Establish basic file browser and index fallback (#2662)

Co-authored-by: L. Kärkkäinen <98187+Tronic@users.noreply.github.com>
Co-authored-by: L. Karkkainen <tronic@users.noreply.github.com>
This commit is contained in:
Adam Hopkins
2023-02-05 15:09:04 +02:00
committed by GitHub
parent 30c53b6857
commit 9cb9e88678
45 changed files with 1031 additions and 477 deletions

35
sanic/mixins/base.py Normal file
View File

@@ -0,0 +1,35 @@
from typing import Optional
from sanic.base.meta import SanicMeta
class BaseMixin(metaclass=SanicMeta):
name: str
strict_slashes: Optional[bool]
def _generate_name(self, *objects) -> str:
name = None
for obj in objects:
if obj:
if isinstance(obj, str):
name = obj
break
try:
name = obj.name
except AttributeError:
try:
name = obj.__name__
except AttributeError:
continue
else:
break
if not name: # noqa
raise ValueError("Could not generate a name for handler")
if not name.startswith(f"{self.name}."):
name = f"{self.name}.{name}"
return name

View File

@@ -1,11 +1,6 @@
from ast import NodeVisitor, Return, parse
from contextlib import suppress
from email.utils import formatdate
from functools import partial, wraps
from inspect import getsource, signature
from mimetypes import guess_type
from os import path
from pathlib import Path, PurePath
from textwrap import dedent
from typing import (
Any,
@@ -19,20 +14,15 @@ from typing import (
Union,
cast,
)
from urllib.parse import unquote
from sanic_routing.route import Route
from sanic.base.meta import SanicMeta
from sanic.compat import stat_async
from sanic.constants import DEFAULT_HTTP_CONTENT_TYPE, HTTP_METHODS
from sanic.constants import HTTP_METHODS
from sanic.errorpages import RESPONSE_MAPPING
from sanic.exceptions import FileNotFound, HeaderNotFound, RangeNotSatisfiable
from sanic.handlers import ContentRangeHandler
from sanic.log import error_logger
from sanic.mixins.base import BaseMixin
from sanic.models.futures import FutureRoute, FutureStatic
from sanic.models.handler_types import RouteHandler
from sanic.response import HTTPResponse, file, file_stream, validate_file
from sanic.types import HashableDict
@@ -41,20 +31,14 @@ RouteWrapper = Callable[
]
class RouteMixin(metaclass=SanicMeta):
name: str
class RouteMixin(BaseMixin, metaclass=SanicMeta):
def __init__(self, *args, **kwargs) -> None:
self._future_routes: Set[FutureRoute] = set()
self._future_statics: Set[FutureStatic] = set()
self.strict_slashes: Optional[bool] = False
def _apply_route(self, route: FutureRoute) -> List[Route]:
raise NotImplementedError # noqa
def _apply_static(self, static: FutureStatic) -> Route:
raise NotImplementedError # noqa
def route(
self,
uri: str,
@@ -688,324 +672,6 @@ class RouteMixin(metaclass=SanicMeta):
**ctx_kwargs,
)(handler)
def static(
self,
uri: str,
file_or_directory: Union[str, bytes, PurePath],
pattern: str = r"/?.+",
use_modified_since: bool = True,
use_content_range: bool = False,
stream_large_files: bool = False,
name: str = "static",
host: Optional[str] = None,
strict_slashes: Optional[bool] = None,
content_type: Optional[bool] = None,
apply: bool = True,
resource_type: Optional[str] = None,
):
"""
Register a root to serve files from. The input can either be a
file or a directory. This method will enable an easy and simple way
to setup the :class:`Route` necessary to serve the static files.
:param uri: URL path to be used for serving static content
:param file_or_directory: Path for the Static file/directory with
static files
:param pattern: Regex Pattern identifying the valid static files
:param use_modified_since: If true, send file modified time, and return
not modified if the browser's matches the server's
:param use_content_range: If true, process header for range requests
and sends the file part that is requested
