remove legacy scripts

.github/workflows/roa.yml

@@ -41,13 +41,13 @@ jobs:
           maxlen4=29
           maxlen6=64
           mkdir -p roa_dir
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6}    -o roa_dir/roa46_bird2.conf
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -4 -o roa_dir/roa4_bird2.conf
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -6 -o roa_dir/roa6_bird2.conf
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -j -o roa_dir/roa46.json
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -e -o roa_dir/neonetwork.json
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -r -o roa_dir/rfc8416.json
-          pyenv exec python3 scripts/roa-ng.py --summary --output roa_dir/README.txt
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6}    -o roa_dir/roa46_bird2.conf
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -4 -o roa_dir/roa4_bird2.conf
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -6 -o roa_dir/roa6_bird2.conf
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -j -o roa_dir/roa46.json
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -e -o roa_dir/neonetwork.json
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -r -o roa_dir/rfc8416.json
+          pyenv exec python3 scripts/roa.py --summary --output roa_dir/README.txt
 
       - name: Upload files
         env:

.github/workflows/test-your-pr.yml

@@ -41,9 +41,10 @@ jobs:
           maxlen4=29
           maxlen6=64
           mkdir -p roa_dir
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6}    -o roa_dir/roa46_bird2.conf
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -4 -o roa_dir/roa4_bird2.conf
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -6 -o roa_dir/roa6_bird2.conf
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -j -o roa_dir/roa46.json
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -e -o roa_dir/neonetwork.json
-          pyenv exec python3 scripts/roa-ng.py -m ${maxlen4} -M ${maxlen6} -r -o roa_dir/rfc8416.json
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6}    -o roa_dir/roa46_bird2.conf
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -4 -o roa_dir/roa4_bird2.conf
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -6 -o roa_dir/roa6_bird2.conf
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -j -o roa_dir/roa46.json
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -e -o roa_dir/neonetwork.json
+          pyenv exec python3 scripts/roa.py -m ${maxlen4} -M ${maxlen6} -r -o roa_dir/rfc8416.json
+          pyenv exec python3 scripts/roa.py --summary --output roa_dir/README.txt

