navi/internal/subscription/fetch.go

package subscription

import (
	"context"
	"encoding/base64"
	"fmt"
	"io"
	"net/http"
	"net/url"
	"regexp"
	"strings"
	"time"
	"unicode"

	"vpnclient/internal/config"
	"vpnclient/internal/linknorm"
	"vpnclient/internal/protocols/hysteria2"
	"vpnclient/internal/protocols/naive"
)

// Item is one imported share link.
type Item struct {
	Name     string
	Protocol config.Protocol
	URI      string
}

var (
	reShareLine = regexp.MustCompile(`(?i)((?:hysteria2|hy2|vless|vmess|trojan|awg|amneziawg|wireguard|naive\+https|naive\+quic|naive|quic|https|http)://[^\s"'<>]+)`)
)

// Fetch downloads a subscription body and parses proxy share links.
func Fetch(ctx context.Context, rawURL string) ([]Item, error) {
	rawURL = strings.TrimSpace(rawURL)
	if rawURL == "" {
		return nil, fmt.Errorf("пустой URL подписки")
	}
	u, err := url.Parse(rawURL)
	if err != nil || u.Scheme != "https" {
		return nil, fmt.Errorf("нужен https URL подписки")
	}

	req, err := http.NewRequestWithContext(ctx, http.MethodGet, rawURL, nil)
	if err != nil {
		return nil, err
	}
	req.Header.Set("User-Agent", "ClashMeta/1.18.0")
	req.Header.Set("Accept", "*/*")
	client := &http.Client{Timeout: 45 * time.Second}
	resp, err := client.Do(req)
	if err != nil {
		return nil, fmt.Errorf("не удалось скачать подписку: %w", err)
	}
	defer resp.Body.Close()
	if resp.StatusCode < 200 || resp.StatusCode >= 300 {
		return nil, fmt.Errorf("подписка HTTP %d", resp.StatusCode)
	}
	body, err := io.ReadAll(io.LimitReader(resp.Body, 8<<20))
	if err != nil {
		return nil, err
	}
	items, err := ParseBody(string(body))
	if err != nil {
		return nil, err
	}
	return items, nil
}

// ParseBody accepts plain text lines, base64, or Clash-like blobs with share links.
func ParseBody(body string) ([]Item, error) {
	body = strings.TrimSpace(body)
	if body == "" {
		return nil, fmt.Errorf("пустая подписка")
	}

	candidates := []string{body}
	if decoded, err := decodeMaybeBase64(body); err == nil && decoded != "" && decoded != body {
		candidates = append(candidates, decoded)
		// some panels double-encode
		if decoded2, err2 := decodeMaybeBase64(decoded); err2 == nil && decoded2 != "" && decoded2 != decoded {
			candidates = append(candidates, decoded2)
		}
	}

	var lastErr error
	for _, text := range candidates {
		items, err := parseText(text)
		if err == nil && len(items) > 0 {
			return items, nil
		}
		if err != nil {
			lastErr = err
		}
	}
	if lastErr != nil {
		return nil, lastErr
	}
	return nil, fmt.Errorf("в подписке нет поддерживаемых ссылок")
}

func parseText(text string) ([]Item, error) {
	lines := splitLines(text)
	// Also extract share URIs embedded in YAML/HTML/JSON.
	extracted := reShareLine.FindAllString(text, -1)
	for _, m := range extracted {
		lines = append(lines, strings.TrimRight(m, ".,;)]}\"'"))
	}
	for _, block := range extractClashProxies(text) {
		lines = append(lines, block)
	}

	out := make([]Item, 0, len(lines))
	seen := map[string]int{}
	seenURI := map[string]struct{}{}
	for _, line := range lines {
		line = strings.TrimSpace(line)
		if line == "" || strings.HasPrefix(line, "#") {
			continue
		}
		line = strings.Trim(line, `"'`)
		proto := config.DetectProtocol(line)
		if proto == "" {
			continue
		}
		// Skip bare https:// without credentials (common noise in HTML/YAML).
		// Auth must be in the URI authority (before '/'), not path like Qure@master.
		if proto == config.ProtocolNaive && !naiveLinkHasAuth(line) && !strings.HasPrefix(strings.ToLower(line), "naive+") {
			continue
		}
		normalized, detected, remark, err := linknorm.Normalize(proto, line)
		if err != nil {
			continue
		}
		if _, dup := seenURI[normalized]; dup {
			continue
		}
		seenURI[normalized] = struct{}{}
		name := remark
		if name == "" {
			name = defaultName(detected, normalized)
		}
		base := name
		for n := 2; ; n++ {
			if _, ok := seen[name]; !ok {
				break
			}
			name = fmt.Sprintf("%s-%d", base, n)
		}
		seen[name] = 1
		out = append(out, Item{Name: name, Protocol: detected, URI: normalized})
	}
	if len(out) == 0 {
		return nil, fmt.Errorf("в подписке нет поддерживаемых ссылок")
	}
	return out, nil
}

