hishel 1.0.0.dev2__py3-none-any.whl → 1.0.0.dev3__py3-none-any.whl

hishel/_sync_httpx.py

@@ -0,0 +1,243 @@
+from __future__ import annotations
+
+import ssl
+import typing as t
+from typing import (
+    Iterable,
+    Iterator,
+    Iterator,
+    Union,
+    cast,
+    overload,
+)
+
+from httpx import RequestNotRead
+
+from hishel import SyncCacheProxy, Headers, Request, Response
+from hishel._core._spec import (
+    CacheOptions,
+)
+from hishel._core._storages._sync_base import SyncBaseStorage
+from hishel._core.models import RequestMetadata, extract_metadata_from_headers
+from hishel._utils import (
+    filter_mapping,
+    make_sync_iterator,
+)
+
+try:
+    import httpx
+except ImportError as e:
+    raise ImportError(
+        "httpx is required to use hishel.httpx module. "
+        "Please install hishel with the 'httpx' extra, "
+        "e.g., 'pip install hishel[httpx]'."
+    ) from e
+
+SOCKET_OPTION = t.Union[
+    t.Tuple[int, int, int],
+    t.Tuple[int, int, t.Union[bytes, bytearray]],
+    t.Tuple[int, int, None, int],
+]
+
+# 128 KB
+CHUNK_SIZE = 131072
+
+
+@overload
+def _internal_to_httpx(
+    value: Request,
+) -> httpx.Request: ...
+@overload
+def _internal_to_httpx(
+    value: Response,
+) -> httpx.Response: ...
+def _internal_to_httpx(
+    value: Union[Request, Response],
+) -> Union[httpx.Request, httpx.Response]:
+    """
+    Convert internal Request/Response to httpx.Request/httpx.Response.
+    """
+    if isinstance(value, Request):
+        return httpx.Request(
+            method=value.method,
+            url=value.url,
+            headers=value.headers,
+            stream=_IteratorStream(value.iter_stream()),
+            extensions=value.metadata,
+        )
+    elif isinstance(value, Response):
+        return httpx.Response(
+            status_code=value.status_code,
+            headers=value.headers,
+            stream=_IteratorStream(value.iter_stream()),
+            extensions=value.metadata,
+        )
+
+
+@overload
+def _httpx_to_internal(
+    value: httpx.Request,
+) -> Request: ...
+@overload
+def _httpx_to_internal(
+    value: httpx.Response,
+) -> Response: ...
+def _httpx_to_internal(
+    value: Union[httpx.Request, httpx.Response],
+) -> Union[Request, Response]:
+    """
+    Convert httpx.Request/httpx.Response to internal Request/Response.
+    """
+    headers = Headers(
+        filter_mapping(
+            Headers({key: value for key, value in value.headers.items()}),
+            ["Transfer-Encoding"],
+        )
+    )
+    if isinstance(value, httpx.Request):
+        extension_metadata = RequestMetadata(
+            hishel_refresh_ttl_on_access=value.extensions.get("hishel_refresh_ttl_on_access"),
+            hishel_ttl=value.extensions.get("hishel_ttl"),
+            hishel_spec_ignore=value.extensions.get("hishel_spec_ignore"),
+            hishel_body_key=value.extensions.get("hishel_body_key"),
+        )
+        headers_metadata = extract_metadata_from_headers(value.headers)
+
+        for key, val in extension_metadata.items():
+            if key in value.extensions:
+                headers_metadata[key] = val  # type: ignore
+
+        try:
+            stream = make_sync_iterator([value.content])
+        except RequestNotRead:
+            stream = cast(Iterator[bytes], value.stream)
+
+        return Request(
+            method=value.method,
+            url=str(value.url),
+            headers=headers,
+            stream=stream,
+            metadata=headers_metadata,
+        )
+    elif isinstance(value, httpx.Response):
+        if value.is_stream_consumed and "content-encoding" in value.headers:
+            raise RuntimeError("Can't get the raw stream of a response with `Content-Encoding` header.")
+        stream = (
+            make_sync_iterator([value.content]) if value.is_stream_consumed else value.iter_raw(chunk_size=CHUNK_SIZE)
+        )
+
+        return Response(
+            status_code=value.status_code,
+            headers=headers,
+            stream=stream,
+            metadata={},
+        )
+
+
+class _IteratorStream(httpx.SyncByteStream, httpx.AsyncByteStream):
+    def __init__(self, iterator: Iterator[bytes] | Iterator[bytes]) -> None:
+        self.iterator = iterator
+
+    def __iter__(self) -> Iterator[bytes]:
+        assert isinstance(self.iterator, (Iterator, Iterable))
+        for chunk in self.iterator:
+            yield chunk
+
+
+class SyncCacheTransport(httpx.BaseTransport):
+    def __init__(
+        self,
+        next_transport: httpx.BaseTransport,
+        storage: SyncBaseStorage | None = None,
+        cache_options: CacheOptions | None = None,
+        ignore_specification: bool = False,
+    ) -> None:
+        self.next_transport = next_transport
+        self._cache_proxy: SyncCacheProxy = SyncCacheProxy(
+            request_sender=self.request_sender,
+            storage=storage,
+            cache_options=cache_options,
+            ignore_specification=ignore_specification,
+        )
+        self.storage = self._cache_proxy.storage
+
+    def handle_request(
+        self,
+        request: httpx.Request,
+    ) -> httpx.Response:
+        internal_request = _httpx_to_internal(request)
+        internal_response = self._cache_proxy.handle_request(internal_request)
+        response = _internal_to_httpx(internal_response)
+        return response
+
+    def close(self) -> None:
+        self.next_transport.close()
+        self.storage.close()
+        super().close()
+
+    def request_sender(self, request: Request) -> Response:
+        httpx_request = _internal_to_httpx(request)
+        httpx_response = self.next_transport.handle_request(httpx_request)
+        return _httpx_to_internal(httpx_response)
+
+
+class SyncCacheClient(httpx.Client):
+    def __init__(self, *args: t.Any, **kwargs: t.Any) -> None:
+        self.storage: SyncBaseStorage | None = kwargs.pop("storage", None)
+        self.cache_options: CacheOptions | None = kwargs.pop("cache_options", None)
+        self.ignore_specification: bool = kwargs.pop("ignore_specification", False)
+        super().__init__(*args, **kwargs)
+
+    def _init_transport(
+        self,
+        verify: ssl.SSLContext | str | bool = True,
+        cert: t.Union[str, t.Tuple[str, str], t.Tuple[str, str, str], None] = None,
+        trust_env: bool = True,
+        http1: bool = True,
+        http2: bool = False,
+        limits: httpx.Limits = httpx.Limits(max_connections=100, max_keepalive_connections=20),
+        transport: httpx.BaseTransport | None = None,
+        **kwargs: t.Any,
+    ) -> httpx.BaseTransport:
+        if transport is not None:
+            return transport
+
+        return SyncCacheTransport(
+            next_transport=httpx.HTTPTransport(
+                verify=verify,
+                cert=cert,
+                trust_env=trust_env,
+                http1=http1,
+                http2=http2,
+                limits=limits,
+            ),
+            storage=self.storage,
+            cache_options=self.cache_options,
+            ignore_specification=False,
+        )
+
+    def _init_proxy_transport(
+        self,
+        proxy: httpx.Proxy,
+        verify: ssl.SSLContext | str | bool = True,
+        cert: t.Union[str, t.Tuple[str, str], t.Tuple[str, str, str], None] = None,
+        trust_env: bool = True,
+        http1: bool = True,
+        http2: bool = False,
+        limits: httpx.Limits = httpx.Limits(max_connections=100, max_keepalive_connections=20),
+        **kwargs: t.Any,
+    ) -> httpx.BaseTransport:
+        return SyncCacheTransport(
+            next_transport=httpx.HTTPTransport(
+                verify=verify,
+                cert=cert,
+                trust_env=trust_env,
+                http1=http1,
+                http2=http2,
+                limits=limits,
+                proxy=proxy,
+            ),
+            storage=self.storage,
+            cache_options=self.cache_options,
+            ignore_specification=self.ignore_specification,
+        )

