// SPDX-License-Identifier: GPL-3.0-or-later

package events

import (
	"encoding/csv"
	"fmt"
	"os"
	"regexp"
	"sort"
	"strconv"
	"strings"

	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/model"
)

// ExecRule is the Go representation of Python's declarative exec rule. Every
// specified positive condition is ANDed; values inside a condition are ORed.
type ExecRule struct {
	SID           int
	Message       string
	Severity      string
	Classtype     string
	PathPrefixes  []string
	ExecComms     map[string]bool
	ParentComms   map[string]bool
	ArgvRegex     string
	ParentExclude map[string]bool
	compiled      *regexp.Regexp
}

func NewExecRule(rule ExecRule) (ExecRule, error) {
	if len(rule.PathPrefixes) == 0 && len(rule.ExecComms) == 0 &&
		len(rule.ParentComms) == 0 && rule.ArgvRegex == "" {
		return ExecRule{}, fmt.Errorf("rule sid=%d has no match conditions", rule.SID)
	}
	if rule.Severity == "" {
		rule.Severity = "HIGH"
	}
	if rule.Classtype == "" {
		rule.Classtype = "uncategorized"
	}
	if rule.ArgvRegex != "" {
		compiled, err := regexp.Compile("(?i)" + rule.ArgvRegex)
		if err != nil {
			return ExecRule{}, fmt.Errorf("rule sid=%d argv_regex: %w", rule.SID, err)
		}
		rule.compiled = compiled
	}
	return rule, nil
}

func (r ExecRule) Matches(event ExecEvent) bool {
	if r.ParentExclude[event.ParentComm] {
		return false
	}
	if len(r.PathPrefixes) > 0 && !hasPrefix(event.Filename, r.PathPrefixes) {
		return false
	}
	if len(r.ExecComms) > 0 && !r.ExecComms[event.ExecComm()] {
		return false
	}
	if len(r.ParentComms) > 0 && !r.ParentComms[event.ParentComm] {
		return false
	}
	if r.compiled != nil && !r.compiled.MatchString(event.ArgvString()) {
		return false
	}
	return true
}

func (r ExecRule) Alert(event ExecEvent) model.Alert {
	argv := truncateRunes(event.ArgvString(), 120)
	return model.Alert{
		SID:       r.SID,
		Severity:  r.Severity,
		Signature: "exec_rule." + r.Classtype,
		Classtype: r.Classtype,
		Key:       fmt.Sprintf("exec:%d:%d", r.SID, event.PID),
		Detail: fmt.Sprintf(
			"sid=%d %s — pid=%d ppid=%d parent=%s exec=%s argv=[%s]",
			r.SID, r.Message, event.PID, event.PPID, event.ParentComm,
			event.Filename, argv),
		PIDs: []int{event.PID},
	}
}

type ExecRuleEngine struct {
	Rules []ExecRule
}

func DefaultExecRuleEngine() ExecRuleEngine {
	// Constructors below use only compile-time-valid rules. Panic is preferable
	// to silently dropping a shipped detection if a future edit breaks one.
	rules := make([]ExecRule, 0, len(defaultExecRuleSpecs))
	for _, specification := range defaultExecRuleSpecs {
		rule, err := NewExecRule(specification)
		if err != nil {
			panic(err)
		}
		rules = append(rules, rule)
	}
	return ExecRuleEngine{Rules: rules}
}

func LoadExecRuleEngine(path string) (ExecRuleEngine, error) {
	engine := DefaultExecRuleEngine()
	if path == "" {
		return engine, nil
	}
	if _, err := os.Stat(path); os.IsNotExist(err) {
		return engine, nil
	} else if err != nil {
		return ExecRuleEngine{}, err
	}
	rules, err := loadTOMLExecRules(path)
	if err != nil {
		return ExecRuleEngine{}, err
	}
	engine.Rules = append(engine.Rules, rules...)
	return engine, nil
}

func (e ExecRuleEngine) Match(event ExecEvent) []model.Alert {
	alerts := make([]model.Alert, 0)
	for _, rule := range e.Rules {
		if rule.Matches(event) {
			alerts = append(alerts, rule.Alert(event))
		}
	}
	return alerts
}

var interpreters = stringSet(strings.Fields(
	"sh bash dash zsh ksh ash python python2 python3 perl ruby php lua " +
		"nc ncat netcat socat"))

