1 # Fedi API Block - An aggregator for fetching blocking data from fediverse nodes
2 # Copyright (C) 2023 Free Software Foundation
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.
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.
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/>.
24 from fba.helpers import config
25 from fba.helpers import domain as domain_helper
26 from fba.helpers import tidyup
28 from fba.http import federation
29 from fba.http import network
31 from fba.models import instances
33 logging.basicConfig(level=logging.INFO)
34 logger = logging.getLogger(__name__)
35 #logger.setLevel(logging.DEBUG)
37 def fetch_peers(domain: str, origin: str) -> list:
38 logger.debug("domain='%s',origin='%s' - CALLED!", domain, origin)
39 domain_helper.raise_on(domain)
43 # No CSRF by default, you don't have to add network.api_headers by yourself here
47 logger.debug("Checking CSRF for domain='%s'", domain)
48 headers = csrf.determine(domain, dict())
49 except network.exceptions as exception:
50 logger.warning("Exception '%s' during checking CSRF (fetch_peers,%s) - EXIT!", type(exception), __name__)
51 instances.set_last_error(domain, exception)
55 logger.debug("Fetching '/api/v3/site' from domain='%s' ...", domain)
56 data = network.get_json_api(
60 (config.get("connection_timeout"), config.get("read_timeout"))
63 logger.debug("data[]='%s'", type(data))
64 if "error_message" in data:
65 logger.warning("Could not reach any JSON API: domain='%s'", domain)
66 instances.set_last_error(domain, data)
67 elif "federated_instances" in data["json"] and isinstance(data["json"]["federated_instances"], dict):
68 logger.debug("Found federated_instances for domain='%s'", domain)
69 peers = peers + federation.add_peers(data["json"]["federated_instances"])
71 logger.debug("Marking domain='%s' as successfully handled ...")
72 instances.set_success(domain)
75 logger.warning("Fetching instances for domain='%s' from /instances ...", domain)
76 peers = fetch_instances(domain, origin)
78 except network.exceptions as exception:
79 logger.warning("Exception during fetching JSON: domain='%s',exception[%s]:'%s'", domain, type(exception), str(exception))
80 instances.set_last_error(domain, exception)
82 logger.debug("peers()=%d - EXIT!", len(peers))
85 def fetch_blocks(domain: str, nodeinfo_url: str) -> list:
86 logger.debug("domain='%s,nodeinfo_url='%s' - CALLED!", domain, nodeinfo_url)
87 domain_helper.raise_on(domain)
89 if not isinstance(nodeinfo_url, str):
90 raise ValueError(f"Parameter nodeinfo_url[]='{type(nodeinfo_url)}' is not 'str'")
91 elif nodeinfo_url == "":
92 raise ValueError("Parameter 'nodeinfo_url' is empty")
95 "Blocked Instances".lower(),
96 "Instàncies bloquejades".lower(),
97 "Blocáilte Ásc".lower(),
99 "Blokované instance".lower(),
100 "Geblokkeerde instanties".lower(),
101 "Blockerade instanser".lower(),
102 "Instàncias blocadas".lower(),
103 "Istanze bloccate".lower(),
104 "Instances bloquées".lower(),
105 "Letiltott példányok".lower(),
106 "Instancias bloqueadas".lower(),
107 "Blokeatuta dauden instantziak".lower(),
109 "Peladen Yang Diblokir".lower(),
110 "Blokerede servere".lower(),
111 "Blokitaj nodoj".lower(),
112 "Блокирани Инстанции".lower(),
113 "Blockierte Instanzen".lower(),
114 "Estetyt instanssit".lower(),
115 "Instâncias bloqueadas".lower(),
116 "Zablokowane instancje".lower(),
117 "Blokované inštancie".lower(),
118 "المثلاء المحجوبون".lower(),
119 "Užblokuoti serveriai".lower(),
120 "ブロックしたインスタンス".lower(),
121 "Блокированные Инстансы".lower(),
122 "Αποκλεισμένοι διακομιστές".lower(),
124 "Instâncias bloqueadas".lower(),
130 # json endpoint for newer mastodongs
131 logger.debug("Fetching /instances from domain='%s'", domain)
132 response = network.fetch_response(
136 (config.get("connection_timeout"), config.get("read_timeout"))
139 logger.debug("response.ok='%s',response.status_code=%d,response.text()=%d", response.ok, response.status_code, len(response.text))
140 if response.ok and response.status_code < 300 and response.text != "":
141 logger.debug("Parsing %s Bytes ...", len(response.text))
143 doc = bs4.BeautifulSoup(response.text, "html.parser")
144 logger.debug("doc[]='%s'", type(doc))
146 headers = doc.findAll("h5")
148 logger.debug("Search in %d header(s) ...", len(headers))
149 for header in headers:
150 logger.debug("header[]='%s'", type(header))
151 content = header.contents[0]
153 logger.debug("content[%s]='%s'", type(content), content)
155 logger.debug("domain='%s' has returned empty header='%s' - SKIPPED!", domain, header)
157 elif content.lower() in translations:
158 logger.debug("Found header with blocked instances - BREAK!")
