Sync from v0.13

vllm/connections.py

@@ -0,0 +1,189 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+
+from collections.abc import Mapping, MutableMapping
+from pathlib import Path
+from urllib.parse import urlparse
+
+import aiohttp
+import requests
+
+from vllm.version import __version__ as VLLM_VERSION
+
+
+class HTTPConnection:
+    """Helper class to send HTTP requests."""
+
+    def __init__(self, *, reuse_client: bool = True) -> None:
+        super().__init__()
+
+        self.reuse_client = reuse_client
+
+        self._sync_client: requests.Session | None = None
+        self._async_client: aiohttp.ClientSession | None = None
+
+    def get_sync_client(self) -> requests.Session:
+        if self._sync_client is None or not self.reuse_client:
+            self._sync_client = requests.Session()
+
+        return self._sync_client
+
+    # NOTE: We intentionally use an async function even though it is not
+    # required, so that the client is only accessible inside async event loop
+    async def get_async_client(self) -> aiohttp.ClientSession:
+        if self._async_client is None or not self.reuse_client:
+            self._async_client = aiohttp.ClientSession(trust_env=True)
+
+        return self._async_client
+
+    def _validate_http_url(self, url: str):
+        parsed_url = urlparse(url)
+
+        if parsed_url.scheme not in ("http", "https"):
+            raise ValueError(
+                "Invalid HTTP URL: A valid HTTP URL must have scheme 'http' or 'https'."
+            )
+
+    def _headers(self, **extras: str) -> MutableMapping[str, str]:
+        return {"User-Agent": f"vLLM/{VLLM_VERSION}", **extras}
+
+    def get_response(
+        self,
+        url: str,
+        *,
+        stream: bool = False,
+        timeout: float | None = None,
+        extra_headers: Mapping[str, str] | None = None,
+        allow_redirects: bool = True,
+    ):
+        self._validate_http_url(url)
+
+        client = self.get_sync_client()
+        extra_headers = extra_headers or {}
+
+        return client.get(
+            url,
+            headers=self._headers(**extra_headers),
+            stream=stream,
+            timeout=timeout,
+            allow_redirects=allow_redirects,
+        )
+
+    async def get_async_response(
+        self,
+        url: str,
+        *,
+        timeout: float | None = None,
+        extra_headers: Mapping[str, str] | None = None,
+        allow_redirects: bool = True,
+    ):
+        self._validate_http_url(url)
+
+        client = await self.get_async_client()
+        extra_headers = extra_headers or {}
+
+        return client.get(
+            url,
+            headers=self._headers(**extra_headers),
+            timeout=timeout,
+            allow_redirects=allow_redirects,
+        )
+
+    def get_bytes(
+        self, url: str, *, timeout: float | None = None, allow_redirects: bool = True
+    ) -> bytes:
+        with self.get_response(
+            url, timeout=timeout, allow_redirects=allow_redirects
+        ) as r:
+            r.raise_for_status()
+
+            return r.content
+
+    async def async_get_bytes(
+        self,
+        url: str,
+        *,
+        timeout: float | None = None,
+        allow_redirects: bool = True,
+    ) -> bytes:
+        async with await self.get_async_response(
+            url, timeout=timeout, allow_redirects=allow_redirects
+        ) as r:
+            r.raise_for_status()
+
+            return await r.read()
+
+    def get_text(self, url: str, *, timeout: float | None = None) -> str:
+        with self.get_response(url, timeout=timeout) as r:
+            r.raise_for_status()
+
+            return r.text
+
+    async def async_get_text(
+        self,
+        url: str,
+        *,
+        timeout: float | None = None,
+    ) -> str:
+        async with await self.get_async_response(url, timeout=timeout) as r:
+            r.raise_for_status()
+
+            return await r.text()
+
+    def get_json(self, url: str, *, timeout: float | None = None) -> str:
+        with self.get_response(url, timeout=timeout) as r:
+            r.raise_for_status()
+
+            return r.json()
+
+    async def async_get_json(
+        self,
+        url: str,
+        *,
+        timeout: float | None = None,
+    ) -> str:
+        async with await self.get_async_response(url, timeout=timeout) as r:
+            r.raise_for_status()
+
+            return await r.json()
+
+    def download_file(
+        self,
+        url: str,
+        save_path: Path,
+        *,
+        timeout: float | None = None,
+        chunk_size: int = 128,
+    ) -> Path:
+        with self.get_response(url, timeout=timeout) as r:
+            r.raise_for_status()
+
+            with save_path.open("wb") as f:
+                for chunk in r.iter_content(chunk_size):
+                    f.write(chunk)
+
+        return save_path
+
+    async def async_download_file(
+        self,
+        url: str,
+        save_path: Path,
+        *,
+        timeout: float | None = None,
+        chunk_size: int = 128,
+    ) -> Path:
+        async with await self.get_async_response(url, timeout=timeout) as r:
+            r.raise_for_status()
+
+            with save_path.open("wb") as f:
+                async for chunk in r.content.iter_chunked(chunk_size):
+                    f.write(chunk)
+
+        return save_path
+
+
+global_http_connection = HTTPConnection()
+"""
+The global [`HTTPConnection`][vllm.connections.HTTPConnection] instance used
+by vLLM.
+"""