hishel/_utils.py

@@ -3,7 +3,8 @@ from __future__ import annotations
 import calendar
 import time
 import typing as tp
-from email.utils import parsedate_tz
+from email.utils import formatdate, parsedate_tz
+from pathlib import Path
 from typing import AsyncIterator, Iterable, Iterator
 
 HEADERS_ENCODING = "iso-8859-1"
@@ -50,11 +51,35 @@ def partition(iterable: tp.Iterable[T], predicate: tp.Callable[[T], bool]) -> tp
     return matching, non_matching
 
 
-async def make_async_iterator(iterable: Iterable[bytes]) -> AsyncIterator[bytes]:
+async def make_async_iterator(
+    iterable: Iterable[bytes],
+) -> AsyncIterator[bytes]:
     for item in iterable:
         yield item
 
 
+def filter_mapping(mapping: tp.Mapping[str, T], keys_to_exclude: tp.Iterable[str]) -> tp.Dict[str, T]:
+    """
+        Filter out specified keys from a string-keyed mapping using case-insensitive comparison.
+
+        Args:
+            mapping: The input mapping with string keys to filter.
+            keys_to_exclude: An iterable of string keys to exclude (case-insensitive).
+
+        Returns:
+            A new dictionary with the specified keys excluded.
+
+        Example:
+    ```python
+            original = {'a': 1, 'B': 2, 'c': 3}
+            filtered = filter_mapping(original, ['b'])
+            # filtered will be {'a': 1, 'c': 3}
+    ```
+    """
+    exclude_set = {k.lower() for k in keys_to_exclude}
+    return {k: v for k, v in mapping.items() if k.lower() not in exclude_set}
+
+
 def make_sync_iterator(iterable: Iterable[bytes]) -> Iterator[bytes]:
     for item in iterable:
         yield item
@@ -80,3 +105,25 @@ def snake_to_header(text: str) -> str:
     """
     # Split by underscore, capitalize each word, join with dash, add X- prefix
     return "X-" + "-".join(word.capitalize() for word in text.split("_"))
+
+
+def ensure_cache_dict(base_path: str | None = None) -> Path:
+    _base_path = Path(base_path) if base_path is not None else Path(".cache/hishel")
+    _gitignore_file = _base_path / ".gitignore"
+
+    _base_path.mkdir(parents=True, exist_ok=True)
+
+    if not _gitignore_file.is_file():
+        with open(_gitignore_file, "w", encoding="utf-8") as f:
+            f.write("# Automatically created by Hishel\n*")
+    return _base_path
+
+
+def generate_http_date() -> str:
+    """
+    Generate a Date header value for HTTP responses.
+    Returns date in RFC 1123 format (required by HTTP/1.1).
+
+    Example output: 'Sun, 26 Oct 2025 12:34:56 GMT'
+    """
+    return formatdate(timeval=None, localtime=False, usegmt=True)