Add request filter API for host blocking

Closes https://github.com/qutebrowser/qutebrowser-extensions/issues/8

qutebrowser/api/requests.py

@@ -0,0 +1,37 @@
+# vim: ft=python fileencoding=utf-8 sts=4 sw=4 et:
+
+# Copyright 2018 Florian Bruhin (The Compiler) <mail@qutebrowser.org>
+#
+# This file is part of qutebrowser.
+#
+# qutebrowser is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# qutebrowser is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with qutebrowser.  If not, see <http://www.gnu.org/licenses/>.
+
+"""APIs related to intercepting/blocking requests."""
+
+import typing
+
+import attr
+from PyQt5.QtCore import QUrl
+
+from qutebrowser.extensions import requests
+# pylint: disable=unused-import
+from qutebrowser.extensions.requests import Request
+
+
+def register_filter(reqfilter: requests.RequestFilterType) -> None:
+    """Register a request filter.
+
+    Whenever a request happens, the filter gets called with a Request object.
+    """
+    requests.register_filter(reqfilter)

qutebrowser/browser/adblock.py

@@ -119,9 +119,15 @@ class HostBlocker:
             return False
 
         host = url.host()
-        return ((host in self._blocked_hosts or
+        blocked = ((host in self._blocked_hosts or
-                 host in self._config_blocked_hosts) and
+                    host in self._config_blocked_hosts) and
-                not _is_whitelisted_url(url))
+                   not _is_whitelisted_url(url))
+
+        if blocked:
+            logger.info("Request to {} blocked by host blocker."
+                        .format(url.host()))
+
+        return blocked
 
     def _read_hosts_file(self, filename, target):
         """Read hosts from the given filename.

qutebrowser/browser/webengine/interceptor.py

@@ -26,15 +26,15 @@ from PyQt5.QtWebEngineCore import (QWebEngineUrlRequestInterceptor,
 from qutebrowser.config import config
 from qutebrowser.browser import shared
 from qutebrowser.utils import utils, log, debug
+from qutebrowser.extensions import requests
 
 
 class RequestInterceptor(QWebEngineUrlRequestInterceptor):
 
     """Handle ad blocking and custom headers."""
 
-    def __init__(self, host_blocker, args, parent=None):
+    def __init__(self, args, parent=None):
         super().__init__(parent)
-        self._host_blocker = host_blocker
         self._args = args
 
     def install(self, profile):
@@ -84,9 +84,10 @@ class RequestInterceptor(QWebEngineUrlRequestInterceptor):
                 return
 
         # FIXME:qtwebengine only block ads for NavigationTypeOther?
-        if self._host_blocker.is_blocked(url, first_party):
+        request = requests.Request(first_party_url=first_party,
-            log.webview.info("Request to {} blocked by host blocker.".format(
+                                   request_url=url)
-                url.host()))
+        requests.run_filters(request)
+        if request.is_blocked:
             info.block(True)
 
         for header, value in shared.custom_headers(url=url):

qutebrowser/browser/webengine/webenginetab.py

@@ -60,10 +60,8 @@ def init():
     _qute_scheme_handler.install(webenginesettings.private_profile)
 
     log.init.debug("Initializing request interceptor...")
-    host_blocker = objreg.get('host-blocker')
     args = objreg.get('args')
-    req_interceptor = interceptor.RequestInterceptor(
+    req_interceptor = interceptor.RequestInterceptor(args=args, parent=app)
-        host_blocker, args=args, parent=app)
     req_interceptor.install(webenginesettings.default_profile)
     req_interceptor.install(webenginesettings.private_profile)
 

qutebrowser/browser/webkit/mhtml.py

@@ -39,6 +39,7 @@ from PyQt5.QtCore import QUrl
 from qutebrowser.browser import downloads
 from qutebrowser.browser.webkit import webkitelem
 from qutebrowser.utils import log, objreg, message, usertypes, utils, urlutils
+from qutebrowser.extensions import requests
 
 
 @attr.s
@@ -354,8 +355,9 @@ class _Downloader:
         # qute, see the comments/discussion on
         # https://github.com/qutebrowser/qutebrowser/pull/962#discussion_r40256987
         # and https://github.com/qutebrowser/qutebrowser/issues/1053
-        host_blocker = objreg.get('host-blocker')
+        request = requests.Request(first_party_url=None, request_url=url)
-        if host_blocker.is_blocked(url):
+        requests.run_filters(request)
+        if request.is_blocked:
             log.downloads.debug("Skipping {}, host-blocked".format(url))
             # We still need an empty file in the output, QWebView can be pretty
             # picky about displaying a file correctly when not all assets are

qutebrowser/browser/webkit/network/networkmanager.py

@@ -38,6 +38,7 @@ if MYPY:
 from qutebrowser.utils import (message, log, usertypes, utils, objreg,
                                urlutils, debug)
 from qutebrowser.browser import shared
+from qutebrowser.extensions import requests
 from qutebrowser.browser.webkit import certificateerror
 from qutebrowser.browser.webkit.network import (webkitqutescheme, networkreply,
                                                 filescheme)
@@ -405,10 +406,10 @@ class NetworkManager(QNetworkAccessManager):
                 # the webpage shutdown here.
                 current_url = QUrl()
 
-        host_blocker = objreg.get('host-blocker')
+        request = requests.Request(first_party_url=current_url,
-        if host_blocker.is_blocked(req.url(), current_url):
+                                   request_url=req.url())
-            log.webview.info("Request to {} blocked by host blocker.".format(
+        requests.run_filters(request)
-                req.url().host()))
+        if request.is_blocked:
             return networkreply.ErrorNetworkReply(
                 req, HOSTBLOCK_ERROR_STRING, QNetworkReply.ContentAccessDenied,
                 self)

qutebrowser/extensions/requests.py

@@ -0,0 +1,53 @@
+# vim: ft=python fileencoding=utf-8 sts=4 sw=4 et:
+
+# Copyright 2018 Florian Bruhin (The Compiler) <mail@qutebrowser.org>
+#
+# This file is part of qutebrowser.
+#
+# qutebrowser is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# qutebrowser is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with qutebrowser.  If not, see <http://www.gnu.org/licenses/>.
+
+"""Infrastructure for filtering requests."""
+
+import typing
+
+import attr
+
+
+@attr.s
+class Request:
+
+    """A request which can be blocked."""
+
+    first_party_url = attr.ib()  # type: QUrl
+    request_url = attr.ib()  # type: QUrl
+    is_blocked = attr.ib(False)  # type: bool
+
+    def block(self):
+        """Block this request."""
+        self.is_blocked = True
+
+
+RequestFilterType = typing.Callable[[Request], None]
+
+
+_request_filters = []  # type: typing.List[RequestFilterType]
+
+
+def register_filter(reqfilter: RequestFilterType) -> None:
+    _request_filters.append(reqfilter)
+
+
+def run_filters(info):
+    for reqfilter in _request_filters:
+        reqfilter(info)