Robots/bkmk_robot_base.py

   1 """Base class for robots
   2
   3 This file is a part of Bookmarks database and Internet robot.
   4
   5 """
   6
   7 __author__ = "Oleg Broytman <phd@phdru.name>"
   8 __copyright__ = "Copyright (C) 2000-2014 PhiloSoft Design"
   9 __license__ = "GNU GPL"
  10
  11 __all__ = ['robot_base', 'get_error']
  12
  13
  14 import sys
  15 import time, urllib
  16 from base64 import b64encode
  17 from urlparse import urljoin
  18
  19 from m_lib.net.www.util import parse_time
  20 from m_lib.md5wrapper import md5wrapper
  21
  22 from bkmk_objects import Robot
  23 from parse_html import parse_html
  24
  25
  26 class RedirectException(Exception):
  27    reloc_dict = {
  28       301: "perm.",
  29       302: "temp2.",
  30       303: "temp3.",
  31       307: "temp7.",
  32       "html": "html"
  33    }
  34    def __init__(self, errcode, newurl):
  35       Exception.__init__(self, "(%s) to %s" % (self.reloc_dict[errcode], newurl))
  36       self.url = newurl
  37
  38
  39 def get_error(msg):
  40    if isinstance(msg, str):
  41       return msg
  42
  43    else:
  44       s = []
  45       for i in msg:
  46          s.append("'%s'" % str(i).replace('\n', "\\n"))
  47       return "(%s)" % ' '.join(s)
  48
  49
  50 icons = {} # Icon cache; maps URL to a tuple (content type, data)
  51            # or None if there is no icon.
  52
  53 class robot_base(Robot):
  54    def check_url(self, bookmark):
  55       try:
  56          self.start = int(time.time())
  57          bookmark.icon = None
  58
  59          url_type, url_rest = urllib.splittype(bookmark.href)
  60          url_host, url_path = urllib.splithost(url_rest)
  61          url_path, url_tag  = urllib.splittag(url_path)
  62
  63          url = "%s://%s%s" % (url_type, url_host, url_path)
  64          error, headers, content = self.get(bookmark, url, True)
  65
  66          if error:
  67              bookmark.error = error
  68              return 1
  69
  70          size = 0
  71          last_modified = None
  72
  73          if headers:
  74             try:
  75                size = headers["Content-Length"]
  76             except KeyError:
  77                size = len(content)
  78
  79             try:
  80                last_modified = headers["Last-Modified"]
  81             except KeyError:
  82                pass
  83
  84             if last_modified:
  85                last_modified = parse_time(last_modified)
  86          else:
  87             size = len(content)
  88
  89          if last_modified:
  90             last_modified = str(int(last_modified))
  91          else:
  92             last_modified = bookmark.last_visit
  93
  94          bookmark.size = size
  95          bookmark.last_modified = last_modified
  96
  97          md5 = md5wrapper()
  98          if urllib._urlopener.type == "ftp": # Pass welcome message through MD5
  99             md5.update(self.get_ftp_welcome())
 100
 101          md5.update(content)
 102          bookmark.md5 = str(md5)
 103
 104          if headers:
 105             try:
 106                content_type = headers["Content-Type"]
 107                self.log("   Content-Type: %s" % content_type)
 108                try:
 109                   # extract charset from "text/html; foo; charset=UTF-8, bar; baz;"
 110                   content_type, charset = content_type.split(';', 1)
 111                   content_type = content_type.strip()
 112                   charset = charset.split('=')[1].strip().split(',')[0]
 113                   self.log("   HTTP charset   : %s" % charset)
 114                except (ValueError, IndexError):
 115                   charset = None
 116                   self.log("   no charset in Content-Type header")
 117                for ctype in ("text/html", "application/xhtml+xml"):
 118                   if content_type.startswith(ctype):
 119                       html = True
 120                       break
 121                else:
 122                   html = False
 123                if html:
 124                   parser = parse_html(content, charset, self.log)
 125                   if parser:
 126                       bookmark.real_title = parser.title
 127                       icon = parser.icon
 128                   else:
 129                      icon = None
 130                   if not icon:
 131                      icon = "/favicon.ico"
 132                   icon_url = urljoin("%s://%s%s" % (url_type, url_host, url_path), icon)
 133                   self.log("   looking for icon at: %s" % icon_url)
