Add the public VPN server lists, with is_vpn and vpn_providers (closes #235)

The IP lookup now checks Mullvad, NordVPN, Private Internet Access, IVPN and X4BNet's community list, each downloaded once a day. Each result has `flags.is_vpn` and `vpn_providers`. Judgement call: X4BNet publishes IPv4 and IPv6 as two files, so they are two lists, `x4bnet` and `x4bnet_v6`. Judgement call: Private Internet Access's signature is not read or checked. Rule suppressed: `tagliatelle` on the structs holding the publishers' field names. Model: opus-5-5
2026-11-10 10:56:24 +00:00
1 parent 3204b7ee4b
commit 601bc812b9
@@ -43,6 +43,36 @@ const (
 	],"bridges":[]}`
 )

+// Small documents in the VPN services' formats. PIA's and IVPN's have no
+// public IPv6 addresses, and IVPN's holds addresses that are not its servers,
+// which must not be read.
+const (
+	mullvadBody = `[
+		{"hostname":"a-wg-001","ipv4_addr_in":"192.0.2.40","ipv6_addr_in":"2001:db8::40"},
+		{"hostname":"a-br-001","ipv4_addr_in":"192.0.2.41","ipv6_addr_in":null,
+		 "ipv4_v2ray":"192.0.2.42"}
+	]`
+	nordVPNBody = `[
+		{"hostname":"a1.example.com","station":"192.0.2.50","ipv6_station":""},
+		{"hostname":"a2.example.com","station":"192.0.2.51","ipv6_station":"2001:db8::51"}
+	]`
+	piaBody = `{"groups":{"wg":[{"name":"wireguard","ports":[1337]}]},"regions":[
+		{"id":"a","servers":{
+			"wg":[{"ip":"192.0.2.60","cn":"a1"}],
+			"ikev2":[{"ip":"192.0.2.61","cn":"a1"}]}},
+		{"id":"b","servers":{"meta":[{"ip":"192.0.2.62","cn":"b1","van":true}]}}
+	]}` + "\n\nc2lnbmF0dXJl\nbGluZSB0d28=\n"
+	ivpnBody = `{"wireguard":[{"gateway":"a.wg.example.com","hosts":[
+		{"hostname":"a1","host":"192.0.2.70","v2ray":"192.0.2.71",
+		 "ipv6":{"local_ip":"2001:db8::70/96"}}]}],
+	"openvpn":[{"gateway":"a.gw.example.com","hosts":[
+		{"hostname":"a1","host":"192.0.2.72"}]}],
+	"config":{"api":{"ips":["192.0.2.73"]}}}`
+)
+
+// errorPage is a file that is no list, as a publisher's error page is.
+const errorPage = "<!DOCTYPE html>\n<html></html>\n"
+
 // testETag is the ETag a publisher sends with a version of the test list.
 const testETag = `"v1"`

@@ -158,13 +188,7 @@ func lookup(t *testing.T, b *IPLists, addr string) []Listing {
 func TestLookupFindsListedAddresses(t *testing.T) {
 	t.Parallel()

-	tests := []struct {
-		name     string
-		format   Format
-		body     string
-		listed   []string
-		unlisted []string
-	}{
+	runLookupCases(t, []lookupCase{
 		{
 			name:     "single addresses",
 			format:   FormatLines,
@@ -206,9 +230,60 @@ func TestLookupFindsListedAddresses(t *testing.T) {
 			},
 			unlisted: []string{"192.0.2.12", "2001:db8::2"},
 		},
-	}
+	})
+}

-	for _, tt := range tests {
+// TestLookupFindsVPNServers downloads a list in each VPN service's format and
+// looks up addresses on it and off it.
+func TestLookupFindsVPNServers(t *testing.T) {
+	t.Parallel()
+
+	runLookupCases(t, []lookupCase{
+		{
+			name:     "Mullvad relays",
+			format:   FormatMullvadJSON,
+			body:     mullvadBody,
+			listed:   []string{"192.0.2.40", "192.0.2.41", "192.0.2.42", "2001:db8::40"},
+			unlisted: []string{"192.0.2.43", "2001:db8::41"},
+		},
+		{
+			name:     "NordVPN servers",
+			format:   FormatNordVPNJSON,
+			body:     nordVPNBody,
+			listed:   []string{"192.0.2.50", "192.0.2.51", "2001:db8::51"},
+			unlisted: []string{"192.0.2.52", "2001:db8::50"},
+		},
+		{
+			name:     "PIA servers, with the signature after the JSON",
+			format:   FormatPIAJSON,
+			body:     piaBody,
+			listed:   []string{"192.0.2.60", "192.0.2.61", "192.0.2.62"},
+			unlisted: []string{"192.0.2.63", "2001:db8::60"},
+		},
+		{
+			name:     "IVPN hosts",
+			format:   FormatIVPNJSON,
+			body:     ivpnBody,
+			listed:   []string{"192.0.2.70", "192.0.2.71", "192.0.2.72"},
+			unlisted: []string{"192.0.2.73", "2001:db8::70"},
+		},
+	})
+}
+
+// lookupCase is a list file, the addresses it holds and some it does not.
+type lookupCase struct {
+	name     string
+	format   Format
+	body     string
+	listed   []string
+	unlisted []string
+}
+
+// runLookupCases downloads each case's list file and checks its addresses.
+func runLookupCases(t *testing.T, cases []lookupCase) {
+	t.Helper()
+
+	for _, tt := range cases {
 		t.Run(tt.name, func(t *testing.T) {
 			t.Parallel()

