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) 2000-2023 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 try:
  19     from urllib.parse import splittype, splithost, splittag, urljoin
  20 except ImportError:
  21     from urllib import splittype, splithost, splittag
  22     from urlparse import urljoin
  23
  24 from m_lib.md5wrapper import md5wrapper
  25 from m_lib.net.www.util import parse_time
  26
  27 from bkmk_objects import Robot
  28 from parse_html import parse_html
  29
  30
  31 reloc_dict = {
  32   301: "perm1.",
  33   302: "temp2.",
  34   303: "temp3.",
  35   307: "temp7.",
  36   308: "temp8.",
  37   "html": "html"
  38 }
  39
  40
  41 def get_error(e):
  42     if isinstance(e, str):
  43         return e
  44
  45     else:
  46         s = []
  47         for i in e:
  48             s.append("'%s'" % str(i).replace('\n', "\\n"))
  49         return "(%s)" % ' '.join(s)
  50
  51
  52 # Icon cache; maps URL to a tuple (content type, data)
  53 # or None if there is no icon.
  54 icons = {}
  55
  56
  57 class robot_base(Robot):
  58     timeout = 60
  59
  60     def __init__(self, *args, **kw):
  61         Robot.__init__(self, *args, **kw)
  62         socket.setdefaulttimeout(int(self.timeout))
  63
  64     def check_url(self, bookmark):
  65         try:
  66             self.start = int(time.time())
  67             bookmark.icon = None
  68
  69             url_type, url_rest = splittype(bookmark.href)
  70             url_host, url_path = splithost(url_rest)
  71             url_path, url_tag  = splittag(url_path)  # noqa: E221
  72             #                    multiple spaces before operator
  73
  74             url = "%s://%s%s" % (url_type, url_host, url_path)
  75             error, redirect_code, redirect_to, headers, content = \
  76                 self.get(bookmark, url, True)
  77
  78             if error:
  79                 bookmark.error = error
  80                 return 1
  81
  82             if redirect_code:
  83                 self.set_redirect(bookmark, redirect_code, redirect_to)
  84                 return 1
  85
  86             size = 0
  87             last_modified = None
  88
  89             if headers:
  90                 try:
  91                     size = headers["Content-Length"]
  92                 except KeyError:
  93                     pass
  94
  95                 try:
  96                     last_modified = headers["Last-Modified"]
  97                 except KeyError:
  98                     pass
  99
 100                 if last_modified:
 101                     last_modified = parse_time(last_modified)
 102
 103             if not size:  # Could be None from headers
 104                 size = len(content)
 105
 106             if last_modified:
 107                 last_modified = str(int(last_modified))
 108             else:
 109                 last_modified = bookmark.last_visit
 110
 111             bookmark.size = size
 112             bookmark.last_modified = last_modified
 113
 114             md5 = md5wrapper()
 115             if url_type == "ftp":  # Pass welcome message through MD5
 116                 ftp_welcome = self.get_ftp_welcome()
 117                 if not isinstance(ftp_welcome, bytes):
 118                     ftp_welcome = ftp_welcome.encode('utf-8')
 119                 md5.update(ftp_welcome)
 120
 121             if isinstance(content, bytes):
 122                 md5.update(content)
 123             else:
 124                 md5.update(content.encode('utf-8'))
 125             bookmark.md5 = str(md5)
 126
 127             if headers:
 128                 try:
 129                     content_type = headers["Content-Type"]
 130                     self.log("   Content-Type   : %s" % content_type)
 131                     if content_type is None:
 132                         if 'html' in content.lower():
 133                             content_type = 'text/html'
 134                         else:
 135                             content_type = 'text/plain'
 136                         self.log("   Set Content-Type to: %s"
 137                                  % content_type)
 138                     try:
 139                         # extract charset from
 140                         # "text/html; foo; charset=UTF-8, bar; baz;"
 141                         content_type, charset = content_type.split(';', 1)
 142                         content_type = content_type.strip()
 143                         charset = charset.split('=')[1].strip().split(',')[0]
 144                         self.log("   HTTP charset   : %s" % charset)
 145                     except (ValueError, IndexError):
 146                         charset = None
 147                         self.log("   no charset in Content-Type header")
 148                     is_html = False
 149                     for ctype in ("text/html", "application/xhtml+xml"):
 150                         if content_type.startswith(ctype):
 151                             is_html = True
 152                             break
 153                     content_stripped = content.strip()
 154                     if content_stripped and is_html:
 155                         parser = parse_html(
 156                             content_stripped, charset, self.log)
 157                         if charset:
 158                             bookmark.charset = charset
 159                         elif parser and parser.meta_charset:
 160                             bookmark.charset = parser.meta_charset
 161                         if parser:
 162                             bookmark.real_title = parser.title
 163                             icon = parser.icon
 164                         else:
 165                             icon = None
 166                         if not icon:
 167                             icon = "/favicon.ico"
 168                         icon_url = urljoin(
 169                             "%s://%s%s" % (url_type, url_host, url_path), icon)
 170                         self.log("   looking for icon at: %s" % icon_url)
 171                         if icon_url in icons:
 172                             if icons[icon_url]:
 173                                 bookmark.icon_href = icon_url
 174                                 content_type, bookmark.icon = icons[icon_url]
 175                                 self.log("   cached icon: %s" % content_type)
 176                             else:
 177                                 self.log("   cached icon: no icon")
 178                         else:
 179                             try:
