OpenIsle/backend/src/main/java/com/openisle/controller/RssController.java

package com.openisle.controller;

import com.openisle.model.Post;
import com.openisle.model.Comment;
import com.openisle.model.CommentSort;
import com.openisle.service.PostService;
import com.openisle.service.CommentService;
import lombok.RequiredArgsConstructor;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.safety.Safelist;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;

import com.vladsch.flexmark.ext.autolink.AutolinkExtension;
import com.vladsch.flexmark.ext.tables.TablesExtension;
import com.vladsch.flexmark.ext.gfm.strikethrough.StrikethroughExtension;
import com.vladsch.flexmark.ext.gfm.tasklist.TaskListExtension;
import com.vladsch.flexmark.html.HtmlRenderer;
import com.vladsch.flexmark.parser.Parser;
import com.vladsch.flexmark.util.data.MutableDataSet;

import java.net.URI;
import java.time.ZoneId;
import java.time.ZonedDateTime;
import java.time.format.DateTimeFormatter;
import java.util.*;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

@RestController
@RequiredArgsConstructor
public class RssController {
    private final PostService postService;
    private final CommentService commentService;

    @Value("${app.website-url:https://www.open-isle.com}")
    private String websiteUrl;

    // 兼容 Markdown/HTML 两类图片写法（用于 enclosure）
    private static final Pattern MD_IMAGE = Pattern.compile("!\\[[^\\]]*\\]\\(([^)]+)\\)");
    private static final Pattern HTML_IMAGE = Pattern.compile("<BaseImage[^>]+src=[\"']?([^\"'>]+)[\"']?[^>]*>");

    private static final DateTimeFormatter RFC1123 = DateTimeFormatter.RFC_1123_DATE_TIME;

    // flexmark：Markdown -> HTML
    private static final Parser MD_PARSER;
    private static final HtmlRenderer MD_RENDERER;
    static {
        MutableDataSet opts = new MutableDataSet();
        opts.set(Parser.EXTENSIONS, Arrays.asList(
                TablesExtension.create(),
                AutolinkExtension.create(),
                StrikethroughExtension.create(),
                TaskListExtension.create()
        ));
        // 允许内联 HTML（下游再做 sanitize）
        opts.set(Parser.HTML_BLOCK_PARSER, true);
        MD_PARSER = Parser.builder(opts).build();
        MD_RENDERER = HtmlRenderer.builder(opts).escapeHtml(false).build();
    }

    @GetMapping(value = "/api/rss", produces = "application/rss+xml;charset=UTF-8")
    public String feed() {
        // 建议 20；你现在是 10，这里保留你的 10
        List<Post> posts = postService.listLatestRssPosts(10);
        String base = trimTrailingSlash(websiteUrl);

        StringBuilder sb = new StringBuilder(4096);
        sb.append("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
        sb.append("<rss version=\"2.0\" xmlns:content=\"http://purl.org/rss/1.0/modules/content/\">");
        sb.append("<channel>");
        elem(sb, "title", cdata("OpenIsle RSS"));
        elem(sb, "link", base + "/");
        elem(sb, "description", cdata("Latest posts"));
        ZonedDateTime updated = posts.stream()
                .map(p -> p.getCreatedAt().atZone(ZoneId.systemDefault()))
                .max(Comparator.naturalOrder())
                .orElse(ZonedDateTime.now());
        // channel lastBuildDate（GMT）
        elem(sb, "lastBuildDate", toRfc1123Gmt(updated));

        for (Post p : posts) {
            String link = base + "/posts/" + p.getId();

            // 1) Markdown -> HTML
            String html = renderMarkdown(p.getContent());

            // 2) Sanitize（白名单增强）
            String safeHtml = sanitizeHtml(html);

            // 3) 绝对化 href/src + 强制 rel/target
            String absHtml = absolutifyHtml(safeHtml, base);

            // 4) 纯文本摘要（用于 <description>）
            String plain = textSummary(absHtml, 180);

