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) 2000-2017 PhiloSoft Design"
   9 __license__ = "GNU GPL"
  10
  11 __all__ = ['robot_base', 'get_error']
  12
  13
  14 from base64 import b64encode
  15 import sys
  16 import socket
  17 import time
  18 import urllib
  19 from urlparse import urljoin
  20
  21 from m_lib.md5wrapper import md5wrapper
  22 from m_lib.net.www.util import parse_time
  23
  24 from bkmk_objects import Robot
  25 from parse_html import parse_html
  26
  27
  28 reloc_dict = {
  29   301: "perm.",
  30   302: "temp2.",
  31   303: "temp3.",
  32   307: "temp7.",
  33   "html": "html"
  34 }
  35
  36
  37 def get_error(e):
  38     if isinstance(e, str):
  39         return e
  40
  41     else:
  42         s = []
  43         for i in e:
  44             s.append("'%s'" % str(i).replace('\n', "\\n"))
  45         return "(%s)" % ' '.join(s)
  46
  47
  48 # Icon cache; maps URL to a tuple (content type, data)
  49 # or None if there is no icon.
  50 icons = {}
  51
  52
  53 class robot_base(Robot):
  54     timeout = 60
  55
  56     def __init__(self, *args, **kw):
  57         Robot.__init__(self, *args, **kw)
  58         socket.setdefaulttimeout(int(self.timeout))
  59
  60     def check_url(self, bookmark):
  61         try:
  62             self.start = int(time.time())
  63             bookmark.icon = None
  64
  65             url_type, url_rest = urllib.splittype(bookmark.href)
  66             url_host, url_path = urllib.splithost(url_rest)
  67             url_path, url_tag  = urllib.splittag(url_path)
  68
  69             url = "%s://%s%s" % (url_type, url_host, url_path)
  70             error, redirect_code, redirect_to, headers, content = self.get(bookmark, url, True)
  71
  72             if error:
  73                 bookmark.error = error
  74                 return 1
  75
  76             if redirect_code:
  77                 self.set_redirect(bookmark, redirect_code, redirect_to)
  78                 return 1
  79
  80             size = 0
  81             last_modified = None
  82
  83             if headers:
  84                 try:
  85                     size = headers["Content-Length"]
  86                 except KeyError:
  87                     size = len(content)
  88
  89                 try:
  90                     last_modified = headers["Last-Modified"]
  91                 except KeyError:
  92                     pass
  93
  94                 if last_modified:
  95                     last_modified = parse_time(last_modified)
  96             else:
  97                 size = len(content)
  98
  99             if last_modified:
 100                 last_modified = str(int(last_modified))
 101             else:
 102                 last_modified = bookmark.last_visit
 103
 104             bookmark.size = size
 105             bookmark.last_modified = last_modified
 106
 107             md5 = md5wrapper()
 108             if url_type == "ftp": # Pass welcome message through MD5
 109                 md5.update(self.get_ftp_welcome())
 110
 111             md5.update(content)
 112             bookmark.md5 = str(md5)
 113
 114             if headers:
 115                 try:
 116                     content_type = headers["Content-Type"]
 117                     self.log("   Content-Type: %s" % content_type)
 118                     try:
 119                         # extract charset from "text/html; foo; charset=UTF-8, bar; baz;"
 120                         content_type, charset = content_type.split(';', 1)
 121                         content_type = content_type.strip()
 122                         charset = charset.split('=')[1].strip().split(',')[0]
 123                         self.log("   HTTP charset   : %s" % charset)
 124                     except (ValueError, IndexError):
 125                         charset = None
 126                         self.log("   no charset in Content-Type header")
 127                     for ctype in ("text/html", "application/xhtml+xml"):
 128                         if content_type.startswith(ctype):
 129                             html = True
 130                             break
 131                     else:
 132                         html = False
 133                     if html:
 134                         parser = parse_html(content, charset, self.log)
 135                         if parser:
 136                             bookmark.real_title = parser.title
 137                             icon = parser.icon
 138                         else:
 139                             icon = None
 140                         if not icon:
 141                             icon = "/favicon.ico"
 142                         icon_url = urljoin("%s://%s%s" % (url_type, url_host, url_path), icon)
 143                         self.log("   looking for icon at: %s" % icon_url)
 144                         if icon_url in icons:
 145                             if icons[icon_url]:
 146                                 bookmark.icon_href = icon_url
 147                                 content_type, bookmark.icon = icons[icon_url]
