1 # Copyright (C) 2023 Free Software Foundation
3 # This program is free software: you can redistribute it and/or modify
4 # it under the terms of the GNU Affero General Public License as published
5 # by the Free Software Foundation, either version 3 of the License, or
6 # (at your option) any later version.
8 # This program is distributed in the hope that it will be useful,
9 # but WITHOUT ANY WARRANTY; without even the implied warranty of
10 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11 # GNU Affero General Public License for more details.
13 # You should have received a copy of the GNU Affero General Public License
14 # along with this program. If not, see <https://www.gnu.org/licenses/>.
19 from urllib.parse import urlparse
21 from fba.helpers import blacklist
22 from fba.helpers import config
23 from fba.helpers import domain as domain_helper
25 from fba.http import csrf
26 from fba.http import network
28 from fba.models import instances
30 logging.basicConfig(level=logging.INFO)
31 logger = logging.getLogger(__name__)
44 # "rel" identifiers (no real URLs)
45 _nodeinfo_identifier = [
46 "https://nodeinfo.diaspora.software/ns/schema/2.1",
47 "http://nodeinfo.diaspora.software/ns/schema/2.1",
48 "https://nodeinfo.diaspora.software/ns/schema/2.0",
49 "http://nodeinfo.diaspora.software/ns/schema/2.0",
50 "https://nodeinfo.diaspora.software/ns/schema/1.1",
51 "http://nodeinfo.diaspora.software/ns/schema/1.1",
52 "https://nodeinfo.diaspora.software/ns/schema/1.0",
53 "http://nodeinfo.diaspora.software/ns/schema/1.0",
56 def fetch(domain: str, path: str = None, update_mode: bool = True) -> dict:
57 logger.debug("domain='%s',path='%s',update_mode='%s' - CALLED!", domain, path, update_mode)
58 domain_helper.raise_on(domain)
60 if blacklist.is_blacklisted(domain):
61 raise Exception(f"domain='{domain}' is blacklisted but function was invoked")
62 elif not isinstance(path, str) and path is not None:
63 raise ValueError(f"Parameter path[]='{type(path)}' is not of type 'str'")
64 elif path is not None and not path.startswith("/"):
65 raise ValueError(f"path='{path}' does not start with a slash")
66 elif not isinstance(update_mode, bool) and update_mode is not None:
67 raise ValueError(f"Parameter update_mode[]='{type(update_mode)}' is not of type 'bool'")
69 if path is None and update_mode:
70 logger.debug("Fetching well-known nodeinfo from domain='%s' ...", domain)
71 data = fetch_wellknown_nodeinfo(domain)
73 logger.debug("data[%s](%d)='%s'", type(data), len(data), data)
74 if "exception" in data:
75 logger.warning("Exception returned: '%s', raising again ...", type(data["exception"]))
76 raise data["exception"]
77 elif "error_message" not in data and "json" in data and len(data["json"]) > 0:
78 logger.debug("Invoking instances.set_last_nodeinfo(%s) ...", domain)
79 instances.set_last_nodeinfo(domain)
81 logger.debug("Found data[json]()=%d - EXIT!", len(data['json']))
84 # No CSRF by default, you don't have to add network.api_headers by yourself here
89 logger.debug("Checking CSRF for domain='%s'", domain)
90 headers = csrf.determine(domain, dict())
91 logger.debug("headers()=%d", len(headers))
92 except network.exceptions as exception:
93 logger.warning("Exception '%s' during checking CSRF (nodeinfo,%s) - EXIT!", type(exception), __name__)
94 instances.set_last_error(domain, exception)
97 "error_message": f"exception[{type(exception)}]='{str(exception)}'",
98 "exception" : exception,
101 logger.debug("Checking %d request paths ...", len(_request_paths))
102 for request in _request_paths:
103 logger.debug("request='%s'", request)
104 http_url = f"http://{domain}{str(path) if path is not None else '/'}"
105 https_url = f"https://{domain}{str(path) if path is not None else '/'}"
107 logger.debug("path[%s]='%s',request='%s',http_url='%s',https_url='%s'", type(path), path, request, http_url, https_url)
108 if path is None or path in [request, http_url, https_url]:
109 logger.debug("Fetching request='%s' from domain='%s' ...", request, domain)
110 data = network.get_json_api(
114 timeout=(config.get("nodeinfo_connection_timeout"), config.get("nodeinfo_read_timeout"))
117 logger.debug("data(%d)[]='%s'", len(data), type(data))
118 if "error_message" not in data and "json" in data:
119 logger.debug("Updating last_nodeinfo for domain='%s' ....", domain)
120 instances.set_last_nodeinfo(domain)
122 logger.debug("update_mode='%s'", update_mode)
124 logger.debug("Success: request='%s' - Setting detection_mode=STATIC_CHECK for domain='%s' ...", request, domain)
125 instances.set_detection_mode(domain, "STATIC_CHECK")
127 logger.debug("domain='%s',request='%s'", domain, request)
128 instances.set_nodeinfo_url(domain, f"https://{domain}{request}")
130 logger.debug("BREAK!")
