Implement ETag support (aio-libs#4594)

This change adds an `etag` property to the response object and
`if_match`, `if_none_match` properties to the request object.
Also, it implements ETag support in static routes and fixes a
few bugs found along the way.

Refs:
* https://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.26
* https://tools.ietf.org/html/rfc7232#section-2.3
* https://tools.ietf.org/html/rfc7232#section-6

PR aio-libs#5298 by @greshilov
Resolves aio-libs#4594

Co-Authored-By: Serhiy Storchaka <storchaka@gmail.com>
Co-Authored-By: Andrew Svetlov <andrew.svetlov@gmail.com>

CHANGES/4594.feature

@@ -0,0 +1 @@
+FileResponse now supports ETag.

aiohttp/__init__.py

@@ -38,7 +38,7 @@
 )
 from .cookiejar import CookieJar as CookieJar, DummyCookieJar as DummyCookieJar
 from .formdata import FormData as FormData
-from .helpers import BasicAuth as BasicAuth, ChainMapProxy as ChainMapProxy
+from .helpers import BasicAuth, ChainMapProxy, ETag
 from .http import (
     HttpVersion as HttpVersion,
     HttpVersion10 as HttpVersion10,
@@ -146,6 +146,7 @@
     # helpers
     "BasicAuth",
     "ChainMapProxy",
+    "ETag",
     # http
     "HttpVersion",
     "HttpVersion10",

aiohttp/helpers.py

@@ -52,7 +52,7 @@
 from .log import client_logger, internal_logger
 from .typedefs import PathLike, Protocol  # noqa
 
-__all__ = ("BasicAuth", "ChainMapProxy")
+__all__ = ("BasicAuth", "ChainMapProxy", "ETag")
 
 PY_36 = sys.version_info >= (3, 6)
 PY_37 = sys.version_info >= (3, 7)
@@ -776,3 +776,26 @@ def __bool__(self) -> bool:
     def __repr__(self) -> str:
         content = ", ".join(map(repr, self._maps))
         return f"ChainMapProxy({content})"
+
+
+# https://tools.ietf.org/html/rfc7232#section-2.3
+_ETAGC = r"[!#-}\x80-\xff]+"
+_ETAGC_RE = re.compile(_ETAGC)
+_QUOTED_ETAG = fr'(W/)?"({_ETAGC})"'
+QUOTED_ETAG_RE = re.compile(_QUOTED_ETAG)
+LIST_QUOTED_ETAG_RE = re.compile(fr"({_QUOTED_ETAG})(?:\s*,\s*|$)|(.)")
+
+ETAG_ANY = "*"
+
+
+@attr.s(auto_attribs=True, frozen=True, slots=True)
+class ETag:
+    value: str
+    is_weak: bool = False
+
+
+def validate_etag_value(value: str) -> None:
+    if value != ETAG_ANY and not _ETAGC_RE.fullmatch(value):
+        raise ValueError(
+            f"Value {value!r} is not a valid etag. Maybe it contains '\"'?"
+        )

aiohttp/web_fileresponse.py

@@ -9,14 +9,17 @@
     Any,
     Awaitable,
     Callable,
+    Iterator,
     List,
     Optional,
+    Tuple,
     Union,
     cast,
 )
 
 from . import hdrs
 from .abc import AbstractStreamWriter
+from .helpers import ETAG_ANY, ETag
 from .typedefs import Final, LooseHeaders
 from .web_exceptions import (
     HTTPNotModified,
@@ -100,6 +103,30 @@ async def _sendfile(
         await super().write_eof()
         return writer
 
+    @staticmethod
+    def _strong_etag_match(etag_value: str, etags: Tuple[ETag, ...]) -> bool:
+        if len(etags) == 1 and etags[0].value == ETAG_ANY:
+            return True
+        return any(etag.value == etag_value for etag in etags if not etag.is_weak)
+
+    async def _not_modified(
+        self, request: "BaseRequest", etag_value: str, last_modified: float
+    ) -> Optional[AbstractStreamWriter]:
+        self.set_status(HTTPNotModified.status_code)
+        self._length_check = False
+        self.etag = etag_value  # type: ignore[assignment]
+        self.last_modified = last_modified  # type: ignore[assignment]
+        # Delete any Content-Length headers provided by user. HTTP 304
+        # should always have empty response body
+        return await super().prepare(request)
+
+    async def _precondition_failed(
+        self, request: "BaseRequest"
+    ) -> Optional[AbstractStreamWriter]:
+        self.set_status(HTTPPreconditionFailed.status_code)
+        self.content_length = 0
+        return await super().prepare(request)
+
     async def prepare(self, request: "BaseRequest") -> Optional[AbstractStreamWriter]:
         filepath = self._path
 
@@ -112,20 +139,35 @@ async def prepare(self, request: "BaseRequest") -> Optional[AbstractStreamWriter
                 gzip = True
 
         loop = asyncio.get_event_loop()
-        st = await loop.run_in_executor(None, filepath.stat)
+        st: os.stat_result = await loop.run_in_executor(None, filepath.stat)
 