 148                                 self.log("   cached icon: %s" % content_type)
 149                             else:
 150                                 self.log("   cached icon: no icon")
 151                         else:
 152                             try:
 153                                 _icon_url = icon_url
 154                                 for i in range(8):
 155                                     error, icon_redirect_code, icon_redirect_to, \
 156                                        icon_headers, icon_data = \
 157                                           self.get(bookmark, _icon_url)
 158                                     if icon_redirect_code:
 159                                         _icon_url = icon_redirect_to
 160                                         self.log("   redirect to : %s" % _icon_url)
 161                                     else:
 162                                         if icon_data is None:
 163                                             raise IOError("No icon")
 164                                         break
 165                                 else:
 166                                     raise IOError("Too many redirects")
 167                             except:
 168                                 etype, emsg, tb = sys.exc_info()
 169                                 self.log("   no icon        : %s %s" % (etype, emsg))
 170                                 etype = emsg = tb = None
 171                                 icons[icon_url] = None
 172                             else:
 173                                 content_type = icon_headers["Content-Type"]
 174                                 if content_type.startswith("application/") \
 175                                       or content_type.startswith("image/") \
 176                                       or content_type.startswith("text/plain"):
 177                                     bookmark.icon_href = icon_url
 178                                     self.log("   got icon       : %s" % content_type)
 179                                     if content_type.startswith("application/") \
 180                                           or content_type.startswith("text/plain"):
 181                                         self.log("   non-image content type, assume x-icon")
 182                                         content_type = 'image/x-icon'
 183                                     bookmark.icon = "data:%s;base64,%s" % (content_type, b64encode(icon_data))
 184                                     icons[icon_url] = (content_type, bookmark.icon)
 185                                 else:
 186                                     self.log("   no icon        : bad content type '%s'" % content_type)
 187                                     icons[icon_url] = None
 188                         if parser and parser.refresh:
 189                             refresh = parser.refresh
 190                             try:
 191                                 url = refresh.split('=', 1)[1]
 192                             except IndexError:
 193                                 url = "self"
 194                             try:
 195                                 timeout = float(refresh.split(';')[0])
 196                             except (IndexError, ValueError):
 197                                 self.set_redirect(bookmark, "html", "Bad redirect to %s (%s)" % (url, refresh))
 198                             else:
 199                                 try:
 200                                     timeout = int(refresh.split(';')[0])
 201                                 except ValueError:
 202                                     pass # float timeout
 203                                 self.set_redirect(bookmark, "html", "%s (%s sec)" % (url, timeout))
 204
 205                 except KeyError as key:
 206                     self.log("   no header: %s" % key)
 207
 208         except EOFError:
 209             bookmark.error = "Unexpected EOF (FTP server closed connection)"
 210             self.log('   EOF: %s' % bookmark.error)
 211
 212         except KeyboardInterrupt:
 213             self.log("Keyboard interrupt (^C)")
 214             return 0
 215
 216         except socket.error as e:
 217             bookmark.error = get_error(e)
 218             self.log(bookmark.error)
 219
 220         except:
 221             import traceback
 222             traceback.print_exc()
 223             bookmark.error = "Exception!"
 224             self.log('   Exception: %s' % bookmark.error)
 225
 226         finally:
 227             self.finish_check_url(bookmark)
 228
 229         # Tested
 230         return 1
 231
 232     def set_redirect(self, bookmark, errcode, newurl):
 233         bookmark.moved = "(%s) to %s" % (reloc_dict[errcode], newurl)
 234         self.log('   Moved: %s' % bookmark.moved)
 235
 236     def finish_check_url(self, bookmark):
 237         start = self.start
 238         bookmark.last_tested = str(start)
 239         now = int(time.time())
 240         bookmark.test_time = str(now - start)