unique_web_search 2026.38.0.dev2__tar.gz → 2026.38.0.dev4__tar.gz

{unique_web_search-2026.38.0.dev2 → unique_web_search-2026.38.0.dev4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: unique-web-search
-Version: 2026.38.0.dev2
+Version: 2026.38.0.dev4
 Summary: 
 Author: Andreas Hauri, Gustav Hartz, Rami Azouz
 Author-email: Andreas Hauri <andreas@unique.ch>, Gustav Hartz <gustav.hartz.ext@unique.ch>, Rami Azouz <rami.ext@unique.ch>
@@ -16,7 +16,7 @@ Requires-Dist: tavily-python>=0.7.11,<1
 Requires-Dist: unidecode>=1.4.0,<2
 Requires-Dist: azure-ai-projects>=2.3,<3
 Requires-Dist: azure-identity>=1.25.0,<2
-Requires-Dist: unique-toolkit[monitoring]>=2026.38.0.dev11,<2026.38.0rc0
+Requires-Dist: unique-toolkit[monitoring]>=2026.38.0.dev18,<2026.38.0rc0
 Requires-Dist: azure-core>=1.36.0,<2
 Requires-Dist: google-cloud-aiplatform>=1.128.0,<2
 Requires-Dist: google-auth>=2.43.0,<3
@@ -25,7 +25,7 @@ Requires-Dist: rapidfuzz>=3.0.0,<4
 Requires-Dist: httpx>=0.28.1
 Requires-Dist: certifi>=2025.11.12
 Requires-Dist: jinja2>=3.1.6
-Requires-Dist: unique-search-proxy-core>=2026.38.0.dev3,<2026.38.0rc0
+Requires-Dist: unique-search-proxy-core>=2026.38.0.dev6,<2026.38.0rc0
 Requires-Dist: unique-search-proxy-sdk>=2026.38.0.dev2,<2026.38.0rc0
 Requires-Dist: openai>=1.105.0,<3
 Requires-Python: >=3.12, <4.0

{unique_web_search-2026.38.0.dev2 → unique_web_search-2026.38.0.dev4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "unique_web_search"
-version = "2026.38.0.dev2"
+version = "2026.38.0.dev4"
 description = ""
 readme = "README.md"
 requires-python = ">=3.12,<4.0"
@@ -22,7 +22,7 @@ dependencies = [
     "unidecode>=1.4.0,<2",
     "azure-ai-projects>=2.3,<3",
     "azure-identity>=1.25.0,<2",
-    "unique-toolkit[monitoring]>=2026.38.0.dev11,<2026.38.0rc0",
+    "unique-toolkit[monitoring]>=2026.38.0.dev18,<2026.38.0rc0",
     "azure-core>=1.36.0,<2",
     "google-cloud-aiplatform>=1.128.0,<2",
     "google-auth>=2.43.0,<3",
@@ -31,7 +31,7 @@ dependencies = [
     "httpx>=0.28.1",
     "certifi>=2025.11.12",
     "jinja2>=3.1.6",
-    "unique-search-proxy-core>=2026.38.0.dev3,<2026.38.0rc0",
+    "unique-search-proxy-core>=2026.38.0.dev6,<2026.38.0rc0",
     "unique-search-proxy-sdk>=2026.38.0.dev2,<2026.38.0rc0",
     "openai>=1.105.0,<3",
 ]

{unique_web_search-2026.38.0.dev2 → unique_web_search-2026.38.0.dev4}/src/unique_web_search/service.py

@@ -80,6 +80,7 @@ class WebSearchTool(Tool[WebSearchConfig]):
             company_id=self.event.company_id,
             user_id=self.event.user_id,
             chat_id=self.event.payload.chat_id,
+            user_metadata=self.event.payload.user_metadata or {},
         )
         self.search_engine_service = get_search_engine_service(
             self.config.search_engine_config,

unique_web_search-2026.38.0.dev4/src/unique_web_search/services/client/http_client.py

