[fix] set encoding of README.rst explicit in setup.py (utf-8) (#2117)

## What does this PR do?

When installing searx on a centos-7 image [1] an encoding exception is raised from setup.py::

    command: /usr/local/searx/searx-pyenv/bin/python3 -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/usr/local/searx/searx-src/setup.py'"'"'; __file__='"'"'/usr/local/searx/searx-src/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(__file__);code=f.read().replace('"'"'rn'"'"', '"'"'n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' egg_info --egg-base /tmp/pip-pip-egg-info-pidntzkq
         cwd: /usr/local/searx/searx-src/
    Complete output (7 lines):
    Traceback (most recent call last):
      File "<string>", line 1, in <module>
      File "/usr/local/searx/searx-src/setup.py", line 16, in <module>
        long_description = f.read()
      File "/usr/lib64/python3.6/encodings/ascii.py", line 26, in decode
        return codecs.ascii_decode(input, self.errors)[0]
    UnicodeDecodeError: 'ascii' codec can't decode byte 0xc9 in position 482: ordinal not in range(128)

[1] https://uk.images.linuxcontainers.org/

## Why is this change important?

README.rst uses non ASCII characters _(pronunciation səːks.)_

## How to test this PR locally?

install searx on a centos-7 image from https://uk.images.linuxcontainers.org/

setup.py

@@ -12,7 +12,7 @@
 from version import VERSION_STRING
 import brand
 
-with open('README.rst') as f:
+with open('README.rst', encoding='utf-8') as f:
     long_description = f.read()
 
 with open('requirements.txt') as f: