa_ai.png";
|
||||
import googlePseLogo from "../../../../../public/assets/logos/google_pse.png";
|
||||
import keenableLogo from "../../../../../public/assets/logos/keenable.png";
|
||||
import nimbleLogo from "../../../../../public/assets/logos/nimble.png";
|
||||
import parallelAiLogo from "../../../../../public/assets/logos/parallel_ai.png";
|
||||
import perplexityLogo from "../../../../../public/assets/logos/perplexity.png";
|
||||
|
|
@ -44,6 +45,7 @@ const searchProviderLogoMap: Record<string, string> = {
|
|||
exa_ai: exaAiLogo.src,
|
||||
google_pse: googlePseLogo.src,
|
||||
dataforseo: dataforseoLogo.src,
|
||||
keenable: keenableLogo.src,
|
||||
nimble: nimbleLogo.src,
|
||||
bing_grounding: bingLogo.src,
|
||||
};
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue