apunkt/SurfSense
1
0
Fork 0
mirror of https://github.com/MODSetter/SurfSense.git synced 2026-07-10 22:32:16 +02:00
Code Issues Projects Releases Packages Wiki Activity

chore: bumped version to 0.0.31

Browse source
This commit is contained in:
DESKTOP-RTLN3BA\$punk 2026-07-06 21:43:15 -07:00
parent 8df8565e0a
commit 1c9ab207ef
56 changed files with 520 additions and 190 deletions
Download patch file Download diff file Expand all files Collapse all files

69
surfsense_backend/app/utils/crawl/contacts.py
View file

@ -87,9 +87,30 @@ _NOISE_EMAIL_DOMAINS = frozenset(
# or version-pinned dep (``react@18.2.0.js``) matches the email shape.
_ASSET_TLDS = frozenset(
{
"png", "jpg", "jpeg", "gif", "svg", "webp", "ico", "bmp",
"css", "js", "mjs", "cjs", "ts", "map", "json", "xml",
"woff", "woff2", "ttf", "eot", "otf", "php", "html", "htm",
"png",
"jpg",
"jpeg",
"gif",
"svg",
"webp",
"ico",
"bmp",
"css",
"js",
"mjs",
"cjs",
"ts",
"map",
"json",
"xml",
"woff",
"woff2",
"ttf",
"eot",
"otf",
"php",
"html",
"htm",
}
)
@ -100,19 +121,47 @@ _EMAIL_RE = re.compile(r"[a-zA-Z0-9._%+\-]+@[a-zA-Z0-9.\-]+\.[a-zA-Z]{2,}")
# common locals like hello/info/contact/support.
_PLACEHOLDER_EMAIL_LOCALS = frozenset(
{
"youremail", "yourname", "youraddress", "myemail", "email", "name",
"user", "username", "someone", "somebody", "johndoe", "janedoe",
"firstname", "lastname", "firstnamelastname", "firstlast",
"test", "example", "sample", "placeholder",
"youremail",
"yourname",
"youraddress",
"myemail",
"email",
"name",
"user",
"username",
"someone",
"somebody",
"johndoe",
"janedoe",
"firstname",
"lastname",
"firstnamelastname",
"firstlast",
"test",
"example",
"sample",
"placeholder",
}
)
# Placeholder profile handles left in site templates ("github.com/username").
_PLACEHOLDER_SOCIAL_SEGMENTS = frozenset(
{
"username", "yourusername", "yourhandle", "handle", "user",
"profile", "yourprofile", "yourname", "yourpage", "pagename",
"youraccount", "account", "example", "placeholder", "yourcompany",
"username",
"yourusername",
"yourhandle",
"handle",
"user",
"profile",
"yourprofile",
"yourname",
"yourpage",
"pagename",
"youraccount",
"account",
"example",
"placeholder",
"yourcompany",
}
)