 134                   if icon_url in icons:
 135                      if icons[icon_url]:
 136                         bookmark.icon_href = icon_url
 137                         content_type, bookmark.icon = icons[icon_url]
 138                         self.log("   cached icon: %s" % content_type)
 139                      else:
 140                         self.log("   cached icon: no icon")
 141                   else:
 142                      try:
 143                         _icon_url = icon_url
 144                         for i in range(8):
 145                            try:
 146                               error, icon_headers, icon_data = self.get(bookmark, _icon_url)
 147                            except RedirectException, e:
 148                               _icon_url = e.url
 149                               self.log("   redirect to : %s" % _icon_url)
 150                            else:
 151                               if icon_data is None:
 152                                    raise IOError("No icon")
 153                               break
 154                         else:
 155                            raise IOError("Too many redirects")
 156                      except:
 157                         etype, emsg, tb = sys.exc_info()
 158                         self.log("   no icon        : %s %s" % (etype, emsg))
 159                         etype = emsg = tb = None
 160                         icons[icon_url] = None
 161                      else:
 162                         content_type = icon_headers["Content-Type"]
 163                         if content_type.startswith("application/") \
 164                               or content_type.startswith("image/") \
 165                               or content_type.startswith("text/plain"):
 166                            bookmark.icon_href = icon_url
 167                            self.log("   got icon       : %s" % content_type)
 168                            if content_type.startswith("application/") \
 169                                  or content_type.startswith("text/plain"):
 170                               self.log("   non-image content type, assume x-icon")
 171                               content_type = 'image/x-icon'
 172                            bookmark.icon = "data:%s;base64,%s" % (content_type, b64encode(icon_data))
 173                            icons[icon_url] = (content_type, bookmark.icon)
 174                         else:
 175                            self.log("   no icon        : bad content type '%s'" % content_type)
 176                            icons[icon_url] = None
 177                   if parser and parser.refresh:
 178                      refresh = parser.refresh
 179                      try:
 180                         url = refresh.split('=', 1)[1]
 181                      except IndexError:
 182                         url = "self"
 183                      try:
 184                         timeout = float(refresh.split(';')[0])
 185                      except (IndexError, ValueError):
 186                         raise RedirectException("html", "Bad redirect to %s (%s)" % (url, refresh))
 187                      else:
 188                         try:
 189                            timeout = int(refresh.split(';')[0])
 190                         except ValueError:
 191                            pass # float timeout
 192                         raise RedirectException("html", "%s (%s sec)" % (url, timeout))
 193
 194             except KeyError, key:
 195                self.log("   no header: %s" % key)
 196
 197       except EOFError:
 198          bookmark.error = "Unexpected EOF (FTP server closed connection)"
 199          self.log('   EOF: %s' % bookmark.error)
 200
 201       except RedirectException, msg:
 202          bookmark.moved = str(msg)
 203          self.log('   Moved: %s' % bookmark.moved)
 204
 205       except KeyboardInterrupt:
 206          self.log("Keyboard interrupt (^C)")
 207          return 0
 208
 209       except:
 210          import traceback
 211          traceback.print_exc()
 212          bookmark.error = "Exception!"
 213          self.log('   Exception: %s' % bookmark.error)
 214
 215       finally:
 216          self.finish_check_url(bookmark)
 217
 218       # Tested
 219       return 1
 220
 221    def finish_check_url(self, bookmark):
 222       start = self.start
 223       bookmark.last_tested = str(start)
 224
 225       now = int(time.time())
 226       bookmark.test_time = str(now - start)
 227
 228       self.cleanup()