:param stream_large_files: If true, use the
:func:`StreamingHTTPResponse.file_stream` handler rather
than the :func:`HTTPResponse.file` handler to send the file.
If this is an integer, this represents the threshold size to
switch to :func:`StreamingHTTPResponse.file_stream`
:param name: user defined name used for url_for
:param host: Host IP or FQDN for the service to use
:param strict_slashes: Instruct :class:`Sanic` to check if the request
URLs need to terminate with a */*
:param content_type: user defined content type for header
:return: routes registered on the router
:rtype: List[sanic.router.Route]
"""
name = self._generate_name(name)
if strict_slashes is None and self.strict_slashes is not None:
strict_slashes = self.strict_slashes
if not isinstance(file_or_directory, (str, bytes, PurePath)):
raise ValueError(
f"Static route must be a valid path, not {file_or_directory}"
)
static = FutureStatic(
uri,
file_or_directory,
pattern,
use_modified_since,
use_content_range,
stream_large_files,
name,
host,
strict_slashes,
content_type,
resource_type,
)
self._future_statics.add(static)
if apply:
self._apply_static(static)
def _generate_name(self, *objects) -> str:
name = None
for obj in objects:
if obj:
if isinstance(obj, str):
name = obj
break
try:
name = obj.name
except AttributeError:
try:
name = obj.__name__
except AttributeError:
continue
else:
break
if not name: # noqa
raise ValueError("Could not generate a name for handler")
if not name.startswith(f"{self.name}."):
name = f"{self.name}.{name}"
return name
async def _get_file_path(self, file_or_directory, __file_uri__, not_found):
file_path_raw = Path(unquote(file_or_directory))
root_path = file_path = file_path_raw.resolve()
if __file_uri__:
# Strip all / that in the beginning of the URL to help prevent
# python from herping a derp and treating the uri as an
# absolute path
unquoted_file_uri = unquote(__file_uri__).lstrip("/")
file_path_raw = Path(file_or_directory, unquoted_file_uri)
file_path = file_path_raw.resolve()
if (
file_path < root_path and not file_path_raw.is_symlink()
) or ".." in file_path_raw.parts:
error_logger.exception(
f"File not found: path={file_or_directory}, "
f"relative_url={__file_uri__}"
)
raise not_found
try:
file_path.relative_to(root_path)
except ValueError:
if not file_path_raw.is_symlink():
error_logger.exception(
f"File not found: path={file_or_directory}, "
f"relative_url={__file_uri__}"
)
raise not_found
return file_path
async def _static_request_handler(
self,
file_or_directory,
use_modified_since,
use_content_range,
stream_large_files,
request,
content_type=None,
__file_uri__=None,
):
not_found = FileNotFound(
"File not found",
path=file_or_directory,
relative_url=__file_uri__,
)
# Merge served directory and requested file if provided
file_path = await self._get_file_path(
file_or_directory, __file_uri__, not_found
)
try:
headers = {}
# Check if the client has been sent this file before
# and it has not been modified since
stats = None
if use_modified_since:
stats = await stat_async(file_path)
modified_since = stats.st_mtime
response = await validate_file(request.headers, modified_since)
if response:
return response
headers["Last-Modified"] = formatdate(
modified_since, usegmt=True
)
_range = None
if use_content_range:
_range = None
if not stats:
stats = await stat_async(file_path)
headers["Accept-Ranges"] = "bytes"
headers["Content-Length"] = str(stats.st_size)
if request.method != "HEAD":
try:
_range = ContentRangeHandler(request, stats)
except HeaderNotFound:
pass
else:
del headers["Content-Length"]
headers.update(_range.headers)
if "content-type" not in headers:
content_type = (
content_type
or guess_type(file_path)[0]
or DEFAULT_HTTP_CONTENT_TYPE
)