func naiveLinkHasAuth(line string) bool {
	lower := strings.ToLower(strings.TrimSpace(line))
	for _, p := range []string{"naive+https://", "naive+quic://", "naive://", "https://", "http://", "quic://"} {
		if strings.HasPrefix(lower, p) {
			rest := line[len(p):]
			authEnd := strings.IndexByte(rest, '/')
			authority := rest
			if authEnd >= 0 {
				authority = rest[:authEnd]
			}
			at := strings.IndexByte(authority, '@')
			return at > 0
		}
	}
	return false
}

var reClashProxyName = regexp.MustCompile(`(?m)^[ \t]*-\s+name\s*:\s*`)

// extractClashProxies converts Clash Meta YAML proxy blocks into share URIs
// (vless / trojan / hysteria2) when the subscription has no share-link lines.
func extractClashProxies(text string) []string {
	lower := strings.ToLower(text)
	if !strings.Contains(lower, "type:") {
		return nil
	}
	if !strings.Contains(lower, "vless") && !strings.Contains(lower, "hysteria2") && !strings.Contains(lower, "trojan") {
		return nil
	}
	scope := text
	if sec := yamlTopSection(text, "proxies"); sec != "" {
		scope = sec
	}
	var out []string
	for _, chunk := range splitClashProxyBlocks(scope) {
		if link := clashBlockToShareURI(chunk); link != "" {
			out = append(out, link)
		}
	}
	return out
}

func splitClashProxyBlocks(text string) []string {
	idxs := reClashProxyName.FindAllStringIndex(text, -1)
	if len(idxs) == 0 {
		// Legacy fallback used by older Clash dumps.
		parts := strings.Split(text, "\n- ")
		out := make([]string, 0, len(parts))
		for i, p := range parts {
			if i == 0 {
				continue
			}
			out = append(out, "name: "+p)
		}
		return out
	}
	out := make([]string, 0, len(idxs))
	for i, idx := range idxs {
		start := idx[0]
		end := len(text)
		if i+1 < len(idxs) {
			end = idxs[i+1][0]
		}
		out = append(out, text[start:end])
	}
	return out
}

func clashBlockToShareURI(block string) string {
	switch strings.ToLower(yamlField(block, "type")) {
	case "hysteria2", "hy2":
		return clashHY2URI(block)
	case "vless":
		return clashVLESSURI(block)
	case "trojan":
		return clashTrojanURI(block)
	default:
		return ""
	}
}

func clashHY2URI(block string) string {
	server := yamlField(block, "server")
	port := yamlField(block, "port")
	pass := yamlField(block, "password")
	if pass == "" {
		pass = yamlField(block, "auth")
	}
	if server == "" || port == "" || pass == "" {
		return ""
	}
	u := url.URL{Scheme: "hysteria2", User: url.User(pass), Host: server + ":" + port}
	q := url.Values{}
	if sni := yamlField(block, "sni"); sni != "" {
		q.Set("sni", sni)
	}
	if obfs := yamlField(block, "obfs"); obfs != "" {
		q.Set("obfs", obfs)
	}
	obfsPass := yamlField(block, "obfs-password")
	if obfsPass == "" {
		obfsPass = yamlField(block, "obfs_password")
	}
	if obfsPass != "" {
		q.Set("obfs-password", obfsPass)
	}
	u.RawQuery = q.Encode()
	return withClashName(u.String(), yamlField(block, "name"))
}

