drop Python 2 support

Author: kmike

README.rst

@@ -15,7 +15,7 @@ Splash - A javascript rendering service
    :target: https://gitter.im/scrapinghub/splash
 
 Splash is a javascript rendering service with an HTTP API. It's a lightweight
-browser with an HTTP API, implemented in Python using Twisted and QT.
+browser with an HTTP API, implemented in Python 3 using Twisted and QT5.
 
 It's fast, lightweight and state-less which makes it easy to distribute.
 

docs/api.rst

@@ -111,7 +111,7 @@ allowed_content_types : string : optional
   Comma-separated list of allowed content types.
   If present, Splash will abort any request if the response's content type
   doesn't match any of the content types in this list.
-  Wildcards are supported using the `fnmatch <https://docs.python.org/2/library/fnmatch.html>`_
+  Wildcards are supported using the `fnmatch <https://docs.python.org/3/library/fnmatch.html>`_
   syntax.
 
 .. _arg-forbidden-content-types:
@@ -120,7 +120,7 @@ forbidden_content_types : string : optional
   Comma-separated list of forbidden content types.
   If present, Splash will abort any request if the response's content type
   matches any of the content types in this list.
-  Wildcards are supported using the `fnmatch <https://docs.python.org/2/library/fnmatch.html>`_
+  Wildcards are supported using the `fnmatch <https://docs.python.org/3/library/fnmatch.html>`_
   syntax.
 
 .. _arg-viewport:

docs/index.rst

@@ -3,7 +3,7 @@ Splash - A javascript rendering service
 =======================================
 
 Splash is a javascript rendering service. It's a lightweight web browser
-with an HTTP API, implemented in Python using Twisted and QT. The (twisted)
+with an HTTP API, implemented in Python 3 using Twisted and QT5. The (twisted)
 QT reactor is used to make the service fully asynchronous allowing
 to take advantage of webkit concurrency via QT main loop. Some of Splash
 features:

requirements.txt

@@ -1,4 +1,4 @@
-# install PyQt5 (Splash is tested on PyQT 5.5.1)
+# install PyQt5 (Splash is tested on PyQT 5.8)
 # and the following packages:
 twisted >= 15.5.0, < 16.3.0
 qt5reactor
@@ -7,7 +7,6 @@ adblockparser >= 0.5
 https://github.com/sunu/pyre2/archive/c610be52c3b5379b257d56fc0669d022fd70082a.zip#egg=re2
 xvfbwrapper
 Pillow
-six
 
 # for scripting support
 lupa >= 1.3

setup.py

@@ -68,8 +68,6 @@ def get_version():
     ],
     'classifiers': [
         'Programming Language :: Python',
-        'Programming Language :: Python :: 2',
-        'Programming Language :: Python :: 2.7',
         'Programming Language :: Python :: 3',
         'Programming Language :: Python :: 3.4',
         'Programming Language :: Python :: 3.5',

splash/argument_cache.py

@@ -1,26 +1,9 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 import json
 import hashlib
 from collections import OrderedDict
 
 
-# Add move_to_end method for python 2.7
-# See https://github.com/twitter/commons/blob/master/src/python/twitter/common/collections/ordereddict.py#L284
-if not hasattr(OrderedDict, 'move_to_end'):
-    def move_to_end(self, key, last=True):
-        link_prev, link_next, key = link = self._OrderedDict__map[key]
-        link_prev[1] = link_next
-        link_next[0] = link_prev
-        root = self._OrderedDict__root
-        if last:
-            last = root[0]
-            link[0] = last
-            link[1] = root
-            last[1] = root[0] = link
-    OrderedDict.move_to_end = move_to_end
-
-
 class ArgumentCache(object):
     """
     >>> cache = ArgumentCache()

splash/browser_tab.py

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 import base64
 import functools
 import os
@@ -13,7 +12,6 @@
 from PyQt5.QtWidgets import QApplication
 from twisted.internet import defer
 from twisted.python import log
-import six
 
 from splash import defaults
 from splash.har.qt import cookies2har
@@ -359,7 +357,7 @@ def add_cookie(self, cookie):
     @property
     def url(self):
         """ Current URL """
-        return six.text_type(self.web_page.mainFrame().url().toString())
+        return str(self.web_page.mainFrame().url().toString())
 
     def go(self, url, callback, errback, baseurl=None, http_method='GET',
            body=None, headers=None):
@@ -597,7 +595,7 @@ def _cancel_all_timers(self):
             callback_proxy.use_up()
 
     def _on_url_changed(self, url):
-        self.web_page.har.store_redirect(six.text_type(url.toString()))
+        self.web_page.har.store_redirect(str(url.toString()))
         self._cancel_timers(self._timers_to_cancel_on_redirect)
 
