Python-сообщество

# A basic cache

#   Saving: every time we see a page, we save it to something in /tmp/cache;

#   Reusing: every time we see a request that doesn't have pragma: no-cache,
#   we look in /tmp/cache for the document, and if it's there we serve that

# What to do about:

#   Cache-control: private
#   Cache-Control: No-Cache

# Also, watch for cookie changes

# Browser header ``pragma: no-cache'' is used both for bookmarks and
# for reload button, so how do we distinguish?? :(

from string import *
from proxy3_util import *
from proxy3_filter import *
import os

# Make sure the directory exists
try: os.mkdir('/tmp/cache')
except: pass

CONTENT = {} # URL -> content

class Cache(BufferSomeFilter):
    BUFFER_LEN = 0 # We can't deal with filtering at all
    page_content = None

    def filter(self, s):
        # Save everything that goes by
        if not self.page_content:
            # Insert the headers the first time through
            self.page_content = [join(self.args['serverheaders'].headers,''), '\r\n']
        self.page_content.append(s)

        return s
    
    def close(self):
        BufferSomeFilter.close(self)
        
        url = self.args['url']
        page_content = join(self.page_content, '')
        CONTENT[url]= page_content
        print 'saving', url
        
def event_process_request(obj, site, document, request, headers):
    if (find(headers.getheader('Pragma', ''), 'no-cache') < 0 and
        not headers.getheader('Content-length', '') and
        CONTENT.has_key(site+document)):
        # We have the content, and it's not marked as "no-cache" by the browser
        print 'reusing', site+document
        obj.outfile.write('HTTP/1.0 200 OK\r\n')
        obj.outfile.write(CONTENT[site+document])
        return []
    
register_filter('*', 'text/html', Cache)
register_filter('*', 'text/plain', Cache)
register_filter('*', 'image/gif', Cache)
register_filter('*', 'image/jpg', Cache)