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             charset = None
 115             if headers:
 116                 try:
 117                     content_type = headers["Content-Type"]
 118                     self.log("   Content-Type   : %s" % content_type)
 119                     if content_type is None:
 120                         if b'html' in content.lower():
 121                             content_type = 'text/html'
 122                         else:
 123                             content_type = 'text/plain'
 124                         self.log("   Set Content-Type to: %s"
 125                                  % content_type)
 126                     try:
 127                         # extract charset from
 128                         # "text/html; charset=UTF-8, foo; bar"
 129                         content_type, charset = content_type.split(';', 1)
 130                         content_type = content_type.strip()
 131                         charset = charset.split('=')[1].strip().split(',')[0]
 132                         self.log("   HTTP charset   : %s" % charset)
 133                     except (ValueError, IndexError):
 134                         charset = None
 135                         self.log("   no charset in Content-Type header")
 136                     is_html = False
 137                     for ctype in ("text/html", "application/xhtml+xml"):
 138                         if content_type.startswith(ctype):
 139                             is_html = True
 140                             break
 141                     content_stripped = content.strip()
 142                     if content_stripped and charset:
 143                         content_stripped = content_stripped.decode(
 144                             charset, 'replace')
 145                     if content_stripped and is_html:
 146                         parser = parse_html(
 147                             content_stripped, charset, self.log)
 148                         if charset:
 149                             bookmark.charset = charset
 150                         elif parser and parser.meta_charset:
 151                             bookmark.charset = parser.meta_charset
 152                         if parser:
 153                             bookmark.real_title = parser.title
 154                             icon = parser.icon
 155                         else:
 156                             icon = None
 157                         if not icon:
 158                             icon = "/favicon.ico"
 159                         icon_url = urljoin(
 160                             "%s://%s%s" % (url_type, url_host, url_path), icon)
 161                         self.log("   looking for icon at: %s" % icon_url)
 162                         if icon_url in icons:
 163                             if icons[icon_url]:
 164                                 bookmark.icon_href = icon_url
 165                                 content_type, bookmark.icon = icons[icon_url]
 166                                 self.log("   cached icon: %s" % content_type)
 167                             else:
 168                                 self.log("   cached icon: no icon")
 169                         else:
 170                             try:
 171                                 _icon_url = icon_url
 172                                 for i in range(8):
 173                                     error, icon_redirect_code, \
 174                                         icon_redirect_to, icon_headers, \
 175                                         icon_data = \
 176                                         self.get(bookmark, _icon_url)
 177                                     if icon_redirect_code:
 178                                         _icon_url = icon_redirect_to
 179                                         self.log("   redirect to : %s"
 180                                                  % _icon_url)
 181                                     else:
 182                                         if icon_data is None:
 183                                             raise IOError("No icon")
 184                                         break
 185                                 else:
 186                                     raise IOError("Too many redirects")
 187                             except:
 188                                 etype, emsg, _ = sys.exc_info()
 189                                 self.log("   no icon        : %s %s"
 190                                          % (etype, emsg))
 191                                 etype = emsg = _ = None
 192                                 icons[icon_url] = None
 193                             else:
 194                                 content_type = icon_headers["Content-Type"]
 195                                 if content_type and (
 196                                     content_type.startswith("application/")
 197                                     or content_type.startswith("image/")
 198                                     or content_type.startswith("text/plain")
 199                                 ):
 200                                     bookmark.icon_href = icon_url
 201                                     self.log("   got icon       : %s"
 202                                              % content_type)
 203                                     if (
 204                                         content_type.startswith("application/")
 205                                         or content_type.startswith(
 206                                             "text/plain")
 207                                     ):
 208                                         self.log("   non-image content type,"
 209                                                  " assume x-icon")
 210                                         content_type = 'image/x-icon'
 211                                     if not isinstance(icon_data, bytes):
 212                                         icon_data = icon_data.encode('latin1')
 213                                     bookmark.icon = "data:%s;base64,%s" \
 214                                         % (content_type, b64encode(icon_data))
 215                                     icons[icon_url] = (content_type,
 216                                                        bookmark.icon
 217                                                        )
 218                                 else:
 219                                     self.log("   no icon        :"
 220                                              "bad content type '%s'"
 221                                              % content_type
 222                                              )
 223                                     icons[icon_url] = None
 224                         if parser and parser.refresh:
 225                             refresh = parser.refresh
 226                             try:
 227                                 url = refresh.split('=', 1)[1]
 228                             except IndexError:
 229                                 url = "self"
 230                             try:
 231                                 timeout = float(refresh.split(';')[0])
 232                             except (IndexError, ValueError):
 233                                 self.set_redirect(bookmark, "html",
 234                                                   "Bad redirect to %s (%s)"
 235                                                   % (url, refresh)
 236                                                   )
 237                             else:
 238                                 try:
 239                                     timeout = int(refresh.split(';')[0])
 240                                 except ValueError:
 241                                     pass  # float timeout
 242                                 self.set_redirect(bookmark, "html",
 243                                                   "%s (%s sec)"
 244                                                   % (url, timeout)
 245                                                   )
 246                     elif charset:
 247                         bookmark.charset = charset
 248
 249                     if not content_stripped:
 250                         self.log("   empty response, no content")
 251                     if not is_html:
 252                         self.log("   not html")
 253                 except KeyError as key:
 254                     self.log("   no header: %s" % key)
 255
 256             md5 = md5wrapper()
 257             if url_type == "ftp":  # Pass welcome message through MD5
 258                 ftp_welcome = self.get_ftp_welcome()
 259                 if not isinstance(ftp_welcome, bytes):
 260                     ftp_welcome = ftp_welcome.encode(charset or 'utf-8')
 261                 md5.update(ftp_welcome)
 262
 263             if isinstance(content, bytes):
 264                 md5.update(content)
 265             else:
 266                 md5.update(content.encode(charset or 'utf-8'))
 267             bookmark.md5 = str(md5)
 268
 269         except EOFError:
 270             bookmark.error = "Unexpected EOF (FTP server closed connection)"
 271             self.log('   EOF: %s' % bookmark.error)
 272
 273         except KeyboardInterrupt:
 274             self.log("Keyboard interrupt (^C)")
 275             return 0
 276
 277         except socket.error as e:
 278             bookmark.error = get_error(e)
 279             self.log(bookmark.error)
 280
 281         except:
 282             import traceback
 283             traceback.print_exc()
 284             bookmark.error = "Exception!"
 285             self.log('   Exception: %s' % bookmark.error)
 286
 287         finally:
 288             self.finish_check_url(bookmark)
 289
 290         # Tested
 291         return 1
 292
 293     def set_redirect(self, bookmark, errcode, newurl):
 294         bookmark.moved = moved = "(%s) to %s" % (reloc_dict[errcode], newurl)
 295         try:
 296             moved.encode('ascii')
 297         except UnicodeEncodeError:
 298             try:
 299                 moved = moved.encode(bookmark.charset)
 300             except (LookupError, TypeError, UnicodeEncodeError):
 301                 moved = moved.encode('utf-8')
 302         self.log('   Moved: %s' % moved)
 303
 304     def finish_check_url(self, bookmark):
 305         start = self.start
 306         bookmark.last_tested = str(start)
 307         now = int(time.time())
 308         bookmark.test_time = str(now - start)