@@ -0,0 +1,70 @@
+"""Process-wide HTTP client registry for the web-search tool's direct egress path.
+
+Clients are created lazily on first use via ``client_for`` / the registry — never
+at import time — so per-user proxy configs without a settings username can load.
+"""
+
+from __future__ import annotations
+
+from httpx import AsyncClient
+from pydantic import SecretStr
+from unique_search_proxy_core.context import RequestContext
+from unique_search_proxy_core.http_client import (
+    HttpClientRegistry,
+    ProxySettings,
+    resolver_from_settings,
+)
+
+from unique_web_search.settings import env_settings
+
+_registry: HttpClientRegistry | None = None
+
+
+def proxy_settings_from_env() -> ProxySettings:
+    """Build core ProxySettings from the tool's process env settings."""
+    return ProxySettings(
+        proxy_auth_mode=env_settings.proxy_auth_mode,
+        proxy_protocol=env_settings.proxy_protocol,
+        proxy_host=env_settings.proxy_host,
+        proxy_port=env_settings.proxy_port,
+        proxy_headers={
+            name: SecretStr(value) for name, value in env_settings.proxy_headers.items()
+        },
+        proxy_ssl_ca_bundle_path=env_settings.proxy_ssl_ca_bundle_path,
+        proxy_username=(
+            SecretStr(env_settings.proxy_username)
+            if env_settings.proxy_username is not None
+            else None
+        ),
+        proxy_password=SecretStr(env_settings.proxy_password or ""),
+        proxy_ssl_cert_path=env_settings.proxy_ssl_cert_path,
+        proxy_ssl_key_path=env_settings.proxy_ssl_key_path,
+        proxy_username_source=env_settings.proxy_username_source,
+        proxy_username_metadata_field=env_settings.proxy_username_metadata_field,
+        per_user_proxy_company_ids=env_settings.per_user_proxy_company_ids,
+        http_client_cache_size=env_settings.http_client_cache_size,
+    )
+
+
+def get_http_client_registry() -> HttpClientRegistry:
+    """Return the process-wide registry for direct (non-proxy-service) egress."""
+    global _registry
+    if _registry is None:
+        settings = proxy_settings_from_env()
+        _registry = HttpClientRegistry(
+            settings=settings,
+            resolver=resolver_from_settings(settings),
+        )
+    return _registry
+
+
+async def client_for(context: RequestContext) -> AsyncClient:
+    """Resolve the egress client for a request context."""
+    return await get_http_client_registry().client_for(context)
+
+
+__all__ = [
+    "client_for",
+    "get_http_client_registry",
+    "proxy_settings_from_env",
+]

{unique_web_search-2026.38.0.dev2 → unique_web_search-2026.38.0.dev4}/src/unique_web_search/services/crawlers/basic.py

@@ -18,7 +18,7 @@ from unique_search_proxy_core.crawlers.basic.schema import (
 )
 from unique_toolkit._common.pydantic.rjsf_tags import RJSFMetaTag
 
-from unique_web_search.services.client.proxy_config import async_client
+from unique_web_search.services.client.http_client import client_for
 from unique_web_search.services.crawlers.base import BaseCrawler
 from unique_web_search.services.crawlers.registry import register_crawler
 from unique_web_search.services.crawlers.url_safety import (
@@ -89,29 +89,29 @@ class BasicCrawler(BaseCrawler[BasicConfig]):
 
     @override
     async def _legacy_crawl(self, targets: list[ResolvedCrawlTarget]) -> list[str]:
-        async with async_client(timeout=Timeout(self.config.timeout)) as client:
-            tasks = [self._crawl_url_with_client(client, target) for target in targets]
-            results = await asyncio.gather(*tasks, return_exceptions=True)
-
-            validation_errors = [
-                result
-                for result in results
-                if isinstance(result, CrawlTargetValidationError)
-            ]
-            if validation_errors:
-                raise validation_errors[0]
-
-            markdowns: list[str] = []
-            for result in results:
-                if isinstance(result, BaseException):
-                    markdowns.append(
-                        f"Unexpected error occurred while crawling the URL: {result}"
-                    )
-
-                else:
-                    markdowns.append(result)
-
-            return markdowns
+        client = await client_for(self._request_context)
+        tasks = [self._crawl_url_with_client(client, target) for target in targets]
+        results = await asyncio.gather(*tasks, return_exceptions=True)
+
+        validation_errors = [
+            result
+            for result in results
+            if isinstance(result, CrawlTargetValidationError)
+        ]
+        if validation_errors:
+            raise validation_errors[0]
+
+        markdowns: list[str] = []
+        for result in results:
+            if isinstance(result, BaseException):
+                markdowns.append(
+                    f"Unexpected error occurred while crawling the URL: {result}"
+                )
+
+            else:
+                markdowns.append(result)
+
+        return markdowns
 
     async def _crawl_url_with_client(
         self, client: AsyncClient, target: ResolvedCrawlTarget
@@ -132,6 +132,7 @@ class BasicCrawler(BaseCrawler[BasicConfig]):
             target.request_url,
             headers=request_headers,
             extensions=request_extensions or None,
+            timeout=Timeout(self.config.timeout),
         )
 
         _ = response.raise_for_status()

{unique_web_search-2026.38.0.dev2 → unique_web_search-2026.38.0.dev4}/src/unique_web_search/services/search_engine/google.py

@@ -9,7 +9,7 @@ from unique_search_proxy_core.search_engines.base import SearchEngineType
 from unique_search_proxy_core.search_engines.google.schema import GoogleConfig
 
 from unique_web_search.client_settings import get_google_search_settings
-from unique_web_search.services.client.proxy_config import async_client
+from unique_web_search.services.client.http_client import client_for
 from unique_web_search.services.search_engine.base import SearchEngine, SearchEngineMode
 from unique_web_search.services.search_engine.registry import register_search_engine
 from unique_web_search.services.search_engine.schema import (
@@ -69,8 +69,8 @@ class GoogleSearch(SearchEngine[GoogleConfig]):
             start_index=start_index,
             num_fetch=num_fetch,
         )
-        async with async_client() as client:
-            response = await client.get(**request_params)
+        client = await client_for(self._request_context)
+        response = await client.get(**request_params)
         return response
 
     def _validate_url(self, url: str) -> bool:

