bbot 2.6.0.6856rc0__tar.gz → 2.6.0.6871rc0__tar.gz

{bbot-2.6.0.6856rc0 → bbot-2.6.0.6871rc0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: bbot
-Version: 2.6.0.6856rc0
+Version: 2.6.0.6871rc0
 Summary: OSINT automation for hackers.
 License: GPL-3.0
 Keywords: python,cli,automation,osint,threat-intel,intelligence,neo4j,scanner,python-library,hacking,recursion,pentesting,recon,command-line-tool,bugbounty,subdomains,security-tools,subdomain-scanner,osint-framework,attack-surface,subdomain-enumeration,osint-tool

{bbot-2.6.0.6856rc0 → bbot-2.6.0.6871rc0}/bbot/__init__.py

@@ -1,5 +1,5 @@
 # version placeholder (replaced by poetry-dynamic-versioning)
-__version__ = "v2.6.0.6856rc"
+__version__ = "v2.6.0.6871rc"
 
 from .scanner import Scanner, Preset
 

{bbot-2.6.0.6856rc0 → bbot-2.6.0.6871rc0}/bbot/core/helpers/misc.py

@@ -216,26 +216,29 @@ def split_host_port(d):
     host = None
     port = None
     scheme = None
+
+    # first, try to parse as an IP address
     if is_ip(d):
         return make_ip_type(d), port
 
+    # if not an IP address, try to parse as a host:port
     match = bbot_regexes.split_host_port_regex.match(d)
     if match is None:
-        raise ValueError(f'split_port() failed to parse "{d}"')
+        raise ValueError(f'split_host_port() failed to parse "{d}"')
     scheme = match.group("scheme")
     netloc = match.group("netloc")
     if netloc is None:
-        raise ValueError(f'split_port() failed to parse "{d}"')
+        raise ValueError(f'split_host_port() failed to parse "{d}"')
 
     match = bbot_regexes.extract_open_port_regex.match(netloc)
     if match is None:
-        raise ValueError(f'split_port() failed to parse netloc "{netloc}" (original value: {d})')
+        raise ValueError(f'split_host_port() failed to parse netloc "{netloc}" (original value: {d})')
 
     host = match.group(2)
     if host is None:
         host = match.group(1)
     if host is None:
-        raise ValueError(f'split_port() failed to locate host in netloc "{netloc}" (original value: {d})')
+        raise ValueError(f'split_host_port() failed to locate host in netloc "{netloc}" (original value: {d})')
 
     port = match.group(3)
     if port is None and scheme is not None:

{bbot-2.6.0.6856rc0 → bbot-2.6.0.6871rc0}/bbot/core/helpers/regexes.py

@@ -23,13 +23,28 @@ num_regex = re.compile(r"\d+")
 _ipv4_regex = r"(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)(?:\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)){3}"
 ipv4_regex = re.compile(_ipv4_regex, re.I)
 
-# IPv6 is complicated, so we have accommodate alternative patterns,
-# :(:[A-F0-9]{1,4}){1,7} == ::1, ::ffff:1
-# ([A-F0-9]{1,4}:){1,7}: == 2001::, 2001:db8::, 2001:db8:0:1:2:3::
-# ([A-F0-9]{1,4}:){1,6}:([A-F0-9]{1,4}) == 2001::1, 2001:db8::1, 2001:db8:0:1:2:3::1
-# ([A-F0-9]{1,4}:){7,7}([A-F0-9]{1,4}) == 1:1:1:1:1:1:1:1, ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
-
-_ipv6_regex = r"(:(:[A-F0-9]{1,4}){1,7}|([A-F0-9]{1,4}:){1,7}:|([A-F0-9]{1,4}:){1,6}:([A-F0-9]{1,4})|([A-F0-9]{1,4}:){7,7}([A-F0-9]{1,4}))"
+# IPv6 regex breakdown:
+#
+# (?:                            # —— address body ——
+# We have to individually account for all possible variations of: "N left hextets :: M right hextets" with N+M ≤ 8 or fully expanded 8 hextets.
+#   (?:[A-F0-9]{1,4}:){7}[A-F0-9]{1,4}        # 8 hextets, no compression.
+# | (?:[A-F0-9]{1,4}:){1,7}:                  # 1–7 left, then "::" (0 right).
+# | (?:[A-F0-9]{1,4}:){1,6}:[A-F0-9]{1,4}     # 1–6 left, "::", 1 right.
+# | (?:[A-F0-9]{1,4}:){1,5}(?::[A-F0-9]{1,4}){1,2}  # 1–5 left, "::", 1–2 right.
+# | (?:[A-F0-9]{1,4}:){1,4}(?::[A-F0-9]{1,4}){1,3}  # 1–4 left, "::", 1–3 right.
+# | (?:[A-F0-9]{1,4}:){1,3}(?::[A-F0-9]{1,4}){1,4}  # 1–3 left, "::", 1–4 right.
+# | (?:[A-F0-9]{1,4}:){1,2}(?::[A-F0-9]{1,4}){1,5}  # 1–2 left, "::", 1–5 right.
+# | [A-F0-9]{1,4}:(?::[A-F0-9]{1,4}){1,6}           # 1 left,    "::", 1–6 right.
+# | :(?::[A-F0-9]{1,4}){1,7}                        # 0 left,    "::", 1–7 right.
+# | ::                                              # all zeros.
+# )
+#
+# Notes:
+# - Does not match IPv4-embedded forms (e.g., ::ffff:192.0.2.1).
+# - Does not match zone IDs (e.g., %eth0).
+# - Pure syntax check; will not validate special ranges.
+
+_ipv6_regex = r"(?:(?:[A-F0-9]{1,4}:){7}[A-F0-9]{1,4}|(?:[A-F0-9]{1,4}:){1,7}:|(?:[A-F0-9]{1,4}:){1,6}:[A-F0-9]{1,4}|(?:[A-F0-9]{1,4}:){1,5}(?::[A-F0-9]{1,4}){1,2}|(?:[A-F0-9]{1,4}:){1,4}(?::[A-F0-9]{1,4}){1,3}|(?:[A-F0-9]{1,4}:){1,3}(?::[A-F0-9]{1,4}){1,4}|(?:[A-F0-9]{1,4}:){1,2}(?::[A-F0-9]{1,4}){1,5}|[A-F0-9]{1,4}:(?::[A-F0-9]{1,4}){1,6}|:(?::[A-F0-9]{1,4}){1,7}|::)"
 ipv6_regex = re.compile(_ipv6_regex, re.I)
 
