feat(asn): IP→ASN enrichment via iptoasn.com bulk dump

Mirrors decnet/geoip/ end-to-end: paths/base/factory/lookup at the
package level, iptoasn/ subpackage holds the data-source-specific
fetch+parse+provider. AsnLookup is bisect-indexed over (start, end,
AsnInfo) ranges with a pickled cache invalidated on raw-file mtime
bump.

Why iptoasn (and not bgp.tools / Team Cymru): public-domain dump,
zero attribution, no UA mandate, daily refresh — keeps DECNET stealth
intact (the geoip/rir module's "never identify as DECNET" comment
applies the same way here). bgp.tools' ToS would have required an
identifying UA, conflicting with feedback_stealth.

Public surface: decnet.asn.enrich_ip(ip) -> (asn, name, source) or
all-None on miss/disabled. Same shape as decnet.geoip.enrich_ip so
the profiler can compose them in one call site.
This commit is contained in:
2026-04-25 03:58:58 -04:00
parent 77a19ffe9f
commit 010568e558
14 changed files with 789 additions and 0 deletions

0
tests/asn/__init__.py Normal file
View File

22
tests/asn/conftest.py Normal file
View File

@@ -0,0 +1,22 @@
"""Per-package fixtures — sandbox the ASN provider into a tmp dir so no
real /var/lib/decnet paths get touched and no real iptoasn URL gets
fetched."""
from __future__ import annotations
from pathlib import Path
import pytest
@pytest.fixture(autouse=True)
def _asn_sandbox(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> Path:
monkeypatch.setenv("DECNET_ASN_ENABLED", "true")
monkeypatch.setenv("DECNET_ASN_ROOT", str(tmp_path))
import decnet.asn as _a
import decnet.asn.factory as _f
import decnet.asn.paths as _p
monkeypatch.setattr(_p, "ASN_ROOT", tmp_path)
_a._lookup = None
_a._provider_name = None
_f.reset_cache()
return tmp_path

74
tests/asn/test_lookup.py Normal file
View File

@@ -0,0 +1,74 @@
"""AsnLookup index tests."""
from __future__ import annotations
import ipaddress
from pathlib import Path
from decnet.asn.lookup import AsnInfo, AsnLookup
def _ip(s: str) -> int:
return int(ipaddress.IPv4Address(s))
def _fixture_lookup() -> AsnLookup:
return AsnLookup.from_ranges([
(_ip("8.8.8.0"), _ip("8.8.8.255"), AsnInfo(15169, "GOOGLE")),
(_ip("1.0.0.0"), _ip("1.0.0.255"), AsnInfo(13335, "CLOUDFLARENET")),
(_ip("46.101.0.0"), _ip("46.101.255.255"), AsnInfo(14061, "DIGITALOCEAN")),
])
def test_asn_hits_known_ranges() -> None:
lookup = _fixture_lookup()
assert lookup.asn("8.8.8.8").asn == 15169
assert lookup.asn("1.0.0.5").name == "CLOUDFLARENET"
assert lookup.asn("46.101.10.20").asn == 14061
def test_asn_misses_gap() -> None:
lookup = _fixture_lookup()
assert lookup.asn("9.0.0.0") is None
def test_asn_private_returns_none() -> None:
lookup = _fixture_lookup()
for ip in ("10.0.0.1", "192.168.1.1", "172.16.0.1", "127.0.0.1", "0.0.0.0"):
assert lookup.asn(ip) is None, ip
def test_asn_ipv6_returns_none() -> None:
lookup = _fixture_lookup()
assert lookup.asn("2001:db8::1") is None
assert lookup.asn("::1") is None
def test_asn_invalid_returns_none() -> None:
lookup = _fixture_lookup()
assert lookup.asn("not-an-ip") is None
assert lookup.asn("") is None
def test_lookup_roundtrips_through_pickle(tmp_path: Path) -> None:
lookup = _fixture_lookup()
cache = tmp_path / "idx.pkl"
lookup.save(cache)
loaded = AsnLookup.load(cache)
assert len(loaded) == len(lookup)
assert loaded.asn("8.8.8.8").asn == 15169
assert loaded.asn("8.8.8.8").name == "GOOGLE"
def test_from_ranges_last_writer_wins_on_collision() -> None:
lookup = AsnLookup.from_ranges([
(_ip("1.0.0.0"), _ip("1.0.0.255"), AsnInfo(1, "first")),
(_ip("1.0.0.0"), _ip("1.0.0.255"), AsnInfo(2, "second")),
])
assert lookup.asn("1.0.0.5").asn == 2
def test_boundary_inclusive() -> None:
lookup = _fixture_lookup()
assert lookup.asn("8.8.8.0").asn == 15169
assert lookup.asn("8.8.8.255").asn == 15169
assert lookup.asn("8.8.9.0") is None