{unique_web_search-2026.38.0.dev2 → unique_web_search-2026.38.0.dev4}/src/unique_web_search/settings.py

@@ -93,6 +93,12 @@ class Base(BaseSettings):
     proxy_username: str | None = None
     proxy_password: str | None = None
 
+    ## Username source for per-user corporate proxies
+    proxy_username_source: Literal["settings", "user_metadata"] = "settings"
+    proxy_username_metadata_field: str = "userName"
+    per_user_proxy_company_ids: list[str] = []
+    http_client_cache_size: int = 128
+
     ## For SSL/TLS authentication
     proxy_ssl_cert_path: str | None = None
     proxy_ssl_key_path: str | None = None

unique_web_search-2026.38.0.dev2/src/unique_web_search/services/client/proxy_config.py

@@ -1,119 +0,0 @@
-import logging
-from functools import partial
-
-from httpx import AsyncClient
-from pydantic import BaseModel
-
-from unique_web_search.settings import env_settings
-
-_LOGGER = logging.getLogger(__name__)
-
-
-class ProxyConfig(BaseModel):
-    verify: bool | str
-    proxy: str | None
-    headers: dict[str, str] | None
-    cert: tuple[str, str] | str | None = None
-    trust_env: bool = False
-
-
-def _get_proxy_host_and_port() -> tuple[str, int]:
-    proxy_host = env_settings.proxy_host
-    proxy_port = env_settings.proxy_port
-    assert proxy_host is not None and proxy_port is not None, (
-        "Proxy host and port are required"
-    )
-
-    return proxy_host, proxy_port
-
-
-def _build_proxy_url_with_username_password() -> str:
-    proxy_host, proxy_port = _get_proxy_host_and_port()
-
-    proxy_username = env_settings.proxy_username
-    proxy_password = env_settings.proxy_password
-    assert proxy_username is not None and proxy_password is not None, (
-        "Proxy username and password are required"
-    )
-
-    return f"{env_settings.proxy_protocol}://{proxy_username}:{proxy_password}@{proxy_host}:{proxy_port}"
-
-
-def _build_proxy_url_with_tls() -> str:
-    proxy_host, proxy_port = _get_proxy_host_and_port()
-    return f"{env_settings.proxy_protocol}://{proxy_host}:{proxy_port}"
-
-
-def _get_cert_args() -> tuple[str, str] | str:
-    proxy_ssl_cert_path = env_settings.proxy_ssl_cert_path
-    assert proxy_ssl_cert_path is not None, "Proxy SSL cert path is required"
-
-    proxy_ssl_key_path = env_settings.proxy_ssl_key_path
-
-    if proxy_ssl_key_path is None:
-        _LOGGER.warning(
-            "Proxy SSL key path is not set. Assuming cert path includes key path"
-        )
-        return proxy_ssl_cert_path
-    else:
-        return proxy_ssl_cert_path, proxy_ssl_key_path
-
-
-def _get_none_proxy_kwargs() -> ProxyConfig:
-    _LOGGER.info("Proxy auth mode: none. Using no proxy")
-    return ProxyConfig(
-        proxy=None,
-        headers=None,
-        verify=True,
-        trust_env=True,
-        cert=None,
-    )
-
-
-def _get_username_password_proxy_kwargs() -> ProxyConfig:
-    proxy_url = _build_proxy_url_with_username_password()
-    _LOGGER.info(
-        "Proxy auth mode: username_password. Using proxy with username and password"
-    )
-    return ProxyConfig(
-        proxy=proxy_url,
-        headers=env_settings.proxy_headers,
-        verify=env_settings.proxy_ssl_ca_bundle_path or True,
-    )
-
-
-def _get_ssl_tls_proxy_kwargs() -> ProxyConfig:
-    proxy_url = _build_proxy_url_with_tls()
-    cert_args = _get_cert_args()
-    _LOGGER.info("Proxy auth mode: ssl_tls. Using proxy with SSL/TLS")
-    return ProxyConfig(
-        proxy=proxy_url,
-        cert=cert_args,
-        headers=env_settings.proxy_headers,
-        verify=env_settings.proxy_ssl_ca_bundle_path or True,
-    )
-
-
-def _build_client_kwargs() -> ProxyConfig:
-    match env_settings.proxy_auth_mode:
-        case "none":
-            return _get_none_proxy_kwargs()
-        case "username_password":
-            return _get_username_password_proxy_kwargs()
-        case "ssl_tls":
-            return _get_ssl_tls_proxy_kwargs()
-        case _:
-            raise ValueError(f"Invalid proxy auth mode: {env_settings.proxy_auth_mode}")
-
-
-_client_kwargs = _build_client_kwargs()
-
-
-async_client = partial(
-    AsyncClient,
-    proxy=_client_kwargs.proxy,
-    headers=_client_kwargs.headers,
-    verify=_client_kwargs.verify,
-    trust_env=_client_kwargs.trust_env,
-    cert=_client_kwargs.cert,
-)