fba/federation/mastodon.py

   1 # Fedi API Block - An aggregator for fetching blocking data from fediverse nodes
   2 # Copyright (C) 2023 Free Software Foundation
   3 #
   4 # This program is free software: you can redistribute it and/or modify
   5 # it under the terms of the GNU Affero General Public License as published
   6 # by the Free Software Foundation, either version 3 of the License, or
   7 # (at your option) any later version.
   8 #
   9 # This program is distributed in the hope that it will be useful,
  10 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 # GNU Affero General Public License for more details.
  13 #
  14 # You should have received a copy of the GNU Affero General Public License
  15 # along with this program.  If not, see <https://www.gnu.org/licenses/>.
  16
  17 import bs4
  18 import validators
  19
  20 from fba import blacklist
  21 from fba import blocks
  22 from fba import config
  23 from fba import fba
  24
  25 language_mapping = {
  26     # English -> English
  27     "Silenced instances"            : "Silenced servers",
  28     "Suspended instances"           : "Suspended servers",
  29     "Limited instances"             : "Limited servers",
  30     "Filtered media"                : "Filtered media",
  31     # Mappuing German -> English
  32     "Gesperrte Server"              : "Suspended servers",
  33     "Gefilterte Medien"             : "Filtered media",
  34     "Stummgeschaltete Server"       : "Silenced servers",
  35     # Japanese -> English
  36     "停止済みのサーバー"            : "Suspended servers",
  37     "制限中のサーバー"              : "Limited servers",
  38     "メディアを拒否しているサーバー": "Filtered media",
  39     "サイレンス済みのサーバー"      : "Silenced servers",
  40     # ??? -> English
  41     "שרתים מושעים"                  : "Suspended servers",
  42     "מדיה מסוננת"                   : "Filtered media",
  43     "שרתים מוגבלים"                 : "Silenced servers",
  44     # French -> English
  45     "Serveurs suspendus"            : "Suspended servers",
  46     "Médias filtrés"                : "Filtered media",
  47     "Serveurs limités"              : "Limited servers",
  48     "Serveurs modérés"              : "Limited servers",
  49 }
  50
  51 def fetch_blocks_from_about(domain: str) -> dict:
  52     # DEBUG: print(f"DEBUG: domain='{domain}' - CALLED!")
  53     if type(domain) != str:
  54         raise ValueError(f"Parameter domain[]={type(domain)} is not 'str'")
  55     elif domain == "":
  56         raise ValueError(f"Parameter 'domain' is empty")
  57
  58     # DEBUG: print("DEBUG: Fetching mastodon blocks from domain:", domain)
  59     blocklist = {
  60         "Suspended servers": [],
  61         "Filtered media"   : [],
  62         "Limited servers"  : [],
  63         "Silenced servers" : [],
  64     }
  65
  66     try:
  67         doc = bs4.BeautifulSoup(
  68             fba.get_response(domain, "/about/more", fba.headers, (config.get("connection_timeout"), config.get("read_timeout"))).text,
  69             "html.parser",
  70         )
  71     except BaseException as e:
  72         print("ERROR: Cannot fetch from domain:", domain, e)
  73         fba.update_last_error(domain, e)
  74         return {}
  75
  76     for header in doc.find_all("h3"):
  77         header_text = fba.tidyup_reason(header.text)
  78
  79         # DEBUG: print(f"DEBUG: header_text='{header_text}'")
  80         if header_text in language_mapping:
  81             # DEBUG: print(f"DEBUG: header_text='{header_text}'")
  82             header_text = language_mapping[header_text]
  83         else:
  84             print(f"WARNING: header_text='{header_text}' not found in language mapping table")
  85
  86         if header_text in blocklist or header_text.lower() in blocklist:
  87             # replaced find_next_siblings with find_all_next to account for instances that e.g. hide lists in dropdown menu
  88             for line in header.find_all_next("table")[0].find_all("tr")[1:]:
  89                 blocklist[header_text].append(
  90                     {
  91                         "domain": fba.tidyup_domain(line.find("span").text),
  92                         "hash"  : fba.tidyup_domain(line.find("span")["title"][9:]),
  93                         "reason": fba.tidyup_domain(line.find_all("td")[1].text),
  94                     }
  95                 )
  96         else:
  97             print(f"WARNING: header_text='{header_text}' not found in blocklist()={len(blocklist)}")
  98
  99     # DEBUG: print("DEBUG: Returning blocklist for domain:", domain)
 100     return {
 101         "reject"        : blocklist["Suspended servers"],
 102         "media_removal" : blocklist["Filtered media"],
 103         "followers_only": blocklist["Limited servers"] + blocklist["Silenced servers"],
 104     }
 105
 106 def fetch_blocks(domain: str, origin: str, nodeinfo_url: str):
 107     print(f"DEBUG: domain='{domain}',origin='{origin}',nodeinfo_url='{nodeinfo_url}' - CALLED!")
 108     if type(domain) != str:
 109         raise ValueError(f"Parameter domain[]={type(domain)} is not 'str'")
 110     elif domain == "":
 111         raise ValueError(f"Parameter 'domain' is empty")
 112     elif type(origin) != str and origin != None:
 113         raise ValueError(f"Parameter origin[]={type(origin)} is not 'str'")
 114     elif origin == "":
 115         raise ValueError(f"Parameter 'origin' is empty")
 116     elif type(nodeinfo_url) != str:
 117         raise ValueError(f"Parameter nodeinfo_url[]={type(nodeinfo_url)} is not 'str'")
 118     elif nodeinfo_url == "":
 119         raise ValueError(f"Parameter 'nodeinfo_url' is empty")
 120
 121     try:
 122         # json endpoint for newer mastodongs
 123         blockdict = list()
 124         try:
 125             json = {
 126                 "reject"        : [],
 127                 "media_removal" : [],
 128                 "followers_only": [],
 129                 "report_removal": []
 130             }
 131
 132             # handling CSRF, I've saw at least one server requiring it to access the endpoint
 133             # DEBUG: print("DEBUG: Fetching meta:", domain)
 134             meta = bs4.BeautifulSoup(
 135                 fba.get_response(domain, "/", fba.headers, (config.get("connection_timeout"), config.get("read_timeout"))).text,
 136                 "html.parser",
 137             )
 138             try:
 139                 csrf = meta.find("meta", attrs={"name": "csrf-token"})["content"]
 140                 # DEBUG: print("DEBUG: Adding CSRF token:", domain, csrf)
 141                 reqheaders = {**fba.api_headers, **{"X-CSRF-Token": csrf}}
 142             except BaseException as e:
 143                 # DEBUG: print("DEBUG: No CSRF token found, using normal headers:", domain, e)
 144                 reqheaders = fba.api_headers
 145
 146             # DEBUG: print("DEBUG: Querying API domain_blocks:", domain)
 147             blocklist = fba.get_response(domain, "/api/v1/instance/domain_blocks", reqheaders, (config.get("connection_timeout"), config.get("read_timeout"))).json()
 148
 149             print(f"INFO: Checking {len(blocklist)} entries from domain='{domain}',software='mastodon' ...")
 150             for block in blocklist:
 151                 entry = {
 152                     'domain': block['domain'],
 153                     'hash'  : block['digest'],
 154                     'reason': block['comment']
 155                 }
 156
 157                 # DEBUG: print("DEBUG: severity,domain,hash,comment:", block['severity'], block['domain'], block['digest'], block['comment'])
 158                 if block['severity'] == 'suspend':
 159                     # DEBUG: print(f"DEBUG: Adding entry='{entry}' with severity='{block['severity']}' ...")
 160                     json['reject'].append(entry)
 161                 elif block['severity'] == 'silence':
 162                     # DEBUG: print(f"DEBUG: Adding entry='{entry}' with severity='{block['severity']}' ...")
 163                     json['followers_only'].append(entry)
 164                 elif block['severity'] == 'reject_media':
 165                     # DEBUG: print(f"DEBUG: Adding entry='{entry}' with severity='{block['severity']}' ...")
 166                     json['media_removal'].append(entry)
 167                 elif block['severity'] == 'reject_reports':
 168                     # DEBUG: print(f"DEBUG: Adding entry='{entry}' with severity='{block['severity']}' ...")
 169                     json['report_removal'].append(entry)
 170                 else:
 171                     print("WARNING: Unknown severity:", block['severity'], block['domain'])
 172
 173         except BaseException as e:
 174             # DEBUG: print(f"DEBUG: Failed, trying mastodon-specific fetches: domain='{domain}',exception[{type(e)}]={str(e)}")
 175             json = fetch_blocks_from_about(domain)
 176
 177         print(f"INFO: Checking {len(json.items())} entries from domain='{domain}',software='mastodon' ...")