-        modsince = request.if_modified_since
-        if modsince is not None and st.st_mtime <= modsince.timestamp():
-            self.set_status(HTTPNotModified.status_code)
-            self._length_check = False
-            # Delete any Content-Length headers provided by user. HTTP 304
-            # should always have empty response body
-            return await super().prepare(request)
+        etag_value = f"{st.st_mtime_ns:x}-{st.st_size:x}"
+        last_modified = st.st_mtime
+
+        # https://tools.ietf.org/html/rfc7232#section-6
+        ifmatch = request.if_match
+        if ifmatch is not None and not self._strong_etag_match(etag_value, ifmatch):
+            return await self._precondition_failed(request)
 
         unmodsince = request.if_unmodified_since
-        if unmodsince is not None and st.st_mtime > unmodsince.timestamp():
-            self.set_status(HTTPPreconditionFailed.status_code)
-            return await super().prepare(request)
+        if (
+            unmodsince is not None
+            and ifmatch is None
+            and st.st_mtime > unmodsince.timestamp()
+        ):
+            return await self._precondition_failed(request)
+
+        ifnonematch = request.if_none_match
+        if ifnonematch is not None and self._strong_etag_match(etag_value, ifnonematch):
+            return await self._not_modified(request, etag_value, last_modified)
+
+        modsince = request.if_modified_since
+        if (
+            modsince is not None
+            and ifnonematch is None
+            and st.st_mtime <= modsince.timestamp()
+        ):
+            return await self._not_modified(request, etag_value, last_modified)
 
         if hdrs.CONTENT_TYPE not in self.headers:
             ct, encoding = mimetypes.guess_type(str(filepath))
@@ -216,6 +258,8 @@ async def prepare(self, request: "BaseRequest") -> Optional[AbstractStreamWriter
             self.headers[hdrs.CONTENT_ENCODING] = encoding
         if gzip:
             self.headers[hdrs.VARY] = hdrs.ACCEPT_ENCODING
+
+        self.etag = etag_value  # type: ignore[assignment]
         self.last_modified = st.st_mtime  # type: ignore[assignment]
         self.content_length = count
 

aiohttp/web_request.py

@@ -31,7 +31,16 @@
 
 from . import hdrs
 from .abc import AbstractStreamWriter
-from .helpers import DEBUG, ChainMapProxy, HeadersMixin, reify, sentinel
+from .helpers import (
+    DEBUG,
+    ETAG_ANY,
+    LIST_QUOTED_ETAG_RE,
+    ChainMapProxy,
+    ETag,
+    HeadersMixin,
+    reify,
+    sentinel,
+)
 from .http_parser import RawRequestMessage
 from .http_writer import HttpVersion
 from .multipart import BodyPartReader, MultipartReader
@@ -495,6 +504,52 @@ def if_unmodified_since(self) -> Optional[datetime.datetime]:
         """
         return self._http_date(self.headers.get(hdrs.IF_UNMODIFIED_SINCE))
 
+    @staticmethod
+    def _etag_values(etag_header: str) -> Iterator[ETag]:
+        """Extract `ETag` objects from raw header."""
+        if etag_header == ETAG_ANY:
+            yield ETag(
+                is_weak=False,
+                value=ETAG_ANY,
+            )
+        else:
+            for match in LIST_QUOTED_ETAG_RE.finditer(etag_header):
+                is_weak, value, garbage = match.group(2, 3, 4)
+                # Any symbol captured by 4th group means
+                # that the following sequence is invalid.
+                if garbage:
+                    break
+
+                yield ETag(
+                    is_weak=bool(is_weak),
+                    value=value,
+                )
+
+    @classmethod
+    def _if_match_or_none_impl(
+        cls, header_value: Optional[str]
+    ) -> Optional[Tuple[ETag, ...]]:
+        if not header_value:
+            return None
+
+        return tuple(cls._etag_values(header_value))
+
+    @reify
+    def if_match(self) -> Optional[Tuple[ETag, ...]]:
+        """The value of If-Match HTTP header, or None.
+
+        This header is represented as a `tuple` of `ETag` objects.
+        """
+        return self._if_match_or_none_impl(self.headers.get(hdrs.IF_MATCH))
+
+    @reify
+    def if_none_match(self) -> Optional[Tuple[ETag, ...]]:
+        """The value of If-None-Match HTTP header, or None.
+
+        This header is represented as a `tuple` of `ETag` objects.
+        """
+        return self._if_match_or_none_impl(self.headers.get(hdrs.IF_NONE_MATCH))
+
     @reify
     def if_range(self) -> Optional[datetime.datetime]:
         """The value of If-Range HTTP header, or None.