scripts/roa-ng.py

@@ -1,327 +0,0 @@
-#!/usr/bin/env python3
-import argparse
-import json
-import time
-from collections import defaultdict
-from contextlib import redirect_stdout
-from io import StringIO
-from ipaddress import IPv4Network, IPv6Network, ip_network
-from itertools import combinations
-from pathlib import Path
-
-import toml
-
-NEO_NETWORK_POOL = [ip_network("10.127.0.0/16"), ip_network("fd10:127::/32")]
-
-
-def pick(entity: dict, fields: [str], **kwargs: dict):
-    new_entity = {}
-    for field in fields:
-        new_entity[field] = entity.get(field)
-    for old_field, new_field in kwargs.items():
-        new_entity[new_field] = entity.get(old_field)
-    return new_entity
-
-
-def is_neo_network(address):
-    return any(
-        address.version == neo.version and address.subnet_of(neo)
-        for neo in NEO_NETWORK_POOL
-    )
-
-
-def is_neo_network_asn(asn: int):
-    return 4201270000 <= asn <= 4201279999
-
-
-def is_dn42_asn(asn: int):
-    return 4242420000 <= asn <= 4242429999
-
-
-def iter_toml_file(path: str):
-    for item in Path(path).iterdir():
-        if not item.is_file() or item.suffix != ".toml":
-            continue
-        yield item, toml.loads(item.read_text())
-
-
-def load_entities():
-    return {item.stem: entity for item, entity in iter_toml_file("entity")}
-
-
-def load_asn(entities: dict):
-    def assert_entity(entity, asn):
-        owner = entity.get("owner")
-        source = entity.get("source")
-        if is_neo_network_asn(asn):
-            source = "NeoNetwork"
-        elif is_dn42_asn(asn):
-            source = "DN42"
-        entity["source"] = source
-        assert owner in entities
-        assert source in ["NeoNetwork", "DN42", "Internet"]
-        return entity
-
-    mapping = {
-        int(item.stem.lstrip("AS")): entity for item, entity in iter_toml_file("asn")
-    }
-    return {asn: assert_entity(entity, asn) for asn, entity in mapping.items()}
-
-
-def node_to_asn(orignal_asn_set: set):
-    node_table = dict()
-    for _, entities in iter_toml_file("node"):
-        mapping = {name: entity["asn"] for (name, entity) in entities.items()}
-        asn_set = set(mapping.values())
-        assert orignal_asn_set & asn_set == asn_set
-        node_table.update(mapping)
-    return node_table
-
-
-def assert_peer(nodes: set):
-    for item, entities in iter_toml_file("peer"):
-        peers = set(entities["to-peer"])
-        assert item.stem in nodes
-        assert nodes & peers == peers
-
-
-def route_to_roa(asn_table: dict):
-    def make_route():
-        for item, entity in iter_toml_file("route"):
-            asn = int(item.stem.lstrip("AS"))
-            for prefix, fields in entity.items():
-                if fields["type"] not in ("loopback", "subnet"):
-                    continue
-                fields["asn"] = asn
-                fields["prefix"] = ip_network(prefix, strict=True)
-                supernet = fields.get("supernet")
-                fields["supernet"] = (
-                    ip_network(supernet, strict=True) if supernet else None
-                )
-                assert fields["name"]
-                assert is_neo_network(fields["prefix"])
-                assert not fields["supernet"] or is_neo_network(fields["supernet"])
-                yield fields
-
-    entities = (
-        pick(route, ["asn", "name", "prefix", "supernet"]) for route in make_route()
-    )
-    entities = sorted(entities, key=lambda item: item["asn"])
-    prefixes = [item["prefix"] for item in entities]
-    for net1, net2 in combinations(
-        sorted(entities, key=lambda net: net["prefix"].prefixlen), 2
-    ):
-        if not net1["prefix"].overlaps(net2["prefix"]):
-            continue
-        assert net1["prefix"] != net2["prefix"]
-        assert net1["prefix"].supernet_of(net2["prefix"])
-        s1net, s2net = (net1["supernet"], net2["supernet"])
-        assert s2net  # please include supernet = <cidr> in your route
-        # if net1(the bigger net) has a supernet s1net, then s1net and net1
-        # will be checked or must have been checked, same for net2
-        assert not s1net or s1net in prefixes  # net1.supernet is garbage
-        assert s2net == net1["prefix"] or s2net in prefixes  # net2.supernet is garbage
-    return entities
-
-
-def prehandle_roa(asn_table: dict, args):
-    roa = route_to_roa(asn_table)
-    max_prefixlen = IPv4Network(0).max_prefixlen
-    roa4 = filter(lambda item: isinstance(item["prefix"], IPv4Network), roa)
-    roa6 = filter(lambda item: isinstance(item["prefix"], IPv6Network), roa)
-    if args.ipv4:
-        roa6 = []
-    elif args.ipv6:
-        roa4 = []
-    roa4 = [
-        r
-        for r in roa4
-        if r["prefix"].prefixlen <= args.max or r["prefix"].prefixlen == max_prefixlen
-    ]
-    roa6 = [r for r in roa6 if r["prefix"].prefixlen <= args.max6]
-    for r in roa4:
-        r["maxLength"] = args.max
-        if r["prefix"].prefixlen == max_prefixlen:
-            r["maxLength"] = max_prefixlen
-    for r in roa6:
-        r["maxLength"] = args.max6
-    for r in (*roa4, *roa6):
-        r["prefix"] = r["prefix"].with_prefixlen
-    return roa4, roa6
-
-
-def make_export(roa4, roa6):
-    entities = load_entities()
-    asn_table = load_asn(entities)
-    current = int(time.time())
-    output = {
-        "metadata": {"generated": current, "valid": current + 14 * 86400},
-        "people": {
-            owner: {"info": entity, "asns": []} for owner, entity in entities.items()
-        },
-    }
-    for asn, asn_info in asn_table.items():
-        owner = asn_info["owner"]
-        asn_item = {
-            "asn": asn,
-            "name": asn_info["name"],
-            "source": asn_info["source"],
-            "routes": {