func clashVLESSURI(block string) string {
	server := yamlField(block, "server")
	port := yamlField(block, "port")
	uuid := yamlField(block, "uuid")
	if server == "" || port == "" || uuid == "" {
		return ""
	}
	network := strings.ToLower(yamlField(block, "network"))
	if network == "" {
		network = "tcp"
	}
	if network == "websocket" {
		network = "ws"
	}
	reality := yamlSection(block, "reality-opts")
	hasReality := reality != "" || strings.Contains(strings.ToLower(block), "reality-opts")
	tlsFlag := strings.ToLower(yamlField(block, "tls"))
	security := "none"
	switch {
	case hasReality:
		security = "reality"
	case tlsFlag == "true" || tlsFlag == "1":
		security = "tls"
	case tlsFlag == "false" || tlsFlag == "0":
		security = "none"
	default:
		if s := strings.ToLower(yamlField(block, "security")); s != "" {
			security = s
		}
	}
	q := url.Values{}
	q.Set("type", network)
	q.Set("security", security)
	if sni := firstNonEmpty(yamlField(block, "servername"), yamlField(block, "sni")); sni != "" {
		q.Set("sni", sni)
	}
	if fp := yamlField(block, "client-fingerprint"); fp != "" {
		q.Set("fp", fp)
	}
	if flow := yamlField(block, "flow"); flow != "" {
		q.Set("flow", flow)
	}
	if alpn := yamlStringOrList(block, "alpn"); alpn != "" {
		q.Set("alpn", alpn)
	}
	if pe := yamlField(block, "packet-encoding"); pe != "" {
		q.Set("packetEncoding", pe)
	}
	switch network {
	case "ws":
		ws := yamlSection(block, "ws-opts")
		src := ws
		if src == "" {
			src = block
		}
		if path := yamlField(src, "path"); path != "" {
			q.Set("path", path)
		}
		if host := clashWSHost(ws); host != "" {
			q.Set("host", host)
		}
	case "grpc", "gun":
		grpc := yamlSection(block, "grpc-opts")
		if sn := firstNonEmpty(yamlField(grpc, "grpc-service-name"), yamlField(block, "servername")); sn != "" {
			q.Set("serviceName", sn)
		}
	case "xhttp", "splithttp":
		xh := yamlSection(block, "xhttp-opts")
		src := xh
		if src == "" {
			src = block
		}
		if path := yamlField(src, "path"); path != "" {
			q.Set("path", path)
		}
		if host := yamlField(xh, "host"); host != "" {
			q.Set("host", host)
		}
		if mode := yamlField(xh, "mode"); mode != "" {
			q.Set("mode", mode)
		}
	}
	if security == "reality" {
		if pbk := firstNonEmpty(yamlField(reality, "public-key"), yamlField(block, "public-key")); pbk != "" {
			q.Set("pbk", pbk)
		}
		sid := firstNonEmpty(yamlField(reality, "short-id"), yamlField(block, "short-id"))
		q.Set("sid", sid)
		if spx := yamlField(reality, "spider-x"); spx != "" {
			q.Set("spx", spx)
		}
	}
	if skip := strings.ToLower(yamlField(block, "skip-cert-verify")); skip == "true" || skip == "1" {
		q.Set("allowInsecure", "1")
	}
	u := url.URL{Scheme: "vless", User: url.User(uuid), Host: server + ":" + port, RawQuery: q.Encode()}
	return withClashName(u.String(), yamlField(block, "name"))
}

func clashTrojanURI(block string) string {
	server := yamlField(block, "server")
	port := yamlField(block, "port")
	pass := yamlField(block, "password")
	if server == "" || port == "" || pass == "" {
		return ""
	}
	network := strings.ToLower(yamlField(block, "network"))
	if network == "" {
		network = "tcp"
	}
	if network == "websocket" {
		network = "ws"
	}
	security := "tls"
	if tlsFlag := strings.ToLower(yamlField(block, "tls")); tlsFlag == "false" || tlsFlag == "0" {
		security = "none"
	}
	q := url.Values{}
	q.Set("type", network)
	q.Set("security", security)
	if sni := firstNonEmpty(yamlField(block, "sni"), yamlField(block, "servername")); sni != "" {
		q.Set("sni", sni)
	}
	if fp := yamlField(block, "client-fingerprint"); fp != "" {
		q.Set("fp", fp)
	}
	if alpn := yamlStringOrList(block, "alpn"); alpn != "" {
		q.Set("alpn", alpn)
	}
	if network == "ws" {
		ws := yamlSection(block, "ws-opts")
		src := ws
		if src == "" {
			src = block
		}
		if path := yamlField(src, "path"); path != "" {
			q.Set("path", path)
		}
		if host := clashWSHost(ws); host != "" {
			q.Set("host", host)
		}
	}
	if skip := strings.ToLower(yamlField(block, "skip-cert-verify")); skip == "true" || skip == "1" {
		q.Set("allowInsecure", "1")
	}
	u := url.URL{Scheme: "trojan", User: url.User(pass), Host: server + ":" + port, RawQuery: q.Encode()}
	return withClashName(u.String(), yamlField(block, "name"))
}

func clashWSHost(wsOpts string) string {
	headers := yamlSection(wsOpts, "headers")
	return firstNonEmpty(yamlField(headers, "Host"), yamlField(headers, "host"), yamlField(wsOpts, "host"))
}