162 logger.debug("found[]='%s'", type(found))
164 logger.debug("domain='%s' is not blocking any instances - EXIT!", domain)
167 blocking = found.find_next(["ul","table"]).findAll("a")
168 logger.debug("Found %d blocked instance(s) ...", len(blocking))
170 logger.debug("tag[]='%s'", type(tag))
171 blocked = tidyup.domain(tag.contents[0])
172 logger.debug("blocked='%s'", blocked)
174 if not utils.is_domain_wanted(blocked):
175 logger.debug("blocked='%s' is not wanted - SKIPPED!", blocked)
178 logger.debug("Appending blocker='%s',blocked='%s',block_level='reject'", domain, blocked)
183 "block_level": "reject",
186 except network.exceptions as exception:
187 logger.warning("domain='%s',exception[%s]:'%s'", domain, type(exception), str(exception))
188 instances.set_last_error(domain, exception)
190 logger.debug("blocklist()=%d - EXIT!", len(blocklist))
193 def fetch_instances(domain: str, origin: str) -> list:
194 logger.debug("domain='%s',origin='%s' - CALLED!", domain, origin)
195 domain_helper.raise_on(domain)
200 # json endpoint for newer mastodongs
201 logger.debug("Fetching /instances from domain='%s'", domain)
202 response = network.fetch_response(
206 (config.get("connection_timeout"), config.get("read_timeout"))
209 logger.debug("response.ok='%s',response.status_code=%d,response.text()=%d", response.ok, response.status_code, len(response.text))
210 if response.ok and response.status_code < 300 and response.text != "":
211 logger.debug("Parsing %s Bytes ...", len(response.text))
213 doc = bs4.BeautifulSoup(response.text, "html.parser")
214 logger.debug("doc[]='%s'", type(doc))
216 headers = doc.findAll("h5")
217 logger.debug("Checking %d headers ...", len(headers))
218 for header in headers:
219 logger.debug("header[%s]='%s'", type(header), header)
221 rows = header.find_next(["ul","table"]).findAll("a")
222 logger.debug("Found %d blocked instance(s) ...", len(rows))
224 logger.debug("tag[]='%s'", type(tag))
225 text = tag.contents[0] if isinstance(tag.contents[0], str) else tag.contents[0].text
226 peer = tidyup.domain(text)
227 logger.debug("peer='%s'", peer)
230 logger.debug("peer is empty - SKIPPED!")
232 elif not utils.is_domain_wanted(peer):
233 logger.debug("peer='%s' is not wanted - SKIPPED!", peer)
236 logger.debug("peer='%s' already added - SKIPPED!", peer)
239 logger.debug("Appending peer='%s' ...", peer)
242 logger.debug("Marking domain='%s' as successfully handled ...")
243 instances.set_success(domain)
245 except network.exceptions as exception:
246 logger.warning("domain='%s',exception[%s]:'%s'", domain, type(exception), str(exception))
247 instances.set_last_error(domain, exception)
249 logger.debug("peers()=%d - EXIT!", len(peers))