 _ip_range_regexes = (
@@ -173,7 +188,9 @@ button_tag_regex2 = re.compile(
 )
 tag_attribute_regex = re.compile(r"<[^>]*(?:href|action|src)\s*=\s*[\"\']?(?!mailto:)([^\'\"\>]+)[\"\']?[^>]*>")
 
-valid_netloc = r"[^\s!@#$%^&()=/?\\'\";~`<>]+"
+_invalid_netloc_chars = r"\s!@#$%^&()=/?\\'\";~`<>"
+# first char must not be a colon, even though it's a valid char for a netloc
+valid_netloc = r"[^" + (_invalid_netloc_chars + ":") + r"]{1}[^" + _invalid_netloc_chars + "]*"
 
 _split_host_port_regex = r"(?:(?P<scheme>[a-z0-9]{1,20})://)?(?:[^?]*@)?(?P<netloc>" + valid_netloc + ")"
 split_host_port_regex = re.compile(_split_host_port_regex, re.I)

{bbot-2.6.0.6856rc0 → bbot-2.6.0.6871rc0}/bbot/test/test_step_1/test_helpers.py

@@ -155,6 +155,7 @@ async def test_helpers_misc(helpers, scan, bbot_scanner, bbot_httpserver):
     assert helpers.extract_host("https://[dead::beef]:22?a=b") == ("dead::beef", "https://[", "]:22?a=b")
     assert helpers.extract_host("https://[dead::beef]/?a=b") == ("dead::beef", "https://[", "]/?a=b")
     assert helpers.extract_host("https://[dead::beef]?a=b") == ("dead::beef", "https://[", "]?a=b")
+    assert helpers.extract_host("https://[::1]") == ("::1", "https://[", "]")
     assert helpers.extract_host("ftp://username:password@my-ftp.com/my-file.csv") == (
         "my-ftp.com",
         "ftp://username:password@",

{bbot-2.6.0.6856rc0 → bbot-2.6.0.6871rc0}/bbot/test/test_step_1/test_regexes.py

@@ -6,9 +6,6 @@ from bbot.core.helpers import regexes
 from bbot.errors import ValidationError
 from bbot.core.event.helpers import EventSeed
 
-# NOTE: :2001:db8:: will currently cause an exception...
-# e.g. raised unknown error: split_port() failed to parse netloc ":2001:db8::"
-
 
 def test_ip_regexes():
     bad_ip = [
@@ -23,6 +20,15 @@ def test_ip_regexes():
         "2001:db8:g::",  # includes non-hex character,
         "2001.db8.80",  # weird dot separated thing that might actually resolve as a DNS_NAME
         "9e:3e:53:29:43:64",  # MAC address, poor regex patterning will often detect these.
+        "2001:db8:1:2:3:4:5",  # only 7 groups, no zero-compression
+        "2001:db8:1:2:3:4:5:6:7",  # too many groups
+        "2001:db8::1::1",  # multiple ::
+        "2001:db8::zzzz",  # non-hex character
+        "2001:db8::12345",  # hex value too long
+        ":2001:db8::1",  # starts with :
+        ":2001:db8::",  # starts with :
+        "cafe:80",  # looks like open port
+        "12:34:56:78:9A:BC",  # mac address
     ]
 
     good_ip = [
@@ -46,6 +52,17 @@ def test_ip_regexes():
         "1::1",
         "ffff::ffff",
         "ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff",
+        "2001:db8::ff00:42:8329",
+        "2001:0db8:0000:0000:0000:0000:0000:0001",
+        "2001:db8:0:0:0:0:0:1",
+        "2001:db8::1",
+        "2001:db8::dead:beef",
+        "2001:db8:1:2:3:4:5:6",
+        "2001:db8:1:2:3:4:5:ffff",
+        "::",
+        "::ffff",
+        "::dead:beef",
+        "::DEAD:BEEF",
     ]
 
     ip_address_regexes = regexes.event_type_regexes["IP_ADDRESS"]
@@ -61,7 +78,7 @@ def test_ip_regexes():
                 if ip.startswith("["):
                     assert ip == "[2001:db8::]:80"
                 else:
-                    assert ip == "203.0.113.0:80"
+                    assert ip in ("cafe:80", "203.0.113.0:80")
                 continue
             if event_type == "DNS_NAME":
                 if ip.startswith("2001"):

{bbot-2.6.0.6856rc0 → bbot-2.6.0.6871rc0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "bbot"
-version = "v2.6.0.6856rc"
+version = "v2.6.0.6871rc"
 description = "OSINT automation for hackers."
 authors = [
     "TheTechromancer",