var webDBServers = stringSet(strings.Fields(
	"nginx apache apache2 httpd php-fpm php php7 php8 lighttpd caddy " +
		"node nodejs tomcat catalina mysqld mariadbd postgres redis-server"))

var defaultExecRuleSpecs = []ExecRule{
	{
		SID: 100001, Message: "Program executed from a world-writable directory",
		Severity: "CRITICAL", Classtype: "fileless-execution",
		PathPrefixes: []string{"/tmp/", "/dev/shm/", "/var/tmp/"},
	},
	{
		SID: 100002, Message: "Reverse-shell command pattern in execve arguments",
		Severity: "CRITICAL", Classtype: "c2-reverse-shell",
		ArgvRegex: `/dev/(tcp|udp)/|\b(ba|da|z)?sh\b[^|]*\s-i\b|\bn(c|cat|etcat)\b.*\s-e\b|\bsocat\b.*\bexec|python[0-9.]*\b.*(pty\.spawn|socket\.socket)|perl\b.*\bSocket\b`,
	},
	{
		SID:      100003,
		Message:  "Web/DB service spawned a shell or interpreter (possible RCE/webshell)",
		Severity: "CRITICAL", Classtype: "web-rce",
		ExecComms: interpreters, ParentComms: webDBServers,
	},
	{
		SID: 100004, Message: "Download piped directly to a shell (ingress tool transfer)",
		Severity: "HIGH", Classtype: "ingress-tool-transfer",
		ArgvRegex: `\b(curl|wget|fetch)\b.*\|\s*(ba|da|z)?sh\b`,
	},
}

type rawExecRule struct {
	values map[string]string
}

// loadTOMLExecRules implements the intentionally narrow [[exec_rules]] subset
// consumed by Python. The sidecar stays stdlib-only; multiline string arrays,
// quoted strings, comments, and arbitrary table ordering are supported.
func loadTOMLExecRules(path string) ([]ExecRule, error) {
	raw, err := os.ReadFile(path)
	if err != nil {
		return nil, err
	}
	tables, err := parseExecRuleTables(string(raw))
	if err != nil {
		return nil, err
	}
	rules := make([]ExecRule, 0, len(tables))
	for _, table := range tables {
		sidText, ok := table.values["sid"]
		if !ok {
			return nil, fmt.Errorf("exec_rules entry missing sid")
		}
		sid, err := strconv.Atoi(strings.TrimSpace(sidText))
		if err != nil {
			return nil, fmt.Errorf("exec_rules sid: %w", err)
		}
		message, err := optionalString(table.values, "msg", "")
		if err != nil {
			return nil, err
		}
		severity, err := optionalString(table.values, "severity", "HIGH")
		if err != nil {
			return nil, err
		}
		severity = strings.ToUpper(severity)
		if severity != "MEDIUM" && severity != "HIGH" && severity != "CRITICAL" {
			severity = "HIGH"
		}
		classtype, err := optionalString(table.values, "classtype", "uncategorized")
		if err != nil {
			return nil, err
		}
		argvRegex, err := optionalString(table.values, "argv_regex", "")
		if err != nil {
			return nil, err
		}
		pathPrefixes, err := optionalStringArray(table.values, "path_prefixes")
		if err != nil {
			return nil, err
		}
		execComms, err := optionalStringArray(table.values, "exec_comm")
		if err != nil {
			return nil, err
		}
		parentComms, err := optionalStringArray(table.values, "parent_comm")
		if err != nil {
			return nil, err
		}
		parentExclude, err := optionalStringArray(table.values, "parent_exclude")
		if err != nil {
			return nil, err
		}
		rule, err := NewExecRule(ExecRule{
			SID: sid, Message: message, Severity: severity, Classtype: classtype,
			PathPrefixes: pathPrefixes, ExecComms: stringSet(execComms),
			ParentComms: stringSet(parentComms), ArgvRegex: argvRegex,
			ParentExclude: stringSet(parentExclude),
		})
		if err != nil {
			return nil, err
		}
		rules = append(rules, rule)
	}
	return rules, nil
}