57
tests/asn/test_parse.py Normal file
View File

@@ -0,0 +1,57 @@
"""Parser tests for the iptoasn TSV dump."""
from __future__ import annotations
import gzip
import ipaddress
from pathlib import Path
from decnet.asn.iptoasn.parse import parse_file
_FIXTURE_TSV = (
"1.0.0.0\t1.0.0.255\t13335\tUS\tCLOUDFLARENET\n"
"8.8.8.0\t8.8.8.255\t15169\tUS\tGOOGLE\n"
# ASN 0 sentinel — must be skipped.
"100.64.0.0\t100.127.255.255\t0\tNone\tNot routed\n"
# Malformed addresses — skipped.
"garbage\tnonsense\t12345\tXX\twhatever\n"
# Reversed range (end < start) — skipped.
"10.0.0.10\t10.0.0.5\t99999\tXX\tBackwards\n"
# Valid row with empty description.
"46.101.0.0\t46.101.255.255\t14061\tDE\t\n"
)
def test_parse_plain_tsv(tmp_path: Path) -> None:
fixture = tmp_path / "ip2asn-v4.tsv"
fixture.write_text(_FIXTURE_TSV)
ranges = list(parse_file(fixture))
asns = {r[2].asn for r in ranges}
assert asns == {13335, 15169, 14061}
def test_parse_gzipped(tmp_path: Path) -> None:
fixture = tmp_path / "ip2asn-v4.tsv.gz"
with gzip.open(fixture, "wt", encoding="utf-8") as fh:
fh.write(_FIXTURE_TSV)
ranges = list(parse_file(fixture))
asns = {r[2].asn for r in ranges}
assert 13335 in asns and 15169 in asns
def test_parse_range_boundaries(tmp_path: Path) -> None:
fixture = tmp_path / "ip2asn-v4.tsv"
fixture.write_text(_FIXTURE_TSV)
ranges = [r for r in parse_file(fixture) if r[2].asn == 15169]
assert len(ranges) == 1
start, end, info = ranges[0]
assert start == int(ipaddress.IPv4Address("8.8.8.0"))
assert end == int(ipaddress.IPv4Address("8.8.8.255"))
assert info.name == "GOOGLE"
def test_parse_empty_description_kept(tmp_path: Path) -> None:
fixture = tmp_path / "ip2asn-v4.tsv"
fixture.write_text(_FIXTURE_TSV)
ranges = [r for r in parse_file(fixture) if r[2].asn == 14061]
assert ranges[0][2].name == ""

View File

@@ -0,0 +1,95 @@
"""IptoasnProvider + factory + public API tests."""
from __future__ import annotations
import gzip
from pathlib import Path
import pytest
def _seed_fixture(root: Path, content: str = "8.8.8.0\t8.8.8.255\t15169\tUS\tGOOGLE\n") -> None:
target = root / "ip2asn-v4.tsv.gz"
with gzip.open(target, "wt", encoding="utf-8") as fh:
fh.write(content)
def test_factory_returns_iptoasn_by_default() -> None:
from decnet.asn.factory import get_provider
provider = get_provider()
assert provider.name == "iptoasn"
def test_factory_rejects_unknown_provider(monkeypatch: pytest.MonkeyPatch) -> None:
from decnet.asn import factory
monkeypatch.setenv("DECNET_ASN_PROVIDER", "nope")
factory.reset_cache()
with pytest.raises(ValueError):
factory.get_provider()
def test_provider_build_lookup_empty_when_no_files(tmp_path: Path) -> None:
from decnet.asn.iptoasn.provider import IptoasnProvider
p = IptoasnProvider()
lookup = p.build_lookup()
assert len(lookup) == 0
assert lookup.asn("8.8.8.8") is None
def test_provider_build_lookup_reads_present_file(tmp_path: Path) -> None:
from decnet.asn.iptoasn.provider import IptoasnProvider
_seed_fixture(tmp_path)
p = IptoasnProvider()
lookup = p.build_lookup()
info = lookup.asn("8.8.8.8")
assert info is not None
assert info.asn == 15169
assert info.name == "GOOGLE"
def test_provider_uses_cache_when_fresh(tmp_path: Path) -> None:
from decnet.asn.iptoasn.provider import IptoasnProvider
_seed_fixture(tmp_path)
p = IptoasnProvider()
a = p.build_lookup()
assert (tmp_path / ".iptoasn_index.pkl").exists()
p2 = IptoasnProvider()
b = p2.build_lookup()
assert len(b) == len(a)
def test_enrich_ip_short_circuits_when_disabled(monkeypatch: pytest.MonkeyPatch) -> None:
import decnet.asn as asn
monkeypatch.setenv("DECNET_ASN_ENABLED", "false")
assert asn.enrich_ip("8.8.8.8") == (None, None, None)
def test_enrich_ip_returns_asn_and_source(tmp_path: Path) -> None:
from decnet.asn import enrich_ip
_seed_fixture(tmp_path)
asn, name, src = enrich_ip("8.8.8.8")
assert asn == 15169
assert name == "GOOGLE"
assert src == "iptoasn"
def test_enrich_ip_private_returns_none(tmp_path: Path) -> None:
from decnet.asn import enrich_ip
_seed_fixture(tmp_path)
assert enrich_ip("192.168.1.1") == (None, None, None)
def test_enrich_ip_unannounced_returns_none(tmp_path: Path) -> None:
from decnet.asn import enrich_ip
_seed_fixture(tmp_path)
# 9.0.0.0 isn't in our fixture range — no BGP announcement we know of.
assert enrich_ip("9.0.0.0") == (None, None, None)