#!/usr/bin/python
from config import directory
import os, os.path, time

#########################################################################
#		       Remove items from the cache			#
#########################################################################

def clear(entry=None):
  from glob import glob
  files = glob(directory.cache+"index.*") + glob(directory.cache+"all.*")
  date=time.localtime()

  if entry:
    if directory.cache+"index.rss" in files:
      files.remove(directory.cache+"index.rss")
    files += glob(directory.cache+"comments.*")
    files += glob(directory.cache+entry+".*")
    date=time.localtime(os.stat(directory.data + entry+".txt").st_mtime)

    from entry import post
    link=directory.cache+post(entry).link()
    files += glob('/'.join(link.split('/')[:-1]+['*']))

    from atomize import atomize
    atomize(entry).close()

    import search
    search.decache(entry)

  date = time.strftime("%Y %m %d",date).split()
  files += [directory.cache+"/index.html5"]
  for i in range(0,4):
    path="/".join(date[:i])
    files += [directory.cache+path+"/index.html"]
    files += [directory.cache+path+"/index.atom"]
    if not entry: 
      files += [directory.cache+"archives/"+path+"/index.html"]
      files += [directory.cache+"archives/"+path+"/index.atom"]

  for file in files:
    if os.path.exists(file):
     if os.path.isdir(file):
       os.remove(file+"/index.html")
       if not os.listdir(file): os.rmdir(file)
     else:
       os.remove(file)

#########################################################################
#			Add an item to the cache			#
#########################################################################

def save(path, data):
  data=str(data)
  if path.endswith('/'): return data

  if isinstance(path,list): path="/".join(path)
  if path[0]=='/': path=path[1:]

  mime_boundary = data.find('\n\n')

  if data[:mime_boundary].split()[0].lower()<>'status:':

    # create any required directories
    dir=path[:path.rfind('/')+1]
    if dir and not os.path.exists(directory.cache+dir): 
      dir=dir.split('/')
      for i in range(0,len(dir)):
        parent="/".join(dir[:i+1])
        if not os.path.exists(directory.cache+parent):
          try:
            os.mkdir(directory.cache+parent)
          except:
            pass # timing issue
          os.chmod(directory.cache+parent, 0777)

    # safety first
    if os.path.exists(directory.cache+path): return data

    # save the data to the cache directory
    file = open(directory.cache+path, "w")
    file.write(data[mime_boundary+2:])
    file.close()
    os.chmod(directory.cache+path, 0666)

    if path=='index.atom': delta("index")
    if path=='comments.atom': delta("comments")

  return data
  
#########################################################################
#		 Precompute deltas for RFC 3229 support			#
#########################################################################

def delta(feed):
  from glob import glob
  import httplib, re

  # get current page
  connection = httplib.HTTPConnection("intertwingly.net")
  connection.request("GET", "/blog/%s.atom" % feed)
  response = connection.getresponse()
  headers = response.msg.items()
  etag = response.getheader('ETag')

  # etag is vital to this... remove W/ prefix
  if not etag: return
  etag=etag[etag.find('"'):]

  # divide feed into entries
  pattern=re.compile("(\s*<entry>.*?</entry>)",re.DOTALL)
  latest=re.split(pattern,response.read())
  if not latest: return

  # remove existing "asis" files... updating the base files access times
  for asisfile in glob(directory.history + feed + '*.asis'):
    if os.path.exists(asisfile): 
      asis_stats=os.stat(asisfile)
      os.remove(asisfile)
      basefile=asisfile.replace('.asis','.base')
      if os.path.exists(basefile): 
        base_stats=os.stat(basefile)
        os.utime(basefile,(asis_stats.st_atime,base_stats.st_mtime))

  # for each basefile in history
  for basefile in glob(directory.history + feed + '*.base'):

    # don't compute delta on current file
    if etag and basefile.find(etag)>=0: continue

    # parse and split basefile, retaining stats 
    stats=os.stat(basefile)
    base=re.split(pattern, open(basefile).read())
    os.utime(basefile,(stats.st_atime,stats.st_mtime))

    # asis = latest - base
    asis=latest[:]
    for i in range(1,len(asis),2):
      if asis[i] in base: asis[i]=''
    asis=''.join(asis)

    # write out asis file, with headers
    asisfile=open(basefile.replace('.base','.asis'),'w')
    asisfile.write("Status: 226 IM Used\r\n")
    for (name,value) in headers:
      if name.lower()=='etag': value=etag
      if name.lower()=='content-length': value=str(len(asis))
      if name.lower()=='vary': value += ",If-None-Match"
      name='-'.join(map(lambda(x):x.capitalize(), name.split('-')))
      asisfile.write("%s: %s\r\n" % (name, value))
    asisfile.write("IM: feed\r\n")
    asisfile.write("\r\n%s" % asis)
    asisfile.close()

  # write out new base file
  # basefile=open('%s%s.%s.base' % (directory.history, feed, etag), 'w')
  # basefile.write(''.join(latest))
  # basefile.close()

#########################################################################
#                     Serve an item from the cache			#
#########################################################################

def serve(path, query=None):
  import search
  if search.isStale(path, query): return None

  from httplib import HTTP
  h=HTTP('www.intertwingly.net')
  h.putrequest('GET','/blog/'+path)
  h.putheader('Host','www.intertwingly.net')
  h.putheader('Accept-Encoding',os.environ.get('HTTP_ACCEPT_ENCODING',''))
  h.putheader('Accept',os.environ.get('HTTP_ACCEPT',''))
  h.endheaders()
  (code,msg,headers)=h.getreply()
  if code>=400:
    return None
  else:
    result="Status: %s %s\r\n%s\r\n" % (code,msg,''.join(headers.headers))
    result+=h.file.read()
    h.file.close()
    return result
  
#########################################################################
#	      Main: clear specified entries from the cache		#
#########################################################################

if __name__ == '__main__':
  import sys
  if len(sys.argv)==1:
    clear()
  else:
     for entry in sys.argv[1:]:
       clear(entry)