func withClashName(link, name string) string {
	if name == "" {
		return link
	}
	return link + "#" + url.PathEscape(name)
}

func yamlTopSection(text, key string) string {
	re := regexp.MustCompile(`(?im)^` + regexp.QuoteMeta(key) + `\s*:\s*(?:#.*)?$`)
	loc := re.FindStringIndex(text)
	if loc == nil {
		return ""
	}
	rest := text[loc[1]:]
	end := regexp.MustCompile(`(?m)^[A-Za-z0-9_-]+\s*:`).FindStringIndex(rest)
	if end == nil {
		return rest
	}
	return rest[:end[0]]
}

func yamlSection(block, key string) string {
	lines := strings.Split(block, "\n")
	keyLower := strings.ToLower(key) + ":"
	capturing := false
	parentIndent := 0
	var out []string
	for _, line := range lines {
		indent := 0
		for _, r := range line {
			if r == ' ' || r == '\t' {
				indent++
			} else {
				break
			}
		}
		trimmed := strings.TrimSpace(line)
		if !capturing {
			if strings.HasPrefix(strings.ToLower(trimmed), keyLower) {
				capturing = true
				parentIndent = indent
			}
			continue
		}
		if trimmed != "" && indent <= parentIndent {
			break
		}
		out = append(out, line)
	}
	return strings.Join(out, "\n")
}

func yamlField(block, key string) string {
	// Allow list marker: "  - name: …"
	// Use [ \t]* after ':' so we do NOT consume newlines (or else "alpn:\n - h2" becomes "- h2").
	re := regexp.MustCompile(`(?im)^\s*(?:-\s+)?` + regexp.QuoteMeta(key) + `:[ \t]*["']?([^"'#\n\r]+)["']?`)
	m := re.FindStringSubmatch(block)
	if len(m) < 2 {
		return ""
	}
	return strings.TrimSpace(m[1])
}

func yamlStringOrList(block, key string) string {
	single := strings.TrimSpace(yamlField(block, key))
	// Scalar must not look like a YAML list item pulled from the next line.
	if single != "" && single != "|" && single != ">" && !strings.HasPrefix(single, "-") {
		if strings.HasPrefix(single, "[") {
			inner := strings.Trim(single, "[]")
			parts := strings.Split(inner, ",")
			out := make([]string, 0, len(parts))
			for _, p := range parts {
				p = strings.Trim(strings.TrimSpace(p), `"'`)
				if p != "" {
					out = append(out, p)
				}
			}
			return strings.Join(out, ",")
		}
		return single
	}
	section := yamlSection(block, key)
	var items []string
	for _, line := range strings.Split(section, "\n") {
		t := strings.TrimSpace(line)
		if !strings.HasPrefix(t, "-") {
			continue
		}
		v := strings.TrimSpace(strings.TrimPrefix(t, "-"))
		v = strings.Trim(v, `"'`)
		if i := strings.IndexByte(v, '#'); i >= 0 {
			v = strings.TrimSpace(v[:i])
		}
		if v != "" {
			items = append(items, v)
		}
	}
	return strings.Join(items, ",")
}

func firstNonEmpty(vals ...string) string {
	for _, v := range vals {
		if strings.TrimSpace(v) != "" {
			return strings.TrimSpace(v)
		}
	}
	return ""
}

func decodeMaybeBase64(s string) (string, error) {
	s = strings.Map(func(r rune) rune {
		if unicode.IsSpace(r) {
			return -1
		}
		return r
	}, s)
	if s == "" {
		return "", fmt.Errorf("empty")
	}
	raw, err := base64.StdEncoding.DecodeString(s)
	if err != nil {
		raw, err = base64.RawStdEncoding.DecodeString(s)
	}
	if err != nil {
		raw, err = base64.URLEncoding.DecodeString(s)
	}
	if err != nil {
		raw, err = base64.RawURLEncoding.DecodeString(s)
	}
	if err != nil {
		return "", err
	}
	return string(raw), nil
}

func splitLines(s string) []string {
	s = strings.ReplaceAll(s, "\r\n", "\n")
	s = strings.ReplaceAll(s, "\r", "\n")
	return strings.Split(s, "\n")
}

func defaultName(proto config.Protocol, uri string) string {
	host := ""
	switch proto {
	case config.ProtocolHysteria2:
		if h, _, err := hysteria2.HostPort(uri); err == nil {
			host = h
		}
	default:
		if u, _, err := naive.ParseShareLink(uri); err == nil {
			if parsed, err := url.Parse(u); err == nil {
				host = parsed.Hostname()
			}
		}
	}
	if host == "" {
		host = string(proto)
	}
	return string(proto) + "-" + host
}