-                "ipv4": [
-                    pick(roa, ["prefix", "maxLength"])
-                    for roa in roa4
-                    if roa["asn"] == asn
-                ],
-                "ipv6": [
-                    pick(roa, ["prefix", "maxLength"])
-                    for roa in roa6
-                    if roa["asn"] == asn
-                ],
-            },
-        }
-        output["people"][owner]["asns"].append(asn_item)
-    return json.dumps(output, indent=2)
-
-
-def make_json(roa4, roa6):
-    current = int(time.time())
-    output = {
-        "metadata": {
-            "counts": len(roa4) + len(roa6),
-            "generated": current,
-            "valid": current + 14 * 86400,
-        },
-        "roas": [
-            {"asn": "AS%d" % roa["asn"], **pick(roa, ["prefix", "maxLength"])}
-            for roa in (*roa4, *roa6)
-        ],
-    }
-    return json.dumps(output, indent=2)
-
-
-def make_rfc8416(roa4, roa6):
-    output = {
-        "slurmVersion": 1,
-        "validationOutputFilters": {"prefixFilters": [], "bgpsecFilters": []},
-        "locallyAddedAssertions": {
-            "bgpsecAssertions": [],
-            "prefixAssertions": [
-                pick(
-                    roa, ["asn", "prefix"], maxLength="maxPrefixLength", name="comment",
-                )
-                for roa in (*roa4, *roa6)
-            ],
-        },
-    }
-    return json.dumps(output, indent=2)
-
-
-def make_roa_records(roa4, roa6):
-    records = [
-        "route {asn} max {prefix} as {maxLength};".format_map(roa)
-        for roa in (*roa4, *roa6)
-    ]
-    return "\n".join(["# NeoNetwork ROA tool", "", *records])
-
-
-def make_summary():
-    entities = load_entities()
-    asn_table = load_asn(entities)
-    node_table = node_to_asn(set(asn_table.keys()))
-    stream = StringIO()
-    with redirect_stdout(stream):
-        print("Entity table:")
-        print("{:20}{:20}{}".format("Name", "Telegram", "Email"))
-        for entity in sorted(
-            entities.values(), key=lambda entity: entity["name"].lower(),
-        ):
-            contact = entity.get("contact", {})
-            email = contact.get("email", "")
-            telegram = contact.get("telegram", "")
-            print("{:20}{:20}{}".format(entity["name"], telegram, email))
-        print()
-        print("AS table:")
-        print("{:15}{:<17}{:20}{}".format("Source", "ASN", "Owner", "Name"))
-        for asn, entity in sorted(asn_table.items(), key=lambda item: item[0]):
-            print(
-                "{:15}AS{:<15}{:20}{}".format(
-                    entity["source"], asn, entity["owner"], entity["name"]
-                )
-            )
-        print()
-        print("Node table:")
-        print("{:<17}{}".format("ASN", "Name"))
-        for name, asn in sorted(node_table.items(), key=lambda item: item[1]):
-            print("AS{:<15}{}".format(asn, name))
-        print()
-        print("Peer table:")
-        peers = {
-            item.stem: entity["to-peer"] for item, entity in iter_toml_file("peer")
-        }
-        peers = [
-            (upstream, downstream)
-            for upstream, downstream_list in peers.items()
-            for downstream in downstream_list
-        ]
-        print("{:>20} ~ {}".format("Upstream", "Downstream"))
-        for upstream, downstream in peers:
-            print("{:>20} ~ {}".format(upstream, downstream))
-        print()
-        print("Route table:")
-        print("{:17}{:30}{:30}{}".format("ASN", "Name", "Prefix", "Supernet"))
-        for entity in route_to_roa(asn_table):
-            entity["prefix"] = str(entity["prefix"])
-            entity["supernet"] = str(entity["supernet"]) if entity["supernet"] else ""
-            print("AS{asn:<15}{name:30}{prefix:30}{supernet}".format_map(entity))
-    return stream.getvalue()
-
-
-def main(args):
-    entities = load_entities()
-    asn_table = load_asn(entities)
-    node_table = node_to_asn(set(asn_table.keys()))
-    assert_peer(set(node_table.keys()))
-    roa4, roa6 = prehandle_roa(asn_table, args)
-    if args.export:
-        return make_export(roa4, roa6)
-    elif args.json:
-        return make_json(roa4, roa6)
-    elif args.rfc8416:
-        return make_rfc8416(roa4, roa6)
-    elif args.summary:
-        return make_summary()
-    else:
-        return make_roa_records(roa4, roa6)
-
-
-if __name__ == "__main__":
-    parser = argparse.ArgumentParser(description="NeoNetwork ROA tool")
-    parser.add_argument(
-        "-m", "--max", type=int, default=29, help="set ipv4 max prefix length"
-    )
-    parser.add_argument(
-        "-M", "--max6", type=int, default=64, help="set ipv6 max prefix length"
-    )
-    parser.add_argument("-j", "--json", action="store_true", help="output json")
-    parser.add_argument("-r", "--rfc8416", action="store_true", help="output rfc8416")
-    parser.add_argument("-s", "--summary", action="store_true", help="output summary")
-    parser.add_argument("-o", "--output", default="", help="write output to file")
-    parser.add_argument("-4", "--ipv4", action="store_true", help="print ipv4 only")
-    parser.add_argument("-6", "--ipv6", action="store_true", help="print ipv6 only")
-    parser.add_argument(
-        "-e", "--export", action="store_true", help="export registry to json"
-    )
-    args = parser.parse_args()
-    if (
-        args.max < 0
-        or args.max6 < 0
-        or args.max > IPv4Network(0).max_prefixlen
-        or args.max6 > IPv6Network(0).max_prefixlen
-    ):
-        parser.error("check your max prefix length")
-    output = main(args)
-    if not args.output or args.output == "-":
-        print(output)
-    elif output:
-        Path(args.output).write_text(output)
-        print("written to", args.output)