 178         for block_level, blocklist in json.items():
 179             # DEBUG: print("DEBUG: domain,block_level,blocklist():", domain, block_level, len(blocklist))
 180             block_level = fba.tidyup_domain(block_level)
 181
 182             # DEBUG: print("DEBUG: AFTER-block_level:", block_level)
 183             if block_level == "":
 184                 print("WARNING: block_level is empty, domain:", domain)
 185                 continue
 186
 187             # DEBUG: print(f"DEBUG: Checking {len(blocklist)} entries from domain='{domain}',software='mastodon',block_level='{block_level}' ...")
 188             for block in blocklist:
 189                 blocked, blocked_hash, reason = block.values()
 190                 # DEBUG: print("DEBUG: blocked,hash,reason:", blocked, blocked_hash, reason)
 191                 blocked = fba.tidyup_domain(blocked)
 192                 # DEBUG: print("DEBUG: AFTER-blocked:", blocked)
 193
 194                 if blocked == "":
 195                     print("WARNING: blocked is empty:", domain)
 196                     continue
 197                 elif blacklist.is_blacklisted(blocked):
 198                     # DEBUG: print(f"DEBUG: blocked='{blocked}' is blacklisted - skipping!")
 199                     continue
 200                 elif blocked.count("*") > 0:
 201                     # Doing the hash search for instance names as well to tidy up DB
 202                     fba.cursor.execute(
 203                         "SELECT domain, origin, nodeinfo_url FROM instances WHERE hash = ? LIMIT 1", [blocked_hash]
 204                     )
 205                     searchres = fba.cursor.fetchone()
 206
 207                     if searchres == None:
 208                         print(f"WARNING: Cannot deobsfucate blocked='{blocked}',blocked_hash='{blocked_hash}' - SKIPPED!")
 209                         continue
 210
 211                     # DEBUG: print("DEBUG: Updating domain: ", searchres[0])
 212                     blocked = searchres[0]
 213                     origin = searchres[1]
 214                     nodeinfo_url = searchres[2]
 215
 216                     # DEBUG: print("DEBUG: Looking up instance by domain:", blocked)
 217                     if not validators.domain(blocked):
 218                         print(f"WARNING: blocked='{blocked}',software='mastodon' is not a valid domain name - skipped!")
 219                         continue
 220                     elif not fba.is_instance_registered(blocked):
 221                         # DEBUG: print(f"DEBUG: Domain blocked='{blocked}' wasn't found, adding ..., domain='{domain}',origin='{origin}',nodeinfo_url='{nodeinfo_url}'")
 222                         fba.add_instance(blocked, domain, inspect.currentframe().f_code.co_name, nodeinfo_url)
 223                 elif not validators.domain(blocked):
 224                     print(f"WARNING: blocked='{blocked}',software='mastodon' is not a valid domain name - skipped!")
 225                     continue
 226
 227                 # DEBUG: print("DEBUG: Looking up instance by domain:", blocked)
 228                 if not validators.domain(blocked):
 229                     print(f"WARNING: blocked='{blocked}',software='mastodon' is not a valid domain name - skipped!")
 230                     continue
 231                 elif not fba.is_instance_registered(blocked):
 232                     # DEBUG: print("DEBUG: Hash wasn't found, adding:", blocked, domain)
 233                     fba.add_instance(blocked, domain, inspect.currentframe().f_code.co_name, nodeinfo_url)
 234
 235                 blocking = blocked if blocked.count("*") <= 1 else blocked_hash
 236                 # DEBUG: print(f"DEBUG: blocking='{blocking}',blocked='{blocked}',blocked_hash='{blocked_hash}'")
 237
 238                 if not blocks.is_instance_blocked(domain, blocked, block_level):
 239                     # DEBUG: print("DEBUG: Blocking:", domain, blocked, block_level)
 240                     blocks.add_instance(domain, blocking, reason, block_level)
 241
 242                     if block_level == "reject":
 243                         blockdict.append({
 244                             "blocked": blocked,
 245                             "reason" : reason
 246                         })
 247                 else:
 248                     # DEBUG: print(f"DEBUG: Updating block last seen and reason for domain='{domain}',blocking='{blocking}' ...")
 249                     blocks.update_last_seen(domain, blocking, block_level)
 250                     blocks.update_reason(reason, domain, blocking, block_level)
 251
 252         # DEBUG: print("DEBUG: Committing changes ...")
 253         fba.connection.commit()
 254     except Exception as e:
 255         print(f"ERROR: domain='{domain}',software='mastodon',exception[{type(e)}]:'{str(e)}'")
 256
 257     # DEBUG: print("DEBUG: EXIT!")