            // 5) enclosure（首图，已绝对化）
            String enclosure = firstImage(p.getContent());
            if (enclosure == null) {
                // 如果 Markdown 没有图，尝试从渲染后的 HTML 再抓一次
                enclosure = firstImage(absHtml);
            }
            if (enclosure != null) {
                enclosure = absolutifyUrl(enclosure, base);
            }

            // 6) 构造优雅的附加区块（原文链接 + 精选评论），编入 <content:encoded>
            List<Comment> topComments = commentService
                    .getCommentsForPost(p.getId(), CommentSort.MOST_INTERACTIONS);
            topComments = topComments.subList(0, Math.min(10, topComments.size()));
            String footerHtml = buildFooterHtml(base, link, topComments);

            sb.append("<item>");
            elem(sb, "title", cdata(nullSafe(p.getTitle())));
            elem(sb, "link", link);
            sb.append("<guid isPermaLink=\"true\">").append(link).append("</guid>");
            elem(sb, "pubDate", toRfc1123Gmt(p.getCreatedAt().atZone(ZoneId.systemDefault())));
            // 摘要
            elem(sb, "description", cdata(plain));
            // 全文（HTML）：正文 + 优雅的 Markdown 区块（已转 HTML）
            sb.append("<content:encoded><![CDATA[")
                    .append(absHtml)
                    .append(footerHtml)
                    .append("]]></content:encoded>");
            // 首图 enclosure（图片类型）
            if (enclosure != null) {
                sb.append("<enclosure url=\"").append(escapeXml(enclosure)).append("\" type=\"")
                        .append(getMimeType(enclosure)).append("\" />");
            }
            sb.append("</item>");
        }

        sb.append("</channel></rss>");
        return sb.toString();
    }

    /* ===================== Markdown → HTML ===================== */

    private static String renderMarkdown(String md) {
        if (md == null || md.isEmpty()) return "";
        return MD_RENDERER.render(MD_PARSER.parse(md));
    }

    /* ===================== Sanitize & 绝对化 ===================== */

    private static String sanitizeHtml(String html) {
        if (html == null) return "";
        Safelist wl = Safelist.relaxed()
                .addTags(
                        "pre","code","figure","figcaption","picture","source",
                        "table","thead","tbody","tr","th","td",
                        "h1","h2","h3","h4","h5","h6",
                        "hr","blockquote"
                )
                .addAttributes("a", "href", "title", "target", "rel")
                .addAttributes("img", "src", "alt", "title", "width", "height")
                .addAttributes("source", "srcset", "type", "media")
                .addAttributes("code", "class")
                .addAttributes("pre", "class")
                .addProtocols("a", "href", "http", "https", "mailto")
                .addProtocols("img", "src", "http", "https", "data")
                .addProtocols("source", "srcset", "http", "https");
        // 清除所有 on* 事件、style（避免阅读器环境差异）
        return Jsoup.clean(html, wl);
    }

    private static String absolutifyHtml(String html, String baseUrl) {
        if (html == null || html.isEmpty()) return "";
        Document doc = Jsoup.parseBodyFragment(html, baseUrl);
        // a[href]
        for (Element a : doc.select("a[href]")) {
            String href = a.attr("href");
            String abs = absolutifyUrl(href, baseUrl);
            a.attr("href", abs);
            // 强制外链安全属性
            a.attr("rel", "noopener noreferrer nofollow");
            a.attr("target", "_blank");
        }
        // img[src]
        for (Element img : doc.select("img[src]")) {
            String src = img.attr("src");
            String abs = absolutifyUrl(src, baseUrl);
            img.attr("src", abs);
        }
        // source[srcset] （picture/webp）
        for (Element s : doc.select("source[srcset]")) {
            String srcset = s.attr("srcset");
            s.attr("srcset", absolutifySrcset(srcset, baseUrl));
        }
        return doc.body().html();
    }

    private static String absolutifyUrl(String url, String baseUrl) {
        if (url == null || url.isEmpty()) return url;
        String u = url.trim();
        if (u.startsWith("//")) {
            return "https:" + u;
        }
        if (u.startsWith("#")) {
            // 保留页面内锚点：拼接到首页（也可拼接到当前帖子的 link，但此处无上下文）
            return baseUrl + "/" + u;
        }
        try {
            URI base = URI.create(ensureTrailingSlash(baseUrl));
            URI abs = base.resolve(u);
            return abs.toString();
        } catch (Exception e) {
            return url;
        }
    }