133 logger.warning("Failed fetching nodeinfo from domain='%s',status_code='%s',error_message='%s'", domain, data['status_code'], data['error_message'])
135 logger.debug("data()=%d - EXIT!", len(data))
138 def fetch_wellknown_nodeinfo(domain: str) -> dict:
139 logger.debug("domain='%s' - CALLED!", domain)
140 domain_helper.raise_on(domain)
142 if blacklist.is_blacklisted(domain):
143 raise Exception(f"domain='{domain}' is blacklisted but function was invoked")
145 # No CSRF by default, you don't have to add network.api_headers by yourself here
149 logger.debug("Checking CSRF for domain='%s'", domain)
150 headers = csrf.determine(domain, dict())
151 except network.exceptions as exception:
152 logger.warning("Exception '%s' during checking CSRF (fetch_wellknown_nodeinfo,%s) - EXIT!", type(exception), __name__)
153 instances.set_last_error(domain, exception)
156 "error_message": type(exception),
157 "exception" : exception,
162 logger.debug("Fetching .well-known info for domain='%s'", domain)
163 for path in ["/.well-known/x-nodeinfo2", "/.well-known/nodeinfo"]:
164 logger.debug("Fetching path='%s' from domain='%s' ...", path, domain)
165 data = network.get_json_api(
169 (config.get("nodeinfo_connection_timeout"), config.get("nodeinfo_read_timeout"))
172 logger.debug("data(%d)[]='%s'", len(data), type(data))
173 if "error_message" not in data and "json" in data and len(data["json"]) > 0:
174 logger.debug("path='%s' returned valid json()=%d - BREAK!", path, len(data["json"]))
177 logger.debug("data[]='%s'", type(data))
178 if "exception" in data:
179 logger.warning("domain='%s' returned exception '%s' - RAISE!", domain, str(data["exception"]))
180 raise data["exception"]
181 elif "error_message" in data:
182 logger.warning("domain='%s' returned error message: '%s' - EXIT!", domain, data["error_message"])
184 elif "json" not in data:
185 logger.warning("domain='%s' returned no 'json' key - EXIT!", domain)
189 logger.debug("infos(%d)[]='%s' has been returned", len(infos), type(infos))
192 logger.debug("Marking domain='%s' as successfully handled ...", domain)
193 instances.set_success(domain)
195 logger.debug("Checking %d nodeinfo ids ...", len(_nodeinfo_identifier))
196 for niid in _nodeinfo_identifier:
199 logger.debug("Checking niid='%s' for infos[links]()=%d ...", niid, len(infos["links"]))
200 for link in infos["links"]:
201 logger.debug("link[%s]='%s'", type(link), link)
202 if not isinstance(link, dict) or not "rel" in link:
203 logger.debug("link[]='%s' is not of type 'dict' or no element 'rel' found - SKIPPED!", type(link))
205 elif link["rel"] != niid:
206 logger.debug("link[re]='%s' does not matched niid='%s' - SKIPPED!", link["rel"], niid)
208 elif "href" not in link:
209 logger.warning("link[rel]='%s' has no element 'href' - SKIPPED!", link["rel"])
211 elif link["href"] in [None, ""]:
212 logger.debug("link[href]='%s' is empty, link[rel]='%s' - SKIPPED!", link["href"], link["rel"])
214 elif not validators.url(link["href"]):
215 logger.warning("link[href]='%s' is not a valid domain - SKIPPED!", link["href"])
218 # Default is that 'href' has a complete URL, but some hosts don't send that
219 logger.debug("link[rel]='%s' matches niid='%s'", link["rel"], niid)
220 url = link["href"].lower()
222 logger.debug("Parsing url='%s' ...", url)
223 components = urlparse(url)
225 logger.debug("components[%s]='%s'", type(components), components)
226 if components.scheme == "" and components.netloc == "":
227 logger.warning("link[href]='%s' has no scheme and host name in it, prepending from domain='%s'", link['href'], domain)
228 url = f"https://{domain}{url}"
229 components = urlparse(url)
230 elif components.netloc == "":
231 logger.warning("link[href]='%s' has no netloc set, setting domain='%s'", link["href"], domain)
232 url = f"{components.scheme}://{domain}{components.path}"
233 components = urlparse(url)
235 domain2 = components.netloc.lower().split(":")[0]
236 logger.debug("domain2='%s'", domain2)
237 if not domain_helper.is_wanted(domain2):
238 logger.debug("domain2='%s' is not wanted - SKIPPED!", domain2)
241 logger.debug("Fetching nodeinfo from url='%s' ...", url)
242 data = network.fetch_api_url(
244 (config.get("connection_timeout"), config.get("read_timeout"))
247 logger.debug("link[href]='%s',data[]='%s'", link["href"], type(data))
248 if "error_message" not in data and "json" in data:
249 logger.debug("Found JSON data()=%d,link[href]='%s' - Setting detection_mode=AUTO_DISCOVERY ...", len(data), link["href"])
250 instances.set_detection_mode(domain, "AUTO_DISCOVERY")
251 instances.set_nodeinfo_url(domain, link["href"])
253 logger.debug("Marking domain='%s' as successfully handled - BREAK!", domain)
254 instances.set_success(domain)
257 logger.debug("Setting last error for domain='%s',data[]='%s'", domain, type(data))
258 instances.set_last_error(domain, data)
260 logger.debug("data()=%d", len(data))
261 if "error_message" not in data and "json" in data:
262 logger.debug("Auto-discovery successful: domain='%s' - BREAK!", domain)
264 elif "server" in infos:
265 logger.debug("Found infos[server][software]='%s'", infos["server"]["software"])
266 instances.set_detection_mode(domain, "AUTO_DISCOVERY")
267 instances.set_nodeinfo_url(domain, f"https://{domain}/.well-known/x-nodeinfo2")
269 logger.debug("Marking domain='%s' as successfully handled ...", domain)
270 instances.set_success(domain)
272 logger.warning("nodeinfo does not contain 'links' or 'server': domain='%s',infos[%s]='%s'", domain, type(infos), infos)
274 logger.debug("Returning data[]='%s' - EXIT!", type(data))