if "charset=" not in content_type and (
content_type.startswith("text/")
or content_type == "application/javascript"
):
content_type += "; charset=utf-8"
headers["Content-Type"] = content_type
if request.method == "HEAD":
return HTTPResponse(headers=headers)
else:
if stream_large_files:
if type(stream_large_files) == int:
threshold = stream_large_files
else:
threshold = 1024 * 1024
if not stats:
stats = await stat_async(file_path)
if stats.st_size >= threshold:
return await file_stream(
file_path, headers=headers, _range=_range
)
return await file(file_path, headers=headers, _range=_range)
except RangeNotSatisfiable:
raise
except FileNotFoundError:
raise not_found
except Exception:
error_logger.exception(
f"Exception in static request handler: "
f"path={file_or_directory}, "
f"relative_url={__file_uri__}"
)
raise
def _register_static(
self,
static: FutureStatic,
):
# TODO: Though sanic is not a file server, I feel like we should
# at least make a good effort here. Modified-since is nice, but
# we could also look into etags, expires, and caching
"""
Register a static directory handler with Sanic by adding a route to the
router and registering a handler.
:param app: Sanic
:param file_or_directory: File or directory path to serve from
:type file_or_directory: Union[str,bytes,Path]
:param uri: URL to serve from
:type uri: str
:param pattern: regular expression used to match files in the URL
:param use_modified_since: If true, send file modified time, and return
not modified if the browser's matches the
server's
:param use_content_range: If true, process header for range requests
and sends the file part that is requested
:param stream_large_files: If true, use the file_stream() handler
rather than the file() handler to send the file
If this is an integer, this represents the
threshold size to switch to file_stream()
:param name: user defined name used for url_for
:type name: str
:param content_type: user defined content type for header
:return: registered static routes
:rtype: List[sanic.router.Route]
"""
if isinstance(static.file_or_directory, bytes):
file_or_directory = static.file_or_directory.decode("utf-8")
elif isinstance(static.file_or_directory, PurePath):
file_or_directory = str(static.file_or_directory)
elif not isinstance(static.file_or_directory, str):
raise ValueError("Invalid file path string.")
else:
file_or_directory = static.file_or_directory
uri = static.uri
name = static.name
# If we're not trying to match a file directly,
# serve from the folder
if not static.resource_type:
if not path.isfile(file_or_directory):
uri = uri.rstrip("/")
uri += "/<__file_uri__:path>"
elif static.resource_type == "dir":
if path.isfile(file_or_directory):
raise TypeError(
"Resource type improperly identified as directory. "
f"'{file_or_directory}'"
)
uri = uri.rstrip("/")
uri += "/<__file_uri__:path>"
elif static.resource_type == "file" and not path.isfile(
file_or_directory
):
raise TypeError(
"Resource type improperly identified as file. "
f"'{file_or_directory}'"
)
elif static.resource_type != "file":
raise ValueError(
"The resource_type should be set to 'file' or 'dir'"
)
# special prefix for static files
# if not static.name.startswith("_static_"):
# name = f"_static_{static.name}"
_handler = wraps(self._static_request_handler)(
partial(
self._static_request_handler,
file_or_directory,
static.use_modified_since,
static.use_content_range,
static.stream_large_files,
content_type=static.content_type,
)
)
route, _ = self.route( # type: ignore
uri=uri,
methods=["GET", "HEAD"],
name=name,
host=static.host,
strict_slashes=static.strict_slashes,
static=True,
)(_handler)
return route
def _determine_error_format(self, handler) -> str:
with suppress(OSError, TypeError):
src = dedent(getsource(handler))

