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