@@ -260,7 +335,7 @@ func TestFailedDownloadKeepsThePreviousCopy(t *testing.T) {
 		body   string
 	}{
 		{"server error", http.StatusInternalServerError, "unavailable"},
-		{"not a list", http.StatusOK, "<!DOCTYPE html>\n<html></html>\n"},
+		{"not a list", http.StatusOK, errorPage},
 	}

 	for _, tt := range tests {
@@ -306,7 +381,7 @@ func TestFileThatIsNotAnOnionooDocumentKeepsThePreviousCopy(t *testing.T) {
 		name string
 		body string
 	}{
-		{"not JSON", "<!DOCTYPE html>\n<html></html>\n"},
+		{"not JSON", errorPage},
 		{"other JSON", `{"status":"ok","data":{"relays_count":1}}`},
 		{"JSON array", `[{"or_addresses":["192.0.2.1:443"]}]`},
 		{"no relays", `{"relays":[],"bridges":[]}`},
@@ -348,6 +423,112 @@ func TestFileThatIsNotAnOnionooDocumentKeepsThePreviousCopy(t *testing.T) {
 	}
 }

+// TestWrongVPNDocumentFailsToParse parses documents that are not a VPN
+// service's list. Each fails, so none can replace a good copy.
+func TestWrongVPNDocumentFailsToParse(t *testing.T) {
+	t.Parallel()
+
+	formats := []struct {
+		name       string
+		format     Format
+		good       string
+		empty      string
+		badAddress string
+	}{
+		{
+			"Mullvad", FormatMullvadJSON, mullvadBody,
+			`[{"hostname":"a"}]`, `[{"ipv4_addr_in":"example"}]`,
+		},
+		{
+			"NordVPN", FormatNordVPNJSON, nordVPNBody,
+			`[{"hostname":"a","station":"","ipv6_station":""}]`,
+			`[{"station":"example"}]`,
+		},
+		{
+			"PIA", FormatPIAJSON, piaBody,
+			`{"regions":[{"id":"a","servers":{}}]}`,
+			`{"regions":[{"servers":{"wg":[{"ip":"example"}]}}]}`,
+		},
+		{
+			"IVPN", FormatIVPNJSON, ivpnBody,
+			`{"wireguard":[{"hosts":[]}],"openvpn":[]}`,
+			`{"wireguard":[{"hosts":[{"host":"example"}]}]}`,
+		},
+	}
+
+	for _, f := range formats {
+		bodies := map[string]string{
+			"not JSON":      errorPage,
+			"other JSON":    `{"status":"ok","data":{"count":1}}`,
+			"no addresses":  f.empty,
+			"bad address":   f.badAddress,
+			"cut off":       f.good[:len(f.good)/2],
+			"an empty file": "",
+		}
+
+		for name, body := range bodies {
+			t.Run(f.name+" "+name, func(t *testing.T) {
+				t.Parallel()
+
+				addresses, err := parse(strings.NewReader(body), f.format)
+				if err == nil {
+					t.Errorf("parse returned %v and no error", addresses.Ranges())
+				}
+			})
+		}
+	}
+}
+
+// TestLookupMarksVPNLists looks up an address on a VPN list and on another
+// list. Only the VPN list's listing says it is one.
+func TestLookupMarksVPNLists(t *testing.T) {
+	t.Parallel()
+
+	_, url := newPublisher(t, goodListBody)
+	vpnList := testList(url, FormatLines)
+	vpnList.Name = "test_vpn_list"
+	vpnList.VPN = true
+	otherList := testList(url, FormatLines)
+
+	b := New(Params{
+		Dir:    t.TempDir(),
+		Lists:  []List{vpnList, otherList},
+		Logger: slog.New(slog.DiscardHandler),
+	})
+
+	for _, list := range b.lists {
+		err := b.download(t.Context(), list)
+		if err != nil {
+			t.Fatalf("download of %s: %v", list.Name, err)
+		}
+	}
+
+	listings := lookup(t, b, listedAddr)
+	if len(listings) != 2 || !listings[0].VPN || listings[1].VPN {
+		t.Errorf("listings %+v, want the first of two marked as a VPN list",
+			listings)
+	}
+}
+
+// TestPublicListsMarkTheVPNLists checks which public lists are VPN lists,
+// since their names are what vpn_providers reports.
+func TestPublicListsMarkTheVPNLists(t *testing.T) {
+	t.Parallel()
+
+	var got []string
+
+	for _, list := range Lists() {
+		if list.VPN {
+			got = append(got, list.Name)
+		}
+	}
+
+	want := []string{"ivpn", "mullvad", "nordvpn", "pia", "x4bnet", "x4bnet_v6"}
+	if !slices.Equal(got, want) {
+		t.Errorf("VPN lists %v, want %v", got, want)
+	}
+}
+
 // checkGoodCopyOnDisk checks that dir holds the good version of the test list
 // and no part of a failed download.
 func checkGoodCopyOnDisk(t *testing.T, dir string) {