aiohttp/web_response.py

@@ -27,7 +27,16 @@
 
 from . import hdrs, payload
 from .abc import AbstractStreamWriter
-from .helpers import PY_38, HeadersMixin, rfc822_formatted_time, sentinel
+from .helpers import (
+    ETAG_ANY,
+    PY_38,
+    QUOTED_ETAG_RE,
+    ETag,
+    HeadersMixin,
+    rfc822_formatted_time,
+    sentinel,
+    validate_etag_value,
+)
 from .http import RESPONSES, SERVER_SOFTWARE, HttpVersion10, HttpVersion11
 from .payload import Payload
 from .typedefs import JSONEncoder, LooseHeaders
@@ -341,6 +350,43 @@ def last_modified(
         elif isinstance(value, str):
             self._headers[hdrs.LAST_MODIFIED] = value
 
+    @property
+    def etag(self) -> Optional[ETag]:
+        quoted_value = self._headers.get(hdrs.ETAG)
+        if not quoted_value:
+            return None
+        elif quoted_value == ETAG_ANY:
+            return ETag(value=ETAG_ANY)
+        match = QUOTED_ETAG_RE.fullmatch(quoted_value)
+        if not match:
+            return None
+        is_weak, value = match.group(1, 2)
+        return ETag(
+            is_weak=bool(is_weak),
+            value=value,
+        )
+
+    @etag.setter
+    def etag(self, value: Optional[Union[ETag, str]]) -> None:
+        if value is None:
+            self._headers.pop(hdrs.ETAG, None)
+        elif (isinstance(value, str) and value == ETAG_ANY) or (
+            isinstance(value, ETag) and value.value == ETAG_ANY
+        ):
+            self._headers[hdrs.ETAG] = ETAG_ANY
+        elif isinstance(value, str):
+            validate_etag_value(value)
+            self._headers[hdrs.ETAG] = f'"{value}"'
+        elif isinstance(value, ETag) and isinstance(value.value, str):
+            validate_etag_value(value.value)
+            hdr_value = f'W/"{value.value}"' if value.is_weak else f'"{value.value}"'
+            self._headers[hdrs.ETAG] = hdr_value
+        else:
+            raise ValueError(
+                f"Unsupported etag type: {type(value)}. "
+                f"etag must be str, ETag or None"
+            )
+
     def _generate_content_type_header(
         self, CONTENT_TYPE: istr = hdrs.CONTENT_TYPE
     ) -> None:
@@ -435,7 +481,7 @@ async def _prepare_headers(self) -> None:
             elif version >= HttpVersion11 and self.status in (100, 101, 102, 103, 204):
                 del headers[hdrs.CONTENT_LENGTH]
 
-        if self.status != 204:
+        if self.status not in (204, 304):
             headers.setdefault(hdrs.CONTENT_TYPE, "application/octet-stream")
         headers.setdefault(hdrs.DATE, rfc822_formatted_time())
         headers.setdefault(hdrs.SERVER, SERVER_SOFTWARE)

docs/client_reference.rst

@@ -1688,6 +1688,23 @@ ClientTimeout
 
    .. versionadded:: 3.3
 
+ETag
+^^^^
+
+.. class:: ETag(name, is_weak=False)
+
+   Represents `ETag` identifier.
+
+   .. attribute:: value
+
+      Value of corresponding etag without quotes.
+
+   .. attribute:: is_weak
+
+      Flag indicates that etag is weak (has `W/` prefix).
+
+   .. versionadded:: 3.8
+
 RequestInfo
 ^^^^^^^^^^^
 

docs/spelling_wordlist.txt

@@ -28,6 +28,7 @@ Dict
 Discord
 Django
 Dup
+ETag
 Facebook
 HTTPException
 HttpProcessingError
@@ -153,6 +154,7 @@ env
 environ
 eof
 epoll
+etag
 facto
 fallback
 fallbacks

docs/web_reference.rst

@@ -336,6 +336,26 @@ and :ref:`aiohttp-web-signals` handlers.
 
       .. versionadded:: 3.1
 
+   .. attribute:: if_match
+
+      Read-only property that returns :class:`ETag` objects specified
+      in the *If-Match* header.
+
+      Returns :class:`tuple` of :class:`ETag` or ``None`` if
+      *If-Match* header is absent.
+
+      .. versionadded:: 3.8
+
+   .. attribute:: if_none_match
+
+      Read-only property that returns :class:`ETag` objects specified
+      *If-None-Match* header.
+
+      Returns :class:`tuple` of :class:`ETag` or ``None`` if
+      *If-None-Match* header is absent.
+
+      .. versionadded:: 3.8
+
    .. attribute:: if_range
 
       Read-only property that returns the date specified in the
@@ -782,6 +802,20 @@ StreamResponse
       as an :class:`int` or a :class:`float` object, and the
       value ``None`` to unset the header.
 
+   .. attribute:: etag
+
+      *ETag* header for outgoing response.
+
+      This property accepts raw :class:`str` values, :class:`ETag`
+      objects and the value ``None`` to unset the header.
+
+      In case of :class:`str` input, etag is considered as strong by default.
+
+      **Do not** use double quotes ``"`` in the etag value,
+      they will be added automatically.
+
+      .. versionadded:: 3.8
+
    .. comethod:: prepare(request)
 
       :param aiohttp.web.Request request: HTTP request object, that the