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