scripts/roa.py

@@ -1,301 +1,327 @@
 #!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-
-from pathlib import Path
+import argparse
+import json
+import time
+from collections import defaultdict
+from contextlib import redirect_stdout
+from io import StringIO
 from ipaddress import IPv4Network, IPv6Network, ip_network
 from itertools import combinations
-import re
+from pathlib import Path
 
-NEONET_ADDR_POOL = ('10.127.0.0/16', 'fd10:127::/32')
-NEONET_ADDR_POOL = [ip_network(neo) for neo in NEONET_ADDR_POOL]
-IS_NEONET = lambda net: bool([True for neo in NEONET_ADDR_POOL if net.version == neo.version and net.subnet_of(neo)])
-AS_IS_NEONET = lambda asn: 4201270000 <= asn <= 4201279999
-assert hasattr(IPv4Network, 'subnet_of') # needs at least python 3.7
+import toml
 
-class BashParser:
-    def __init__(self):
-        self.__pa = None # are we parsing bash array?
-    def end(self):
-        assert not self.__pa # check if array ends properly
-    def parseline(self, line):
-        repl_quotes = lambda t: t.replace('"', '').replace('\'', '')
-        line = line.strip()
-        if '=(' in line:
-            self.__pa = (repl_quotes(line).split('=(')[0], list())
-            return None
-        if self.__pa:
-            if line:
-                if line.endswith(')'):
-                    if line[:-1]:
-                        self.__pa[1].append(repl_quotes(line[:-1]))
-                    ret = self.__pa
-                    self.__pa = None
-                    return ret
-                else:
-                    self.__pa[1].append(repl_quotes(line))
-            return None
-        else:
-            if not line or line.startswith('#'):
-                return None
-            l = line.split('=')
-            assert len(l) >= 2 # is this line key=value syntax?
-            return [l[0], '='.join([repl_quotes(i) for i in l[1:]])]
-bp = BashParser()
+NEO_NETWORK_POOL = [ip_network("10.127.0.0/16"), ip_network("fd10:127::/32")]
 
-def shell2dict(shellscript):
-    fc = dict()
-    for line in shellscript.split('\n'):
-        r = bp.parseline(line)
-        if r:
-            key, val = r
-            fc[key.lower()] = val
-    bp.end()
-    return fc
 
-cwd = Path()
-assert not [d for d in ("asn", "route", "route6", "node", "entity") if not (cwd / d).is_dir()]
+def pick(entity: dict, fields: [str], **kwargs: dict):
+    new_entity = {}
+    for field in fields:
+        new_entity[field] = entity.get(field)
+    for old_field, new_field in kwargs.items():
+        new_entity[new_field] = entity.get(old_field)
+    return new_entity
 
-def str2asn(s_asn, strict=None):
-    """ strict: 1 with prefix AS, other bool(x)==True without prefix AS """
-    s_asn = s_asn.strip().lower()
-    if strict == 1:
-        assert s_asn.startswith('as')
-    elif strict:
-        assert int(s_asn)
-    if s_asn.startswith('as'):
-        s_asn = s_asn[2:]
-    return int(s_asn)
 
-def name2nichdl(name):
-    r, num = re.subn(r'[^0-9A-Z]', '-', name.upper())
-    _r = len(r.replace('-', ''))
-    assert _r >= 3     # has at least 3 effective chars
-    assert r[0] != '-' # starts with [0-9A-Z]
-    assert num < _r    # not too many subs
-    return r
+def is_neo_network(address):
+    return any(
+        address.version == neo.version and address.subnet_of(neo)
+        for neo in NEO_NETWORK_POOL
+    )
 
-def neoneo_get_people():
-    nic_hdl_names = set()
-    people = dict()
-    for f in (cwd / "entity").iterdir():
-        try:
-            if not f.is_file():
-                continue
-            fc = shell2dict(f.read_text())
-            present_keys = ('name', 'desc', 'contact', 'babel', 'auth')
-            assert f.name
-            people[f.name] = {k: fc.get(k) for k in present_keys}
-            nic_hdl = name2nichdl(f.name)
-            assert nic_hdl not in nic_hdl_names # nic_hdl collision
-            nic_hdl_names.add(nic_hdl)
-            people[f.name]['nic_hdl'] = nic_hdl
-            for v in people[f.name].values():
-                assert v is not None
-            auth = people[f.name]['auth']
-            if auth:
-                method, data = auth.split(':')
-                assert method in ('PGP', 'SSH')
-                if method == 'PGP':
-                    assert len(data) == 40 # invaild pgp fingerprint
-                elif method == 'SSH':
-                    assert data.startswith('ssh-') # invalid ssh pubkey
-                people[f.name]['auth'] = f"{'pgp-fingerprint ' if method == 'PGP' else ''}{data.strip()}"
-        except Exception:
-            print("[!] Error while processing file", f)
-            raise
-    return people
-PEOPLE = neoneo_get_people()
 
