1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
|
import uuid
from urllib.parse import urljoin
import httpx
import rfc3987
from bs4 import BeautifulSoup
from flask import current_app
from .. import APP_NAME, VERSION
from .. import DATABASE as db
from .. import HUEY as huey
from ..models import ReceivedWebmention
from ..url_security import AddressResolutionError, is_public_url
class VerificationError(Exception):
"""The source permanently failed ReceivedWebmention verification."""
class SourceGoneError(VerificationError):
"""The source explicitly reports that it has been removed."""
class TemporaryFetchError(Exception):
"""Fetching the source may succeed when retried later."""
IRI_PATTERN = rfc3987.get_compiled_pattern("IRI")
HTML_URL_ATTRIBUTES = {
"href": {"a", "area", "link"},
"src": {
"audio",
"embed",
"iframe",
"img",
'input[type="image" i]',
"script",
"audio source",
"video source",
"track",
"video",
},
"cite": {"blockquote", "del", "ins", "q"},
}
def html_mentions_target(body: bytes, source_url: str, target_url: str) -> bool:
"""Check valid HTML URL attributes for the exact target URL."""
document = BeautifulSoup(body, "html.parser")
base_url = source_url
if (base_element := document.select_one("base[href]")) is not None and isinstance(
base_href := base_element.get("href"), str
):
base_url = urljoin(source_url, base_href.strip())
for attribute, selectors in HTML_URL_ATTRIBUTES.items():
selector = ", ".join(
[
"{selector}[{attribute}]".format(
selector=selector_string, attribute=attribute
)
for selector_string in selectors
]
)
for element in document.select(selector):
if not isinstance(reference := element.get(attribute), str):
continue
if urljoin(base_url, reference.strip()) == target_url:
return True
return False
def text_mentions_target(body: str, target_url: str) -> bool:
"""Check whether plain text contains the exact target IRI."""
return any(match.group() == target_url for match in IRI_PATTERN.finditer(body))
def ensure_public_request(request: httpx.Request) -> None:
"""Prevent requests to non-public network addresses."""
try:
if not is_public_url(str(request.url)):
raise VerificationError("Source resolves to a non-public address")
except ValueError as exc:
raise VerificationError("Source URL has no hostname") from exc
except AddressResolutionError as exc:
raise TemporaryFetchError("Source hostname could not be resolved") from exc
def fetch_source(source_url: str) -> tuple[httpx.Response, bytes]:
"""Fetch a source with limits on redirects, time, and response size."""
with httpx.Client(
headers={
"Accept": "text/html, application/xhtml+xml;q=0.9, text/plain;q=0.8",
"User-Agent": f"{APP_NAME}/{VERSION} ReceivedWebmention",
},
timeout=httpx.Timeout(
current_app.config.get("WEBMENTIONS_SSG_REQUEST_TIMEOUT", 5.0)
),
follow_redirects=True,
max_redirects=current_app.config.get("WEBMENTIONS_SSG_MAX_REDIRECTS", 20),
trust_env=False,
event_hooks={"request": [ensure_public_request]},
) as client:
with client.stream("GET", source_url) as response:
match response.status_code:
case 200:
pass
case 410:
raise SourceGoneError("Source returned HTTP 410")
case status:
if status in {408, 425, 429} or 500 <= status <= 599:
raise TemporaryFetchError(f"Source returned HTTP {status}")
else:
raise VerificationError(f"Source returned HTTP {status}")
max_source_bytes = current_app.config.get(
"WEBMENTIONS_SSG_MAX_SOURCE_BYTES", 1_000_000
)
if (content_length := response.headers.get("Content-Length")) is not None:
try:
if int(content_length) > max_source_bytes:
raise VerificationError("Source document is too large")
except ValueError:
pass
body = bytearray()
for chunk in response.iter_bytes(chunk_size=64 * 1024):
body.extend(chunk)
if len(body) > max_source_bytes:
raise VerificationError("Source document is too large")
return response, bytes(body)
def source_mentions_target(source_url: str, target_url: str) -> bool:
"""Fetch the source and verify it according to its media type."""
response, body = fetch_source(source_url)
media_type = (
response.headers.get("Content-Type", "").partition(";")[0].strip().lower()
)
match media_type:
case "text/html" | "application/xhtml+xml":
return html_mentions_target(body, str(response.url), target_url)
case "text/plain":
try:
decoded_body = body.decode(
response.encoding or "utf-8", errors="replace"
)
except LookupError:
decoded_body = body.decode("utf-8", errors="replace")
return text_mentions_target(decoded_body, target_url)
case _:
raise VerificationError(
f"Unsupported source content type: {media_type or 'missing'}"
)
@huey.task(retries=2, retry_delay=50)
def verify_webmention(webmention_uuid: uuid.UUID) -> None:
"""Verify a ReceivedWebmention and store the result."""
webmention = db.session.get(ReceivedWebmention, webmention_uuid)
if webmention is None:
current_app.logger.warning(
"Cannot verify unknown ReceivedWebmention %s", webmention_uuid
)
return
webmention.status = "verifying"
webmention.failure_reason = None
db.session.commit()
try:
mentions_target = source_mentions_target(webmention.source, webmention.target)
except SourceGoneError as exc:
webmention.status = "deleted"
webmention.failure_reason = str(exc)
except VerificationError as exc:
webmention.status = "failed"
webmention.failure_reason = str(exc)
except (TemporaryFetchError, httpx.RequestError) as exc:
webmention.status = "failed"
webmention.failure_reason = str(exc) or "Source could not be fetched"
db.session.commit()
# Huey retries the task because the exception escapes.
raise
else:
if mentions_target:
webmention.status = "verified"
webmention.failure_reason = None
else:
webmention.status = "deleted"
webmention.failure_reason = "Source does not mention target"
db.session.commit()
|