func parseExecRuleTables(text string) ([]rawExecRule, error) {
	tables := make([]rawExecRule, 0)
	var current *rawExecRule
	var assignment strings.Builder
	depth := 0
	flush := func() error {
		if assignment.Len() == 0 || current == nil {
			assignment.Reset()
			return nil
		}
		key, value, ok := strings.Cut(assignment.String(), "=")
		if !ok {
			return fmt.Errorf("invalid exec rule assignment: %q", assignment.String())
		}
		current.values[strings.TrimSpace(key)] = strings.TrimSpace(value)
		assignment.Reset()
		return nil
	}
	for _, rawLine := range strings.Split(text, "\n") {
		line := strings.TrimSpace(stripTOMLComment(rawLine))
		if line == "" {
			continue
		}
		if line == "[[exec_rules]]" {
			if err := flush(); err != nil {
				return nil, err
			}
			tables = append(tables, rawExecRule{values: map[string]string{}})
			current = &tables[len(tables)-1]
			depth = 0
			continue
		}
		if current == nil {
			continue
		}
		if assignment.Len() > 0 {
			assignment.WriteByte(' ')
		}
		assignment.WriteString(line)
		depth += strings.Count(line, "[") - strings.Count(line, "]")
		if depth <= 0 {
			if err := flush(); err != nil {
				return nil, err
			}
			depth = 0
		}
	}
	if err := flush(); err != nil {
		return nil, err
	}
	return tables, nil
}

func optionalString(values map[string]string, key, fallback string) (string, error) {
	raw, ok := values[key]
	if !ok {
		return fallback, nil
	}
	value, err := tomlString(raw)
	if err != nil {
		return "", fmt.Errorf("exec_rules %s: %w", key, err)
	}
	return value, nil
}

func optionalStringArray(values map[string]string, key string) ([]string, error) {
	raw, ok := values[key]
	if !ok {
		return []string{}, nil
	}
	raw = strings.TrimSpace(raw)
	if len(raw) < 2 || raw[0] != '[' || raw[len(raw)-1] != ']' {
		return nil, fmt.Errorf("exec_rules %s: expected string array", key)
	}
	body := strings.TrimSpace(raw[1 : len(raw)-1])
	if body == "" {
		return []string{}, nil
	}
	body = strings.TrimSpace(strings.TrimSuffix(strings.TrimSpace(body), ","))
	reader := csv.NewReader(strings.NewReader(body))
	reader.TrimLeadingSpace = true
	fields, err := reader.Read()
	if err != nil {
		return nil, fmt.Errorf("exec_rules %s: %w", key, err)
	}
	result := make([]string, 0, len(fields))
	for _, field := range fields {
		field = strings.TrimSpace(field)
		if field == "" {
			continue
		}
		// encoding/csv already removes TOML's ordinary double quotes. Literal
		// single-quoted strings are not CSV syntax, so unwrap those explicitly.
		if len(field) >= 2 && field[0] == '\'' && field[len(field)-1] == '\'' {
			field = field[1 : len(field)-1]
		}
		result = append(result, field)
	}
	return result, nil
}

func tomlString(raw string) (string, error) {
	raw = strings.TrimSpace(raw)
	if len(raw) >= 2 && raw[0] == '\'' && raw[len(raw)-1] == '\'' {
		return raw[1 : len(raw)-1], nil
	}
	value, err := strconv.Unquote(raw)
	if err != nil {
		return "", fmt.Errorf("expected string, got %q", raw)
	}
	return value, nil
}

func stripTOMLComment(line string) string {
	inBasic, inLiteral, escaped := false, false, false
	for index, char := range line {
		if escaped {
			escaped = false
			continue
		}
		if char == '\\' && inBasic {
			escaped = true
			continue
		}
		if char == '"' && !inLiteral {
			inBasic = !inBasic
			continue
		}
		if char == '\'' && !inBasic {
			inLiteral = !inLiteral
			continue
		}
		if char == '#' && !inBasic && !inLiteral {
			return line[:index]
		}
	}
	return line
}

func hasPrefix(value string, prefixes []string) bool {
	for _, prefix := range prefixes {
		if strings.HasPrefix(value, prefix) {
			return true
		}
	}
	return false
}

func stringSet(values []string) map[string]bool {
	result := make(map[string]bool, len(values))
	for _, value := range values {
		result[value] = true
	}
	return result
}

func truncateRunes(value string, limit int) string {
	runes := []rune(value)
	if len(runes) <= limit {
		return value
	}
	return string(runes[:limit])
}

// RuleSIDs is kept sorted for operator catalogs and coverage assertions.
func (e ExecRuleEngine) RuleSIDs() []int {
	result := make([]int, 0, len(e.Rules))
	for _, rule := range e.Rules {
		result = append(result, rule.SID)
	}
	sort.Ints(result)
	return result
}