View File

@@ -1109,7 +1109,6 @@ class StartupMixin(metaclass=SanicMeta):
app: StartupMixin,
server_info: ApplicationServerInfo,
) -> None: # no cov
try:
# We should never get to this point without a server
# This is primarily to keep mypy happy

348
sanic/mixins/static.py Normal file
View File

@@ -0,0 +1,348 @@
from email.utils import formatdate
from functools import partial, wraps
from mimetypes import guess_type
from os import PathLike, path
from pathlib import Path, PurePath
from typing import Optional, Sequence, Set, Union, cast
from urllib.parse import unquote
from sanic_routing.route import Route
from sanic.base.meta import SanicMeta
from sanic.compat import stat_async
from sanic.constants import DEFAULT_HTTP_CONTENT_TYPE
from sanic.exceptions import FileNotFound, HeaderNotFound, RangeNotSatisfiable
from sanic.handlers import ContentRangeHandler
from sanic.handlers.directory import DirectoryHandler
from sanic.log import deprecation, error_logger
from sanic.mixins.base import BaseMixin
from sanic.models.futures import FutureStatic
from sanic.request import Request
from sanic.response import HTTPResponse, file, file_stream, validate_file
class StaticMixin(BaseMixin, metaclass=SanicMeta):
def __init__(self, *args, **kwargs) -> None:
self._future_statics: Set[FutureStatic] = set()
def _apply_static(self, static: FutureStatic) -> Route:
raise NotImplementedError # noqa
def static(
self,
uri: str,
file_or_directory: Union[PathLike, str, bytes],
pattern: str = r"/?.+",
use_modified_since: bool = True,
use_content_range: bool = False,
stream_large_files: Union[bool, int] = False,
name: str = "static",
host: Optional[str] = None,
strict_slashes: Optional[bool] = None,
content_type: Optional[str] = None,
apply: bool = True,
resource_type: Optional[str] = None,
index: Optional[Union[str, Sequence[str]]] = None,
directory_view: bool = False,
directory_handler: Optional[DirectoryHandler] = None,
):
"""
Register a root to serve files from. The input can either be a
file or a directory. This method will enable an easy and simple way
to setup the :class:`Route` necessary to serve the static files.
:param uri: URL path to be used for serving static content
:param file_or_directory: Path for the Static file/directory with
static files
:param pattern: Regex Pattern identifying the valid static files
:param use_modified_since: If true, send file modified time, and return
not modified if the browser's matches the server's
:param use_content_range: If true, process header for range requests
and sends the file part that is requested
:param stream_large_files: If true, use the
:func:`StreamingHTTPResponse.file_stream` handler rather
than the :func:`HTTPResponse.file` handler to send the file.
If this is an integer, this represents the threshold size to
switch to :func:`StreamingHTTPResponse.file_stream`
:param name: user defined name used for url_for
:param host: Host IP or FQDN for the service to use
:param strict_slashes: Instruct :class:`Sanic` to check if the request
URLs need to terminate with a */*
:param content_type: user defined content type for header
:param apply: If true, will register the route immediately
:param resource_type: Explicitly declare a resource to be a "
file" or a "dir"
:param index: When exposing against a directory, index is the name that
will be served as the default file. When multiple files names are
passed, then they will be tried in order.
:param directory_view: Whether to fallback to showing the directory
viewer when exposing a directory
:param directory_handler: An instance of :class:`DirectoryHandler`
that can be used for explicitly controlling and subclassing the
behavior of the default directory handler
:return: routes registered on the router
:rtype: List[sanic.router.Route]
"""
name = self._generate_name(name)
if strict_slashes is None and self.strict_slashes is not None:
strict_slashes = self.strict_slashes
if not isinstance(file_or_directory, (str, bytes, PurePath)):
raise ValueError(
f"Static route must be a valid path, not {file_or_directory}"
)
if isinstance(file_or_directory, bytes):
deprecation(
"Serving a static directory with a bytes string is "
"deprecated and will be removed in v22.9.",
22.9,
)
file_or_directory = cast(str, file_or_directory.decode())
file_or_directory = Path(file_or_directory)
if directory_handler and (directory_view or index):
raise ValueError(
"When explicitly setting directory_handler, you cannot "
"set either directory_view or index. Instead, pass "
"these arguments to your DirectoryHandler instance."
)
if not directory_handler:
directory_handler = DirectoryHandler(
uri=uri,
directory=file_or_directory,
directory_view=directory_view,
index=index,
)
static = FutureStatic(
uri,
file_or_directory,
pattern,
use_modified_since,
use_content_range,
stream_large_files,
name,
host,
strict_slashes,
content_type,
resource_type,
directory_handler,
)
self._future_statics.add(static)
if apply:
self._apply_static(static)
class StaticHandleMixin(metaclass=SanicMeta):
def _apply_static(self, static: FutureStatic) -> Route:
return self._register_static(static)
def _register_static(
self,
static: FutureStatic,
):