 180                                 _icon_url = icon_url
 181                                 for i in range(8):
 182                                     error, icon_redirect_code, \
 183                                         icon_redirect_to, icon_headers, \
 184                                         icon_data = \
 185                                         self.get(bookmark, _icon_url)
 186                                     if icon_redirect_code:
 187                                         _icon_url = icon_redirect_to
 188                                         self.log("   redirect to : %s"
 189                                                  % _icon_url)
 190                                     else:
 191                                         if icon_data is None:
 192                                             raise IOError("No icon")
 193                                         break
 194                                 else:
 195                                     raise IOError("Too many redirects")
 196                             except:
 197                                 etype, emsg, _ = sys.exc_info()
 198                                 self.log("   no icon        : %s %s"
 199                                          % (etype, emsg))
 200                                 etype = emsg = _ = None
 201                                 icons[icon_url] = None
 202                             else:
 203                                 content_type = icon_headers["Content-Type"]
 204                                 if content_type and (
 205                                     content_type.startswith("application/")
 206                                     or content_type.startswith("image/")
 207                                     or content_type.startswith("text/plain")
 208                                 ):
 209                                     bookmark.icon_href = icon_url
 210                                     self.log("   got icon       : %s"
 211                                              % content_type)
 212                                     if (
 213                                         content_type.startswith("application/")
 214                                         or content_type.startswith(
 215                                             "text/plain")
 216                                     ):
 217                                         self.log("   non-image content type,"
 218                                                  " assume x-icon")
 219                                         content_type = 'image/x-icon'
 220                                     if not isinstance(icon_data, bytes):
 221                                         icon_data = icon_data.encode('utf-8')
 222                                     bookmark.icon = "data:%s;base64,%s" \
 223                                         % (content_type, b64encode(icon_data))
 224                                     icons[icon_url] = (content_type,
 225                                                        bookmark.icon
 226                                                        )
 227                                 else:
 228                                     self.log("   no icon        :"
 229                                              "bad content type '%s'"
 230                                              % content_type
 231                                              )
 232                                     icons[icon_url] = None
 233                         if parser and parser.refresh:
 234                             refresh = parser.refresh
 235                             try:
 236                                 url = refresh.split('=', 1)[1]
 237                             except IndexError:
 238                                 url = "self"
 239                             try:
 240                                 timeout = float(refresh.split(';')[0])
 241                             except (IndexError, ValueError):
 242                                 self.set_redirect(bookmark, "html",
 243                                                   "Bad redirect to %s (%s)"
 244                                                   % (url, refresh)
 245                                                   )
 246                             else:
 247                                 try:
 248                                     timeout = int(refresh.split(';')[0])
 249                                 except ValueError:
 250                                     pass  # float timeout
 251                                 self.set_redirect(bookmark, "html",
 252                                                   "%s (%s sec)"
 253                                                   % (url, timeout)
 254                                                   )
 255
 256                     if not content_stripped:
 257                         self.log("   empty response, no content")
 258                     if not is_html:
 259                         self.log("   not html")
 260                 except KeyError as key:
 261                     self.log("   no header: %s" % key)
 262
 263         except EOFError:
 264             bookmark.error = "Unexpected EOF (FTP server closed connection)"
 265             self.log('   EOF: %s' % bookmark.error)
 266
 267         except KeyboardInterrupt:
 268             self.log("Keyboard interrupt (^C)")
 269             return 0
 270
 271         except socket.error as e:
 272             bookmark.error = get_error(e)
 273             self.log(bookmark.error)
 274
 275         except:
 276             import traceback
 277             traceback.print_exc()
 278             bookmark.error = "Exception!"
 279             self.log('   Exception: %s' % bookmark.error)
 280
 281         finally:
 282             self.finish_check_url(bookmark)
 283
 284         # Tested
 285         return 1
 286
 287     def set_redirect(self, bookmark, errcode, newurl):
 288         bookmark.moved = moved = "(%s) to %s" % (reloc_dict[errcode], newurl)
 289         try:
 290             moved.encode('ascii')
 291         except UnicodeEncodeError:
 292             try:
 293                 moved = moved.encode(bookmark.charset)
 294             except (LookupError, TypeError, UnicodeEncodeError):
 295                 moved = moved.encode('utf-8')
 296         self.log('   Moved: %s' % moved)
 297
 298     def finish_check_url(self, bookmark):
 299         start = self.start
 300         bookmark.last_tested = str(start)
 301         now = int(time.time())
 302         bookmark.test_time = str(now - start)