]> git.phdru.name Git - bookmarks_db.git/blob - Robots/bkmk_robot_base.py
df33a26bd43162d92252e04df933bcb378abaffe
[bookmarks_db.git] / 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) 2000-2024 PhiloSoft Design"
9 __license__ = "GNU GPL"
10
11 __all__ = ['robot_base', 'get_error']
12
13
14 from base64 import b64encode
15 from urllib.parse import urlsplit, urljoin
16 import sys
17 import socket
18 import time
19
20 from m_lib.md5wrapper import md5wrapper
21 from m_lib.net.www.util import parse_time
22
23 from bkmk_objects import Robot
24 from parse_html import parse_html
25
26
27 # Fake headers to pretend this is a real browser
28 _user_agent = "Mozilla/5.0 (X11; U; Linux 2.6 i686; en)"
29 " Gecko/20001221 Firefox/2.0.0"
30 _x_user_agent = "bookmarks_db (Python %d.%d.%d)" % sys.version_info[:3]
31
32 request_headers = {
33     'Accept': '*/*',
34     'Accept-Language': 'ru,en',
35     'Cache-Control': 'max-age=300',
36     'Connection': 'close',
37     'Referer': '/',
38     'User-Agent': _user_agent,
39     'X-User-Agent': _x_user_agent,
40 }
41
42
43 reloc_dict = {
44   301: "perm1.",
45   302: "temp2.",
46   303: "temp3.",
47   307: "temp7.",
48   308: "temp8.",
49   "html": "html"
50 }
51
52
53 def get_error(e):
54     if isinstance(e, str):
55         return e
56
57     else:
58         s = []
59         for i in e:
60             s.append("'%s'" % str(i).replace('\n', "\\n"))
61         return "(%s)" % ' '.join(s)
62
63
64 # Icon cache; maps URL to a tuple (content type, data)
65 # or None if there is no icon.
66 icons = {}
67
68
69 class robot_base(Robot):
70     timeout = 60
71
72     def __init__(self, *args, **kw):
73         Robot.__init__(self, *args, **kw)
74         socket.setdefaulttimeout(int(self.timeout))
75
76     def check_url(self, bookmark):
77         try:
78             self.start = int(time.time())
79             bookmark.icon = None
80
81             split_results = urlsplit(bookmark.href)
82             url_type, netloc, url_path, query, url_tag = split_results
83             url_host = split_results.hostname
84
85             url = "%s://%s%s" % (url_type, url_host, url_path)
86             error, redirect_code, redirect_to, headers, content = \
87                 self.get(bookmark, url, True)
88
89             if error:
90                 bookmark.error = error
91                 return 1
92
93             if redirect_code:
94                 self.set_redirect(bookmark, redirect_code, redirect_to)
95                 return 1
96
97             size = 0
98             last_modified = None
99
100             if headers:
101                 try:
102                     size = headers["Content-Length"]
103                 except KeyError:
104                     pass
105
106                 try:
107                     last_modified = headers["Last-Modified"]
108                 except KeyError:
109                     pass
110
111                 if last_modified:
112                     last_modified = parse_time(last_modified)
113
114             if not size:  # Could be None from headers
115                 size = len(content)
116
117             if last_modified:
118                 last_modified = str(int(last_modified))
119             else:
120                 last_modified = bookmark.last_visit
121
122             bookmark.size = size
123             bookmark.last_modified = last_modified
124
125             charset = None
126             if headers:
127                 try:
128                     content_type = headers["Content-Type"]
129                     self.log("   Content-Type   : %s" % content_type)
130                     if content_type is None:
131                         if b'html' in content.lower():
132                             content_type = 'text/html'
133                         else:
134                             content_type = 'text/plain'
135                         self.log("   Set Content-Type to: %s"
136                                  % content_type)
137                     try:
138                         # extract charset from
139                         # "text/html; charset=UTF-8, foo; bar"
140                         content_type, charset = content_type.split(';', 1)
141                         content_type = content_type.strip()
142                         charset = charset.split('=')[1].strip().split(',')[0]
143                         self.log("   HTTP charset   : %s" % charset)
144                     except (ValueError, IndexError):
145                         charset = None
146                         self.log("   no charset in Content-Type header")
147                     is_html = False
148                     for ctype in ("text/html", "application/xhtml+xml"):
149                         if content_type.startswith(ctype):
150                             is_html = True
151                             break
152                     content_stripped = content.strip()
153                     if content_stripped and charset:
154                         try:
155                             content_stripped = content_stripped.decode(
156                                 charset, 'replace')
157                         except LookupError:
158                             charset = None
159                             self.log("   unknown charset "
160                                      "in Content-Type header")
161                     if content_stripped and is_html:
162                         parser = parse_html(
163                             content_stripped, charset, self.log)
164                         if charset:
165                             bookmark.charset = charset
166                         elif parser and parser.meta_charset:
167                             bookmark.charset = parser.meta_charset
168                         if parser:
169                             bookmark.real_title = parser.title
170                             icon = parser.icon
171                         else:
172                             icon = None
173                         if not icon:
174                             icon = "/favicon.ico"
175                         icon_url = urljoin(
176                             "%s://%s%s" % (url_type, url_host, url_path), icon)
177                         self.log("   looking for icon at: %s" % icon_url)
178                         if icon_url in icons:
179                             if icons[icon_url]:
180                                 bookmark.icon_href = icon_url
181                                 content_type, bookmark.icon = icons[icon_url]
182                                 self.log("   cached icon: %s" % content_type)
183                             else:
184                                 self.log("   cached icon: no icon")
185                         else:
186                             try:
187                                 _icon_url = icon_url
188                                 for i in range(8):
189                                     error, icon_redirect_code, \