# TODO: Though sanic is not a file server, I feel like we should
# at least make a good effort here. Modified-since is nice, but
# we could also look into etags, expires, and caching
"""
Register a static directory handler with Sanic by adding a route to the
router and registering a handler.
"""
if isinstance(static.file_or_directory, bytes):
file_or_directory = static.file_or_directory.decode("utf-8")
elif isinstance(static.file_or_directory, PurePath):
file_or_directory = str(static.file_or_directory)
elif not isinstance(static.file_or_directory, str):
raise ValueError("Invalid file path string.")
else:
file_or_directory = static.file_or_directory
uri = static.uri
name = static.name
# If we're not trying to match a file directly,
# serve from the folder
if not static.resource_type:
if not path.isfile(file_or_directory):
uri = uri.rstrip("/")
uri += "/<__file_uri__:path>"
elif static.resource_type == "dir":
if path.isfile(file_or_directory):
raise TypeError(
"Resource type improperly identified as directory. "
f"'{file_or_directory}'"
)
uri = uri.rstrip("/")
uri += "/<__file_uri__:path>"
elif static.resource_type == "file" and not path.isfile(
file_or_directory
):
raise TypeError(
"Resource type improperly identified as file. "
f"'{file_or_directory}'"
)
elif static.resource_type != "file":
raise ValueError(
"The resource_type should be set to 'file' or 'dir'"
)
# special prefix for static files
# if not static.name.startswith("_static_"):
# name = f"_static_{static.name}"
_handler = wraps(self._static_request_handler)(
partial(
self._static_request_handler,
file_or_directory=file_or_directory,
use_modified_since=static.use_modified_since,
use_content_range=static.use_content_range,
stream_large_files=static.stream_large_files,
content_type=static.content_type,
directory_handler=static.directory_handler,
)
)
route, _ = self.route( # type: ignore
uri=uri,
methods=["GET", "HEAD"],
name=name,
host=static.host,
strict_slashes=static.strict_slashes,
static=True,
)(_handler)
return route
async def _static_request_handler(
self,
request: Request,
*,
file_or_directory: PathLike,
use_modified_since: bool,
use_content_range: bool,
stream_large_files: Union[bool, int],
directory_handler: DirectoryHandler,
content_type: Optional[str] = None,
__file_uri__: Optional[str] = None,
):
not_found = FileNotFound(
"File not found",
path=file_or_directory,
relative_url=__file_uri__,
)
# Merge served directory and requested file if provided
file_path = await self._get_file_path(
file_or_directory, __file_uri__, not_found
)
try:
headers = {}
# Check if the client has been sent this file before
# and it has not been modified since
stats = None
if use_modified_since:
stats = await stat_async(file_path)
modified_since = stats.st_mtime
response = await validate_file(request.headers, modified_since)
if response:
return response
headers["Last-Modified"] = formatdate(
modified_since, usegmt=True
)
_range = None
if use_content_range:
_range = None
if not stats:
stats = await stat_async(file_path)
headers["Accept-Ranges"] = "bytes"
headers["Content-Length"] = str(stats.st_size)
if request.method != "HEAD":
try:
_range = ContentRangeHandler(request, stats)
except HeaderNotFound:
pass
else:
del headers["Content-Length"]
headers.update(_range.headers)
if "content-type" not in headers:
content_type = (
content_type
or guess_type(file_path)[0]
or DEFAULT_HTTP_CONTENT_TYPE
)
if "charset=" not in content_type and (
content_type.startswith("text/")
or content_type == "application/javascript"
):
content_type += "; charset=utf-8"
headers["Content-Type"] = content_type
if request.method == "HEAD":
return HTTPResponse(headers=headers)
else:
if stream_large_files:
if isinstance(stream_large_files, bool):
threshold = 1024 * 1024
else:
threshold = stream_large_files
if not stats:
stats = await stat_async(file_path)
if stats.st_size >= threshold:
return await file_stream(
file_path, headers=headers, _range=_range
)
return await file(file_path, headers=headers, _range=_range)
except (IsADirectoryError, PermissionError):
return await directory_handler.handle(request, request.path)
except RangeNotSatisfiable:
raise
except FileNotFoundError:
raise not_found
except Exception:
error_logger.exception(
"Exception in static request handler: "
f"path={file_or_directory}, "
f"relative_url={__file_uri__}"
)
raise
async def _get_file_path(self, file_or_directory, __file_uri__, not_found):
file_path_raw = Path(unquote(file_or_directory))
root_path = file_path = file_path_raw.resolve()
if __file_uri__:
# Strip all / that in the beginning of the URL to help prevent
# python from herping a derp and treating the uri as an
# absolute path
unquoted_file_uri = unquote(__file_uri__).lstrip("/")
file_path_raw = Path(file_or_directory, unquoted_file_uri)
file_path = file_path_raw.resolve()
if (
file_path < root_path and not file_path_raw.is_symlink()
) or ".." in file_path_raw.parts:
error_logger.exception(
f"File not found: path={file_or_directory}, "
f"relative_url={__file_uri__}"
)
raise not_found
try:
file_path.relative_to(root_path)
except ValueError:
if not file_path_raw.is_symlink():
error_logger.exception(
f"File not found: path={file_or_directory}, "
f"relative_url={__file_uri__}"
)
raise not_found
return file_path