-def neonet_get_asns():
-    asns = dict()
-    for f in (cwd / "asn").iterdir():
-        try:
-            if not f.is_file():
-                continue
-            fc = shell2dict(f.read_text())
-            present_keys = ('name', 'owner', 'desc', 'source')
-            required_keys = ('name', 'owner', 'source')
-            asn = str2asn(f.name, 1)
-            asns[asn] = {k: fc.get(k) for k in present_keys}
-            assert fc.get('owner') in PEOPLE
-            for k in required_keys:
-                assert asns[asn].get(k) # required fields missing
-            as_src = asns[asn].get('source')
-            assert as_src in ('NeoNetwork', 'DN42', 'INET')
-            assert AS_IS_NEONET(asn) or as_src != 'NeoNetwork'
-        except Exception:
-            print("[!] Error while processing file", f)
-            raise
-    return asns
-ASNS = neonet_get_asns()
+def is_neo_network_asn(asn: int):
+    return 4201270000 <= asn <= 4201279999
 
-def node2asn():
+
+def is_dn42_asn(asn: int):
+    return 4242420000 <= asn <= 4242429999
+
+
+def iter_toml_file(path: str):
+    for item in Path(path).iterdir():
+        if not item.is_file() or item.suffix != ".toml":
+            continue
+        yield item, toml.loads(item.read_text())
+
+
+def load_entities():
+    return {item.stem: entity for item, entity in iter_toml_file("entity")}
+
+
+def load_asn(entities: dict):
+    def assert_entity(entity, asn):
+        owner = entity.get("owner")
+        source = entity.get("source")
+        if is_neo_network_asn(asn):
+            source = "NeoNetwork"
+        elif is_dn42_asn(asn):
+            source = "DN42"
+        entity["source"] = source
+        assert owner in entities
+        assert source in ["NeoNetwork", "DN42", "Internet"]
+        return entity
+
+    mapping = {
+        int(item.stem.lstrip("AS")): entity for item, entity in iter_toml_file("asn")
+    }
+    return {asn: assert_entity(entity, asn) for asn, entity in mapping.items()}
+
+
+def node_to_asn(orignal_asn_set: set):
     node_table = dict()
-    for f in (cwd / "node").iterdir():
-        try:
-            if not f.is_file():
-                continue
-            fc = shell2dict(f.read_text())
-            asn = str2asn(fc.get('asn'), 2)
-            assert asn in ASNS
-            node_table[f.name] = asn
-        except Exception:
-            print("[!] Error while processing file", f)
-            raise
+    for _, entities in iter_toml_file("node"):
+        mapping = {name: entity["asn"] for (name, entity) in entities.items()}
+        asn_set = set(mapping.values())
+        assert orignal_asn_set & asn_set == asn_set
+        node_table.update(mapping)
     return node_table
-NODE_TABLE = node2asn()
 
-def neonet_peercheck():
-    for f in (cwd / "peer").iterdir():
-        try:
-            if not f.is_file():
-                continue
-            fc = shell2dict(f.read_text())
-            desc = fc.get('desc')
-            assert desc is not None
-            upstream, downstream = f.name.split('~')
-            assert upstream in NODE_TABLE
-            assert downstream in NODE_TABLE
-        except Exception:
-            print("[!] Error while processing file", f)
-            raise
-neonet_peercheck()
 