     def _process_js_result(self, obj, allow_dom):
@@ -942,20 +940,20 @@ def last_http_status(self):
     def _frame_to_dict(self, frame, children=True, html=True):
         g = frame.geometry()
         res = {
-            "url": six.text_type(frame.url().toString()),
-            "requestedUrl": six.text_type(frame.requestedUrl().toString()),
+            "url": str(frame.url().toString()),
+            "requestedUrl": str(frame.requestedUrl().toString()),
             "geometry": (g.x(), g.y(), g.width(), g.height()),
-            "title": six.text_type(frame.title())
+            "title": str(frame.title())
         }
         if html:
-            res["html"] = six.text_type(frame.toHtml())
+            res["html"] = str(frame.toHtml())
 
         if children:
             res["childFrames"] = [
                 self._frame_to_dict(f, True, html)
                 for f in frame.childFrames()
             ]
-            res["frameName"] = six.text_type(frame.frameName())
+            res["frameName"] = str(frame.frameName())
 
         return res
 
@@ -1189,7 +1187,7 @@ def __init__(self, parent=None):
 
     @pyqtSlot(str)
     def log(self, message):
-        self.messages.append(six.text_type(message))
+        self.messages.append(str(message))
 
 
 class _BrowserTabLogger(object):
@@ -1239,7 +1237,7 @@ def log(self, message, min_level=None):
         if min_level is not None and self.verbosity < min_level:
             return
 
-        if isinstance(message, six.text_type):
+        if isinstance(message, str):
             message = message.encode('unicode-escape').decode('ascii')
 
         message = "[%s] %s" % (self.uid, message)

splash/conftest.py

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 from splash import lua
 
 collect_ignore = []

splash/cookies.py

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 from PyQt5.QtCore import QDateTime, Qt
 from PyQt5.QtNetwork import QNetworkRequest, QNetworkCookie, QNetworkCookieJar
 

splash/exceptions.py

@@ -1,7 +1,4 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
-
-
 class BadOption(Exception):
     """ Incorrect HTTP API arguments """
     pass

splash/har/__init__.py

@@ -1,4 +1,2 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
-
 from .utils import get_duration, format_datetime

splash/har/log.py

@@ -1,12 +1,10 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 from collections import namedtuple
 from datetime import datetime
 
 import splash
 from PyQt5.QtCore import PYQT_VERSION_STR, QT_VERSION_STR
 from PyQt5.QtWebKit import qWebKitVersion
-import six
 
 from .utils import get_duration, format_datetime, cleaned_har_entry
 
@@ -48,11 +46,11 @@ def has_entry(self, req_id):
 
     def store_url(self, url):
         """ Call this method when URL is changed. """
-        self.events.append(HarEvent(HAR_URL_CHANGED, six.text_type(url)))
+        self.events.append(HarEvent(HAR_URL_CHANGED, str(url)))
 
     def store_title(self, title):
         """ Call this method when page title is changed. """
-        self.events.append(HarEvent(HAR_TITLE_CHANGED, six.text_type(title)))
+        self.events.append(HarEvent(HAR_TITLE_CHANGED, str(title)))
 
     def store_timing(self, name):
         """
@@ -86,12 +84,12 @@ def todict(self):
     def _get_browser(self):
         return {
             "name": "QWebKit",
-            "version": six.text_type(qWebKitVersion()),
+            "version": str(qWebKitVersion()),
             "comment": "PyQt %s, Qt %s" % (PYQT_VERSION_STR, QT_VERSION_STR),
         }
 
     def _empty_page(self, page_id, started_dt):
-        if not isinstance(started_dt, six.string_types):
+        if not isinstance(started_dt, str):
             started_dt = format_datetime(started_dt)
 
         return {

splash/har/qt.py

@@ -3,12 +3,10 @@
 Module with helper utilities to serialize QWebkit objects to HAR.
 See http://www.softwareishard.com/blog/har-12-spec/.
 """
-from __future__ import absolute_import
 import base64
 
 from PyQt5.QtCore import Qt, QVariant, QUrlQuery
 from PyQt5.QtNetwork import QNetworkRequest
-import six
 
 from splash.qtutils import (
     REQUEST_ERRORS_SHORT,
@@ -62,9 +60,9 @@ def cookie2har(cookie):
     cookie = {
         "name": qt_to_bytes(cookie.name()).decode('utf8', 'replace'),
         "value": qt_to_bytes(cookie.value()).decode('utf8', 'replace'),
-        "path": six.text_type(cookie.path()),
-        "domain": six.text_type(cookie.domain()),
-        "expires": six.text_type(cookie.expirationDate().toString(Qt.ISODate)),
+        "path": str(cookie.path()),
+        "domain": str(cookie.domain()),
+        "expires": str(cookie.expirationDate().toString(Qt.ISODate)),
         "httpOnly": cookie.isHttpOnly(),
         "secure": cookie.isSecure(),
     }
@@ -75,7 +73,7 @@ def cookie2har(cookie):
 
 def querystring2har(url):
     return [
-        {"name": six.text_type(name), "value": six.text_type(value)}
+        {"name": str(name), "value": str(value)}
         for name, value in QUrlQuery(url).queryItems()
     ]
 
@@ -105,7 +103,7 @@ def reply2har(reply, content=None):
 
     content_type = reply.header(QNetworkRequest.ContentTypeHeader)
     if content_type is not None:
-        res["content"]["mimeType"] = six.text_type(content_type)
+        res["content"]["mimeType"] = str(content_type)
 
     content_length = reply.header(QNetworkRequest.ContentLengthHeader)
     if content_length is not None:
@@ -120,15 +118,15 @@ def reply2har(reply, content=None):
 
     status_text = reply.attribute(QNetworkRequest.HttpReasonPhraseAttribute)
     if status_text is not None:
-        if not isinstance(status_text, six.text_type):
+        if not isinstance(status_text, str):
             status_text = qt_to_bytes(status_text).decode('latin1')
         res['statusText'] = status_text
     else:
         res["statusText"] = REQUEST_ERRORS_SHORT.get(reply.error(), "?")
 
     redirect_url = reply.attribute(QNetworkRequest.RedirectionTargetAttribute)
     if redirect_url is not None:
-        res["redirectURL"] = six.text_type(redirect_url.toString())
+        res["redirectURL"] = str(redirect_url.toString())
     else:
         res["redirectURL"] = ""
 
@@ -144,7 +142,7 @@ def request2har(request, operation, outgoing_data=None):
     """ Serialize QNetworkRequest to HAR. """
     return {
         "method": OPERATION_NAMES.get(operation, '?'),
-        "url": six.text_type(request.url().toString()),
+        "url": str(request.url().toString()),
         "httpVersion": "HTTP/1.1",
         "cookies": request_cookies2har(request),
         "queryString": querystring2har(request.url()),

splash/har/schema.py

@@ -2,9 +2,6 @@
 """
 Module for validating HAR data. Uses official HAR JSON Schema.
 """
-from __future__ import absolute_import
-
-
 def validate(instance):
     """ Validate HAR data; raise an exception if it is invalid """
     validator = get_validator()

splash/har/utils.py

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 import base64
 from operator import itemgetter
 import itertools

splash/har_builder.py

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 import copy
 from datetime import datetime
 

splash/html_element.py

@@ -1,4 +1,3 @@
-from __future__ import absolute_import
 from functools import wraps
 
 from splash.exceptions import DOMError

splash/kernel/__init__.py

@@ -1,2 +1 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import

splash/kernel/__main__.py

@@ -9,7 +9,6 @@
     $ python -m splash.kernel
 
 """
-from __future__ import absolute_import
 import sys
 from .kernel import start, install
 

splash/kernel/completer.py

@@ -2,7 +2,6 @@
 """
 Autocompleter for Lua code.
 """
-from __future__ import absolute_import
 import string
 
 from splash.utils import dedupe, to_unicode

splash/kernel/errors.py

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 import lupa
 
 from splash.exceptions import ScriptError

splash/kernel/inspections.py

@@ -2,7 +2,6 @@
 """
 Inspections for Lua code.
 """
-from __future__ import absolute_import
 import os
 import glob
 import json

splash/kernel/kernel.py

@@ -1,7 +1,5 @@
 # -*- coding: utf-8 -*-
-from __future__ import absolute_import
 import os
-import six
 import sys
 
 import lupa
@@ -30,7 +28,7 @@
 
 def install(user=True):
     """ Install IPython kernel specification """
-    name = 'splash-py2' if six.PY2 else 'splash-py3'
+    name = 'splash-py3'
     folder = os.path.join(os.path.dirname(__file__), 'kernels', name)
     install_kernel_spec(folder, kernel_name="splash", user=user, replace=True)
 
@@ -168,7 +166,7 @@ def done(result):
             reply, result, ct = result
             if result:
                 data = {
-                    'text/plain': result if isinstance(result, six.text_type) else str(result),
+                    'text/plain': result if isinstance(result, str) else str(result),
                 }
                 if isinstance(result, BinaryCapsule):
                     if result.content_type in {'image/png', 'image/jpeg'}:

splash/kernel/kernelbase.py

@@ -3,7 +3,6 @@
 Refactored from IPython.kernel.zmq.kernelbase.Kernel with async execution
 support. See https://github.com/ipython/ipython/pull/7713.
 """
-from __future__ import absolute_import
 import functools
 import time
 import sys