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          headers, content, error = self.urlretrieve(bookmark, url, True)
  65
  66          if error:
  67              bookmark.error = error
  68
  69          if content is None:
  70              return 1
  71
  72          size = 0
  73          last_modified = None
  74
  75          if headers:
  76             try:
  77                size = headers["Content-Length"]
  78             except KeyError:
  79                size = len(content)
  80
  81             try:
  82                last_modified = headers["Last-Modified"]
  83             except KeyError:
  84                pass
  85
  86             if last_modified:
  87                last_modified = parse_time(last_modified)
  88          else:
  89             size = len(content)
  90
  91          if last_modified:
  92             last_modified = str(int(last_modified))
  93          else:
  94             last_modified = bookmark.last_visit
  95
  96          bookmark.size = size
  97          bookmark.last_modified = last_modified
  98
  99          md5 = md5wrapper()
 100          if urllib._urlopener.type == "ftp": # Pass welcome message through MD5
 101             md5.update(self.get_ftp_welcome())
 102
 103          md5.update(content)
 104          bookmark.md5 = str(md5)
 105
 106          if headers:
 107             try:
 108                content_type = headers["Content-Type"]
 109                self.log("   Content-Type: %s" % content_type)
 110                try:
 111                   # extract charset from "text/html; foo; charset=UTF-8, bar; baz;"
 112                   content_type, charset = content_type.split(';', 1)
 113                   content_type = content_type.strip()
 114                   charset = charset.split('=')[1].strip().split(',')[0]
 115                   self.log("   HTTP charset   : %s" % charset)
 116                except (ValueError, IndexError):
 117                   charset = None
 118                   self.log("   no charset in Content-Type header")
 119                for ctype in ("text/html", "application/xhtml+xml"):
 120                   if content_type.startswith(ctype):
 121                       html = True
 122                       break
 123                else:
 124                   html = False
 125                if html:
 126                   parser = parse_html(content, charset, self.log)
 127                   if parser:
 128                       bookmark.real_title = parser.title
 129                       icon = parser.icon
 130                   else:
 131                      icon = None
 132                   if not icon:
 133                      icon = "/favicon.ico"
 134                   icon_url = urljoin("%s://%s%s" % (url_type, url_host, url_path), icon)
 135                   self.log("   looking for icon at: %s" % icon_url)
 136                   if icon_url in icons:
 137                      if icons[icon_url]:
 138                         bookmark.icon_href = icon_url
 139                         content_type, bookmark.icon = icons[icon_url]
 140                         self.log("   cached icon: %s" % content_type)
 141                      else:
 142                         self.log("   cached icon: no icon")
 143                   else:
 144                      try:
 145                         _icon_url = icon_url
 146                         for i in range(8):
 147                            try:
 148                               icon_headers, icon_data, error = self.urlretrieve(bookmark, _icon_url)
 149                            except RedirectException, e:
 150                               _icon_url = e.url
 151                               self.log("   redirect to : %s" % _icon_url)
 152                            else:
 153                               if icon_data is None:
 154                                    raise IOError("No icon")
 155                               break
 156                         else:
 157                            raise IOError("Too many redirects")
 158                      except:
 159                         etype, emsg, tb = sys.exc_info()
 160                         self.log("   no icon        : %s %s" % (etype, emsg))
 161                         etype = emsg = tb = None
 162                         icons[icon_url] = None
 163                      else:
 164                         content_type = icon_headers["Content-Type"]
 165                         if content_type.startswith("application/") \
 166                               or content_type.startswith("image/") \
 167                               or content_type.startswith("text/plain"):
 168                            bookmark.icon_href = icon_url
 169                            self.log("   got icon       : %s" % content_type)
 170                            if content_type.startswith("application/") \
 171                                  or content_type.startswith("text/plain"):
 172                               self.log("   non-image content type, assume x-icon")
 173                               content_type = 'image/x-icon'
 174                            bookmark.icon = "data:%s;base64,%s" % (content_type, b64encode(icon_data))
 175                            icons[icon_url] = (content_type, bookmark.icon)
 176                         else:
 177                            self.log("   no icon        : bad content type '%s'" % content_type)
 178                            icons[icon_url] = None
 179                   if parser and parser.refresh:
 180                      refresh = parser.refresh
 181                      try:
 182                         url = refresh.split('=', 1)[1]
 183                      except IndexError:
 184                         url = "self"
 185                      try:
 186                         timeout = float(refresh.split(';')[0])
 187                      except (IndexError, ValueError):
 188                         raise RedirectException("html", "Bad redirect to %s (%s)" % (url, refresh))
 189                      else:
 190                         try:
 191                            timeout = int(refresh.split(';')[0])
 192                         except ValueError:
 193                            pass # float timeout
 194                         raise RedirectException("html", "%s (%s sec)" % (url, timeout))
 195
 196             except KeyError, key:
 197                self.log("   no header: %s" % key)
 198
 199       except EOFError:
 200          bookmark.error = "Unexpected EOF (FTP server closed connection)"
 201          self.log('   EOF: %s' % bookmark.error)
 202
 203       except RedirectException, msg:
 204          bookmark.moved = str(msg)
 205          self.log('   Moved: %s' % bookmark.moved)
 206
 207       except KeyboardInterrupt:
 208          self.log("Keyboard interrupt (^C)")
 209          return 0
 210
 211       except:
 212          import traceback
 213          traceback.print_exc()
 214          bookmark.error = "Exception!"
 215          self.log('   Exception: %s' % bookmark.error)
 216
 217       finally:
 218          self.finish_check_url(bookmark)
 219
 220       # Tested
 221       return 1
 222
 223    def finish_check_url(self, bookmark):
 224       start = self.start
 225       bookmark.last_tested = str(start)
 226
 227       now = int(time.time())
 228       bookmark.test_time = str(now - start)
 229
 230       self.cleanup()