-def neonet_route2roa(dirname, is_ipv6=False):
-    roa_entries = list()
-    for f in (cwd / dirname).iterdir():
-        try:
-            if not f.is_file():
-                continue
-            fc = shell2dict(f.read_text())
-            nettype = IPv6Network if is_ipv6 else IPv4Network
-            get_supernet = lambda s_net: None if not s_net else nettype(s_net, strict=True)
-            roa_entries_key = ("asn", "prefix", "supernet", "netname")
-            if fc.get('type').lower() in ('lo', 'subnet'):
-                asn = str2asn(fc.get('asn'), 2)
-                assert asn in ASNS # asn not in as-dir
-                route = f.name.replace(',', '/')
-                net = nettype(route, strict=True)
-                assert IS_NEONET(net)
-                supernet = get_supernet(fc.get('supernet'))
-                assert not supernet or IS_NEONET(supernet)
-                netname = fc.get('name')
-                assert netname
-                roa_entries.append(dict(zip(roa_entries_key, [asn, net, supernet, netname])))
-            else:
-                raise AssertionError # unknown type
-        except Exception:
-            print("[!] Error while processing file", f)
-            raise
-    roa_entries.sort(key=lambda l: l['asn'])
-    l_prefix = [_roa['prefix'] for _roa in roa_entries]
-    for _net1, _net2 in combinations(roa_entries, 2):
-        net1, net2 = sorted([_net1, _net2], key=lambda net: net['prefix'].prefixlen)
-        if net1['prefix'].overlaps(net2['prefix']):
-            try:
-                assert net1['prefix'] != net2['prefix']
-                assert net1['prefix'].supernet_of(net2['prefix'])
-                s1net, s2net= (net1['supernet'], net2['supernet'])
-                assert s2net # please include SUPERNET=<cidr> in your route
-                # if net1(the bigger net) has a supernet s1net, then s1net and net1
-                # will be checked or must have been checked, same for net2
-                assert not s1net               or s1net in l_prefix # net1.supernet is garbage
-                assert s2net == net1['prefix'] or s2net in l_prefix # net2.supernet is garbage
-            except AssertionError:
-                print("[!] Error: found", net1, "overlaps", net2)
-                raise
-    return roa_entries
+def assert_peer(nodes: set):
+    for item, entities in iter_toml_file("peer"):
+        peers = set(entities["to-peer"])
+        assert item.stem in nodes
+        assert nodes & peers == peers
+
+
+def route_to_roa(asn_table: dict):
+    def make_route():
+        for item, entity in iter_toml_file("route"):
+            asn = int(item.stem.lstrip("AS"))
+            for prefix, fields in entity.items():
+                if fields["type"] not in ("loopback", "subnet"):
+                    continue
+                fields["asn"] = asn
+                fields["prefix"] = ip_network(prefix, strict=True)
+                supernet = fields.get("supernet")
+                fields["supernet"] = (
+                    ip_network(supernet, strict=True) if supernet else None
+                )
+                assert fields["name"]
+                assert is_neo_network(fields["prefix"])
+                assert not fields["supernet"] or is_neo_network(fields["supernet"])
+                yield fields
+
+    entities = (
+        pick(route, ["asn", "name", "prefix", "supernet"]) for route in make_route()
+    )
+    entities = sorted(entities, key=lambda item: item["asn"])
+    prefixes = [item["prefix"] for item in entities]
+    for net1, net2 in combinations(
+        sorted(entities, key=lambda net: net["prefix"].prefixlen), 2
+    ):
+        if not net1["prefix"].overlaps(net2["prefix"]):
+            continue
+        assert net1["prefix"] != net2["prefix"]
+        assert net1["prefix"].supernet_of(net2["prefix"])
+        s1net, s2net = (net1["supernet"], net2["supernet"])
+        assert s2net  # please include supernet = <cidr> in your route
+        # if net1(the bigger net) has a supernet s1net, then s1net and net1
+        # will be checked or must have been checked, same for net2
+        assert not s1net or s1net in prefixes  # net1.supernet is garbage
+        assert s2net == net1["prefix"] or s2net in prefixes  # net2.supernet is garbage
+    return entities
+
+
+def prehandle_roa(asn_table: dict, args):
+    roa = route_to_roa(asn_table)
+    max_prefixlen = IPv4Network(0).max_prefixlen
+    roa4 = filter(lambda item: isinstance(item["prefix"], IPv4Network), roa)
+    roa6 = filter(lambda item: isinstance(item["prefix"], IPv6Network), roa)
+    if args.ipv4:
+        roa6 = []
+    elif args.ipv6:
+        roa4 = []
+    roa4 = [
+        r
+        for r in roa4
+        if r["prefix"].prefixlen <= args.max or r["prefix"].prefixlen == max_prefixlen
+    ]
+    roa6 = [r for r in roa6 if r["prefix"].prefixlen <= args.max6]
+    for r in roa4:
+        r["maxLength"] = args.max
+        if r["prefix"].prefixlen == max_prefixlen:
+            r["maxLength"] = max_prefixlen
+    for r in roa6:
+        r["maxLength"] = args.max6
