package seoser

import (
	"net/url"
	"os"
	"regexp"
	"strings"
)

// BuildRobotsTXT 动态 robots.txt：优先读取 templates/page/robots.txt，并把 Sitemap 写成绝对地址。
func BuildRobotsTXT(siteURL string) string {
	sitemap := AbsURL(siteURL, "/sitemap.xml")
	body := defaultRobotsTXT
	if raw, err := os.ReadFile("templates/page/robots.txt"); err == nil {
		if t := strings.TrimSpace(string(raw)); t != "" {
			body = t
		}
	}
	body = strings.ReplaceAll(body, "\r\n", "\n")
	body = sitemapLineRe.ReplaceAllString(body, "Sitemap: "+sitemap)
	if !strings.Contains(body, "Sitemap:") {
		body = strings.TrimRight(body, "\n") + "\n\nSitemap: " + sitemap + "\n"
	} else if !strings.HasSuffix(body, "\n") {
		body += "\n"
	}
	if host := hostFromSiteURL(siteURL); host != "" && !strings.Contains(body, "Host:") {
		body = strings.TrimRight(body, "\n") + "\nHost: " + host + "\n"
	}
	return body
}

func hostFromSiteURL(siteURL string) string {
	u, err := url.Parse(strings.TrimSpace(siteURL))
	if err != nil || u.Host == "" {
		return ""
	}
	return u.Host
}

var sitemapLineRe = regexp.MustCompile(`(?m)^Sitemap:\s*\S+`)

const defaultRobotsTXT = `User-agent: *
Allow: /

# 非 HTML 接口与工具路由不抓取
Disallow: /api/
Disallow: /ui/
Disallow: /healthz
Disallow: /h5/
Disallow: /enkey
Disallow: /newenkey

# 登录、搜索和账户 HTML 页面允许抓取，以便爬虫读取页面内 noindex。

Sitemap: /sitemap.xml
`