From ba1c7407eaa0c09e93d8f319c9e7e65bdf0187d3 Mon Sep 17 00:00:00 2001
From: Aleksandr Nogikh <nogikh@google.com>
Date: Tue, 17 Jan 2023 19:20:37 +0100
Subject: pkg/subsystem/linux: extract names for subsystems

Extract the short subsystem name from the mailing list email.

Stip the common prefixes and suffixes and make sure there are no
duplicates.

As a fallback, assign the whole list email address as a subsystem name.
---
 pkg/subsystem/linux/names.go | 103 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 103 insertions(+)
 create mode 100644 pkg/subsystem/linux/names.go

(limited to 'pkg/subsystem/linux/names.go')

diff --git a/pkg/subsystem/linux/names.go b/pkg/subsystem/linux/names.go
new file mode 100644
index 000000000..0f422ea05
--- /dev/null
+++ b/pkg/subsystem/linux/names.go
@@ -0,0 +1,103 @@
+// Copyright 2023 syzkaller project authors. All rights reserved.
+// Use of this source code is governed by Apache 2 LICENSE that can be found in the LICENSE file.
+
+package linux
+
+import (
+	"fmt"
+	"regexp"
+
+	"github.com/google/syzkaller/pkg/subsystem/entity"
+)
+
+// setSubsystemNames assigns unique names to the presented subsystems.
+// If it failed to assign a name to a subsystem, the Name field remains empty.
+func setSubsystemNames(list []*entity.Subsystem) error {
+	dups := map[string]bool{}
+	for _, item := range list {
+		// For now, we can only infer name from the list email.
+		if len(item.Lists) == 0 {
+			return fmt.Errorf("no lists for %#v", item)
+		}
+		email := item.Lists[0]
+		name := emailToName(email)
+		if !validateName(name) {
+			return fmt.Errorf("failed to extract a name from %s", email)
+		}
+		if dups[name] {
+			return fmt.Errorf("duplicate subsystem name: %s", item.Name)
+		}
+		item.Name = name
+		dups[name] = true
+	}
+	return nil
+}
+
+func validateName(name string) bool {
+	const (
+		minLen = 2
+		maxLen = 16 // otherwise the email subject can get too big
+	)
+	return len(name) >= minLen && len(name) <= maxLen
+}
+
+func emailToName(email string) string {
+	if name := emailExceptions[email]; name != "" {
+		return name
+	}
+	ret := emailStripRe.FindStringSubmatch(email)
+	if ret == nil {
+		return ""
+	}
+	return ret[1]
+}
+
+func buildEmailStripRe() *regexp.Regexp {
+	raw := `^(?:`
+	for i := 0; i < len(stripPrefixes); i++ {
+		if i > 0 {
+			raw += "|"
+		}
+		raw += regexp.QuoteMeta(stripPrefixes[i])
+	}
+	raw += ")*(.*?)(?:"
+	for i := 0; i < len(stripSuffixes); i++ {
+		if i > 0 {
+			raw += "|"
+		}
+		raw += regexp.QuoteMeta(stripSuffixes[i])
+	}
+	raw += ")*@.*$"
+	return regexp.MustCompile(raw)
+}
+
+var (
+	emailExceptions = map[string]string{
+		"patches@opensource.cirrus.com":             "cirrus",
+		"virtualization@lists.linux-foundation.org": "virt", // the name is too long
+		"dev@openvswitch.org":                       "openvswitch",
+		"devel@acpica.org":                          "acpica",
+		"kernel@dh-electronics.com":                 "dh-electr",
+		"devel@lists.orangefs.org":                  "orangefs",
+		"linux-arm-kernel@axis.com":                 "axis",
+		"Dell.Client.Kernel@dell.com":               "dell",
+		"sound-open-firmware@alsa-project.org":      "sof",
+		"platform-driver-x86@vger.kernel.org":       "x86-drivers",
+		"linux-trace-devel@vger.kernel.org":         "rt-tools",
+		"aws-nitro-enclaves-devel@amazon.com":       "nitro",
+		"brcm80211-dev-list.pdl@broadcom.com":       "brcm80211",
+		"osmocom-net-gprs@lists.osmocom.org":        "osmocom",
+		"netdev@vger.kernel.org":                    "net",
+		"megaraidlinux.pdl@broadcom.com":            "megaraid",
+		"mpi3mr-linuxdrv.pdl@broadcom.com":          "mpi3",
+		"MPT-FusionLinux.pdl@broadcom.com":          "mpt-fusion",
+	}
+	stripPrefixes = []string{"linux-"}
+	stripSuffixes = []string{
+		"-devel", "-dev", "-devs", "-developer", "devel",
+		"-user", "-users",
+		"-discussion", "-discuss", "-list", "-en",
+		"-kernel", "-linux", "-general",
+	}
+	emailStripRe = buildEmailStripRe()
+)
-- 
cgit mrf-deployment