190                                         icon_redirect_to, icon_headers, \
191                                         icon_data = \
192                                         self.get(bookmark, _icon_url)
193                                     if icon_redirect_code:
194                                         _icon_url = icon_redirect_to
195                                         self.log("   redirect to : %s"
196                                                  % _icon_url)
197                                     else:
198                                         if icon_data is None:
199                                             raise IOError("No icon")
200                                         break
201                                 else:
202                                     raise IOError("Too many redirects")
203                             except:
204                                 etype, emsg, _ = sys.exc_info()
205                                 self.log("   no icon        : %s %s"
206                                          % (etype, emsg))
207                                 etype = emsg = _ = None
208                                 icons[icon_url] = None
209                             else:
210                                 content_type = icon_headers["Content-Type"]
211                                 if content_type and (
212                                     content_type.startswith("application/")
213                                     or content_type.startswith("image/")
214                                     or content_type.startswith("text/plain")
215                                 ):
216                                     bookmark.icon_href = icon_url
217                                     self.log("   got icon       : %s"
218                                              % content_type)
219                                     if (
220                                         content_type.startswith("application/")
221                                         or content_type.startswith(
222                                             "text/plain")
223                                     ):
224                                         self.log("   non-image content type,"
225                                                  " assume x-icon")
226                                         content_type = 'image/x-icon'
227                                     if not isinstance(icon_data, bytes):
228                                         icon_data = icon_data.encode('latin1')
229                                     bookmark.icon = "data:%s;base64,%s" \
230                                         % (content_type, b64encode(icon_data))
231                                     icons[icon_url] = (content_type,
232                                                        bookmark.icon
233                                                        )
234                                 else:
235                                     self.log("   no icon        : "
236                                              "bad content type '%s'"
237                                              % content_type
238                                              )
239                                     icons[icon_url] = None
240                         if parser and parser.refresh:
241                             refresh = parser.refresh
242                             try:
243                                 url = refresh.split('=', 1)[1]
244                             except IndexError:
245                                 url = "self"
246                             try:
247                                 timeout = float(refresh.split(';')[0])
248                             except (IndexError, ValueError):
249                                 self.set_redirect(bookmark, "html",
250                                                   "Bad redirect to %s (%s)"
251                                                   % (url, refresh)
252                                                   )
253                             else:
254                                 try:
255                                     timeout = int(refresh.split(';')[0])
256                                 except ValueError:
257                                     pass  # float timeout
258                                 self.set_redirect(bookmark, "html",
259                                                   "%s (%s sec)"
260                                                   % (url, timeout)
261                                                   )
262                     elif charset:
263                         bookmark.charset = charset
264
265                     if not content_stripped:
266                         self.log("   empty response, no content")
267                     if not is_html:
268                         self.log("   not html")
269                 except KeyError as key:
270                     self.log("   no header: %s" % key)
271
272             md5 = md5wrapper()
273             if url_type == "ftp":  # Pass welcome message through MD5
274                 ftp_welcome = self.get_ftp_welcome()
275                 if not isinstance(ftp_welcome, bytes):
276                     ftp_welcome = ftp_welcome.encode(charset or 'utf-8')
277                 md5.update(ftp_welcome)
278
279             if isinstance(content, bytes):
280                 md5.update(content)
281             else:
282                 md5.update(content.encode(charset or 'utf-8'))
283             bookmark.md5 = str(md5)
284
285         except EOFError:
286             bookmark.error = "Unexpected EOF (FTP server closed connection)"
287             self.log('   EOF: %s' % bookmark.error)
288
289         except KeyboardInterrupt:
290             self.log("Keyboard interrupt (^C)")
291             return 0
292
293         except socket.error as e:
294             bookmark.error = get_error(e)
295             self.log(bookmark.error)
296
297         except:
298             import traceback
299             traceback.print_exc()
300             bookmark.error = "Exception!"
301             self.log('   Exception: %s' % bookmark.error)
302
303         finally:
304             self.finish_check_url(bookmark)
305
306         # Tested
307         return 1
308
309     def set_redirect(self, bookmark, errcode, newurl):
310         bookmark.moved = moved = "(%s) to %s" % (reloc_dict[errcode], newurl)
311         try:
312             moved.encode('ascii')
313         except UnicodeEncodeError:
314             try:
315                 moved = moved.encode(bookmark.charset)
316             except (LookupError, TypeError, UnicodeEncodeError):
317                 moved = moved.encode('utf-8')
318         self.log('   Moved: %s' % moved)
319
320     def finish_check_url(self, bookmark):
321         start = self.start
322         bookmark.last_tested = str(start)
323         now = int(time.time())
324         bookmark.test_time = str(now - start)