+    for r in (*roa4, *roa6):
+        r["prefix"] = r["prefix"].with_prefixlen
+    return roa4, roa6
+
+
+def make_export(roa4, roa6):
+    entities = load_entities()
+    asn_table = load_asn(entities)
+    current = int(time.time())
+    output = {
+        "metadata": {"generated": current, "valid": current + 14 * 86400},
+        "people": {
+            owner: {"info": entity, "asns": []} for owner, entity in entities.items()
+        },
+    }
+    for asn, asn_info in asn_table.items():
+        owner = asn_info["owner"]
+        asn_item = {
+            "asn": asn,
+            "name": asn_info["name"],
+            "source": asn_info["source"],
+            "routes": {
+                "ipv4": [
+                    pick(roa, ["prefix", "maxLength"])
+                    for roa in roa4
+                    if roa["asn"] == asn
+                ],
+                "ipv6": [
+                    pick(roa, ["prefix", "maxLength"])
+                    for roa in roa6
+                    if roa["asn"] == asn
+                ],
+            },
+        }
+        output["people"][owner]["asns"].append(asn_item)
+    return json.dumps(output, indent=2)
+
+
+def make_json(roa4, roa6):
+    current = int(time.time())
+    output = {
+        "metadata": {
+            "counts": len(roa4) + len(roa6),
+            "generated": current,
+            "valid": current + 14 * 86400,
+        },
+        "roas": [
+            {"asn": "AS%d" % roa["asn"], **pick(roa, ["prefix", "maxLength"])}
+            for roa in (*roa4, *roa6)
+        ],
+    }
+    return json.dumps(output, indent=2)
+
+
+def make_rfc8416(roa4, roa6):
+    output = {
+        "slurmVersion": 1,
+        "validationOutputFilters": {"prefixFilters": [], "bgpsecFilters": []},
+        "locallyAddedAssertions": {
+            "bgpsecAssertions": [],
+            "prefixAssertions": [
+                pick(
+                    roa, ["asn", "prefix"], maxLength="maxPrefixLength", name="comment",
+                )
+                for roa in (*roa4, *roa6)
+            ],
+        },
+    }
+    return json.dumps(output, indent=2)
+
+
+def make_roa_records(roa4, roa6):
+    records = [
+        "route {asn} max {prefix} as {maxLength};".format_map(roa)
+        for roa in (*roa4, *roa6)
+    ]
+    return "\n".join(["# NeoNetwork ROA tool", "", *records])
+
+
+def make_summary():
+    entities = load_entities()
+    asn_table = load_asn(entities)
+    node_table = node_to_asn(set(asn_table.keys()))
+    stream = StringIO()
+    with redirect_stdout(stream):
+        print("Entity table:")
+        print("{:20}{:20}{}".format("Name", "Telegram", "Email"))
+        for entity in sorted(
+            entities.values(), key=lambda entity: entity["name"].lower(),
+        ):
+            contact = entity.get("contact", {})
+            email = contact.get("email", "")
+            telegram = contact.get("telegram", "")
+            print("{:20}{:20}{}".format(entity["name"], telegram, email))
+        print()
+        print("AS table:")
+        print("{:15}{:<17}{:20}{}".format("Source", "ASN", "Owner", "Name"))
+        for asn, entity in sorted(asn_table.items(), key=lambda item: item[0]):
+            print(
+                "{:15}AS{:<15}{:20}{}".format(
+                    entity["source"], asn, entity["owner"], entity["name"]
+                )
+            )
+        print()
+        print("Node table:")
+        print("{:<17}{}".format("ASN", "Name"))
+        for name, asn in sorted(node_table.items(), key=lambda item: item[1]):
+            print("AS{:<15}{}".format(asn, name))
+        print()
+        print("Peer table:")
+        peers = {
+            item.stem: entity["to-peer"] for item, entity in iter_toml_file("peer")
+        }
+        peers = [
+            (upstream, downstream)
+            for upstream, downstream_list in peers.items()
+            for downstream in downstream_list
+        ]
+        print("{:>20} ~ {}".format("Upstream", "Downstream"))
+        for upstream, downstream in peers:
+            print("{:>20} ~ {}".format(upstream, downstream))
+        print()
+        print("Route table:")
+        print("{:17}{:30}{:30}{}".format("ASN", "Name", "Prefix", "Supernet"))
+        for entity in route_to_roa(asn_table):
+            entity["prefix"] = str(entity["prefix"])
+            entity["supernet"] = str(entity["supernet"]) if entity["supernet"] else ""
+            print("AS{asn:<15}{name:30}{prefix:30}{supernet}".format_map(entity))
+    return stream.getvalue()
+
+
+def main(args):
+    entities = load_entities()
+    asn_table = load_asn(entities)
+    node_table = node_to_asn(set(asn_table.keys()))
+    assert_peer(set(node_table.keys()))
+    roa4, roa6 = prehandle_roa(asn_table, args)
+    if args.export:
+        return make_export(roa4, roa6)
+    elif args.json:
+        return make_json(roa4, roa6)
+    elif args.rfc8416:
+        return make_rfc8416(roa4, roa6)
+    elif args.summary:
+        return make_summary()
+    else:
+        return make_roa_records(roa4, roa6)
+
 
 if __name__ == "__main__":
-    import argparse
-    parser = argparse.ArgumentParser(description='NeoNetwork ROA tool')
-    parser.add_argument('-m', '--max', type=int, default=29, help='set ipv4 max prefix length')
-    parser.add_argument('-M', '--max6', type=int, default=64, help='set ipv6 max prefix length')
-    parser.add_argument('-j', '--json', action='store_true', help='output json')
-    parser.add_argument('-r', '--rfc8416', action='store_true', help='output rfc8416')
-    parser.add_argument('-o', '--output', default='', help='write output to file')
-    parser.add_argument('-4', '--ipv4', action='store_true', help='print ipv4 only')
-    parser.add_argument('-6', '--ipv6', action='store_true', help='print ipv6 only')
-    parser.add_argument('-e', '--export', action='store_true', help='export registry to json')
+    parser = argparse.ArgumentParser(description="NeoNetwork ROA tool")
+    parser.add_argument(
+        "-m", "--max", type=int, default=29, help="set ipv4 max prefix length"
+    )
+    parser.add_argument(
+        "-M", "--max6", type=int, default=64, help="set ipv6 max prefix length"
+    )
+    parser.add_argument("-j", "--json", action="store_true", help="output json")
+    parser.add_argument("-r", "--rfc8416", action="store_true", help="output rfc8416")
+    parser.add_argument("-s", "--summary", action="store_true", help="output summary")
+    parser.add_argument("-o", "--output", default="", help="write output to file")
+    parser.add_argument("-4", "--ipv4", action="store_true", help="print ipv4 only")
+    parser.add_argument("-6", "--ipv6", action="store_true", help="print ipv6 only")
+    parser.add_argument(
+        "-e", "--export", action="store_true", help="export registry to json"
+    )
     args = parser.parse_args()
-    if args.max < 0 or args.max6 < 0 or args.max > IPv4Network(0).max_prefixlen or args.max6 > IPv6Network(0).max_prefixlen:
-        parser.error('check your max prefix length')
-
-    roa4 = roa6 = list()
-    if args.ipv4:
-        roa4 = neonet_route2roa('route')
-    elif args.ipv6:
-        roa6 = neonet_route2roa('route6', True)
-    else:
-        roa4 = neonet_route2roa('route')
-        roa6 = neonet_route2roa('route6', True)
-
-    roa4 = [r for r in roa4 if r['prefix'].prefixlen <= args.max or r['prefix'].prefixlen == IPv4Network(0).max_prefixlen]
-    roa6 = [r for r in roa6 if r['prefix'].prefixlen <= args.max6]
-
-    for r in roa4:
-        if r['prefix'].prefixlen == IPv4Network(0).max_prefixlen:
-            r['maxLength'] = IPv4Network(0).max_prefixlen
-        else:
-            r['maxLength'] = args.max
-    for r in roa6:
-        r['maxLength'] = args.max6
-    for r in (*roa4, *roa6):
-        r['prefix'] = r['prefix'].with_prefixlen
-
-
-    output = ""
-    VALID_KEYS = ('asn', 'prefix', 'maxLength')
-    if args.export:
-        import json, time
-        current = int(time.time())
-        # people has [asns], asn has [route]
-        d_output = {"metadata": {"generated": current, "valid": current+14*86400}, "people": dict()}
-        for asn, asi in ASNS.items():
-            as_route4 = list()
-            as_route6 = list()
-            vkeys = [k for k in VALID_KEYS if k != 'asn']
-            vkeys.append('netname')
-            for roa, as_route in ((roa4, as_route4), (roa6, as_route6)):
-                for r in roa:
-                    if r['asn'] == asn:
-                        as_route.append({k:v for k, v in r.items() if k in vkeys})
-            owner = asi['owner']
-            peopledict = d_output['people'].setdefault(owner, {"info": PEOPLE[owner], "asns": list()})
-            peopledict['asns'].append({"asn": asn, **{k:v for k, v in ASNS[asn].items() if k != 'owner'},
-                                       "routes": {'ipv4': as_route4, 'ipv6': as_route6}})
-        output = json.dumps(d_output, indent=2)
-    elif args.json:
-        import json, time
-        current = int(time.time())
-        d_output = {"metadata": {"counts": len(roa4)+len(roa6), "generated": current, "valid": current+14*86400}, "roas": list()}
-        for r in (*roa4, *roa6):
-            # some preprocessing
-            r['asn'] = "AS%d" % r['asn']
-        for r in (*roa4, *roa6):
-            d_output['roas'].append({k:v for k, v in r.items() if k in VALID_KEYS})
-        output = json.dumps(d_output, indent=2)
-    elif args.rfc8416:
-        import json
-        l_prefix = list()
-        d_output = {'slurmVersion': 1, 'validationOutputFilters': {'prefixFilters': list(),
-                    'bgpsecFilters': list()}, 'locallyAddedAssertions': {'bgpsecAssertions': list(),
-                    'prefixAssertions': l_prefix}}
-        for r in (*roa4, *roa6):
-            vkeys = ["netname", "maxLength", *VALID_KEYS]
-            substk = lambda k: "maxPrefixLength" if k == 'maxLength' else ("comment" if k == "netname" else k)
-            l_prefix.append({substk(k):v for k, v in r.items() if k in vkeys})
-        output = json.dumps(d_output, indent=2)
-    else:
-        output += "# NeoNetwork ROA tool\n"
-        pattern = 'route %s max %d as %d;'
-        l_output = list()
-        rdict2list = lambda d: [d[k] for k in VALID_KEYS]
-        for (asn, prefix, maxlen) in [rdict2list(r) for r in (*roa4, *roa6)]:
-            l_output.append(pattern % (prefix, maxlen, asn))
-        output += '\n'.join(l_output)
-    if not args.output or args.output == '-':
+    if (
+        args.max < 0
+        or args.max6 < 0
+        or args.max > IPv4Network(0).max_prefixlen
+        or args.max6 > IPv6Network(0).max_prefixlen
+    ):
+        parser.error("check your max prefix length")
+    output = main(args)
+    if not args.output or args.output == "-":
         print(output)
-    else:
+    elif output:
         Path(args.output).write_text(output)
-        print('written to', args.output)
+        print("written to", args.output)