Extend taggable-mixin to support different Tag models.
Usage is pretty simple. Tag model is default in Taggable constructor but you
can call it with different model like GHOPTaskType that inherits from Tag model.
Both Taggable and Tag models have been updated and they don't use hardcoded Tag
model anymore and instead use cls of class methods or self.__class__. In case
of Taggable it's self.__tag_model.
import re
from django.utils.text import compress_string
from django.utils.cache import patch_vary_headers
re_accepts_gzip = re.compile(r'\bgzip\b')
class GZipMiddleware(object):
"""
This middleware compresses content if the browser allows gzip compression.
It sets the Vary header accordingly, so that caches will base their storage
on the Accept-Encoding header.
"""
def process_response(self, request, response):
# It's not worth compressing non-OK or really short responses.
if response.status_code != 200 or len(response.content) < 200:
return response
patch_vary_headers(response, ('Accept-Encoding',))
# Avoid gzipping if we've already got a content-encoding.
if response.has_header('Content-Encoding'):
return response
# Older versions of IE have issues with gzipped pages containing either
# Javascript and PDF.
if "msie" in request.META.get('HTTP_USER_AGENT', '').lower():
ctype = response.get('Content-Type', '').lower()
if "javascript" in ctype or ctype == "application/pdf":
return response
ae = request.META.get('HTTP_ACCEPT_ENCODING', '')
if not re_accepts_gzip.search(ae):
return response
response.content = compress_string(response.content)
response['Content-Encoding'] = 'gzip'
response['Content-Length'] = str(len(response.content))
return response