Use stdlib functools.cached_property in Python 3.8+

Python 3.8+ has a functools.cached_property decorator in stdlib.  Use it
in place of third-party cached_property when available.

gixy/core/regexp.py

@@ -3,6 +3,10 @@ import logging
 import re
 import random
 import itertools
+
+try:
+    from functools import cached_property
+except ImportError:
     from cached_property import cached_property
 
 import gixy.core.sre_parse.sre_parse as sre_parse

gixy/directives/block.py

@@ -1,3 +1,6 @@
+try:
+    from functools import cached_property
+except ImportError:
     from cached_property import cached_property
 
 from gixy.directives.directive import Directive

gixy/parser/raw_parser.py

@@ -1,6 +1,10 @@
 import logging
 import codecs
 import six
+
+try:
+    from functools import cached_property
+except ImportError:
     from cached_property import cached_property
 
 from pyparsing import (

requirements.txt

@@ -1,5 +1,5 @@
 pyparsing>=1.5.5
-cached-property>=1.2.0
+cached-property>=1.2.0;python_version<"3.8"
 argparse>=1.4.0
 six>=1.1.0
 Jinja2>=2.8

setup.py

@@ -18,7 +18,7 @@ setup(
     url='https://github.com/yandex/gixy',
     install_requires=[
         'pyparsing>=1.5.5',
-        'cached-property>=1.2.0',
+        'cached-property>=1.2.0;python_version<"3.8"',
         'argparse>=1.4.0;python_version<"3.2"',
         'six>=1.1.0',
         'Jinja2>=2.8',