    private static String absolutifySrcset(String srcset, String baseUrl) {
        if (srcset == null || srcset.isEmpty()) return srcset;
        String[] parts = srcset.split(",");
        List<String> out = new ArrayList<>(parts.length);
        for (String part : parts) {
            String p = part.trim();
            if (p.isEmpty()) continue;
            String[] seg = p.split("\\s+");
            String url = seg[0];
            String size = seg.length > 1 ? seg[1] : "";
            out.add(absolutifyUrl(url, baseUrl) + (size.isEmpty() ? "" : " " + size));
        }
        return String.join(", ", out);
    }

    /* ===================== 摘要 & enclosure ===================== */

    private static String textSummary(String html, int maxLen) {
        if (html == null) return "";
        String text = Jsoup.parse(html).text().replaceAll("\\s+", " ").trim();
        if (text.length() <= maxLen) return text;
        return text.substring(0, maxLen) + "…";
    }

    private String firstImage(String content) {
        if (content == null) return null;
        Matcher m = MD_IMAGE.matcher(content);
        if (m.find()) return m.group(1);
        m = HTML_IMAGE.matcher(content);
        if (m.find()) return m.group(1);
        // 再从纯 HTML 里解析一次（如果传入的是渲染后的）
        try {
            Document doc = Jsoup.parse(content);
            Element img = doc.selectFirst("img[src]");
            if (img != null) return img.attr("src");
        } catch (Exception ignored) {}
        return null;
    }

    private static String getMimeType(String url) {
        String lower = url == null ? "" : url.toLowerCase(Locale.ROOT);
        if (lower.endsWith(".png"))  return "image/png";
        if (lower.endsWith(".gif"))  return "image/gif";
        if (lower.endsWith(".webp")) return "image/webp";
        if (lower.endsWith(".svg"))  return "image/svg+xml";
        if (lower.endsWith(".avif")) return "image/avif";
        if (lower.endsWith(".jpg") || lower.endsWith(".jpeg")) return "image/jpeg";
        // 默认兜底
        return "image/jpeg";
    }

    /* ===================== 附加区块（原文链接 + 精选评论） ===================== */

    /**
     * 将“原文链接 + 精选评论（最多 10 条）”以优雅的 Markdown 形式渲染为 HTML，
     * 并做 sanitize + 绝对化，然后拼入 content:encoded 尾部。
     */
    private static String buildFooterHtml(String baseUrl, String originalLink, List<Comment> topComments) {
        StringBuilder md = new StringBuilder(256);

        // 分割线
        md.append("\n\n---\n\n");

        // 原文链接（强调 + 可点击）
        md.append("**原文链接：** ")
                .append("[").append(originalLink).append("](").append(originalLink).append(")")
                .append("\n\n");

        // 精选评论（仅当有评论时展示）
        if (topComments != null && !topComments.isEmpty()) {
            md.append("### 精选评论（Top ").append(Math.min(10, topComments.size())).append("）\n\n");
            for (Comment c : topComments) {
                String author = usernameOf(c);
                String content = nullSafe(c.getContent()).replace("\r", "");
                // 使用引用样式展示，提升可读性
                md.append("> @").append(author).append(": ").append(content).append("\n\n");
            }
        }

        // 渲染为 HTML，并保持和正文一致的处理流程
        String html = renderMarkdown(md.toString());
        String safe = sanitizeHtml(html);
        return absolutifyHtml(safe, baseUrl);
    }

    private static String usernameOf(Comment c) {
        if (c == null) return "匿名";
        try {
            Object authorObj = c.getAuthor();
            if (authorObj == null) return "匿名";
            // 反射避免直接依赖实体字段名变化（也可直接强转到具体类型）
            String username;
            try {
                username = (String) authorObj.getClass().getMethod("getUsername").invoke(authorObj);
            } catch (Exception e) {
                username = null;
            }
            if (username == null || username.isEmpty()) return "匿名";
            return username;
        } catch (Exception ignored) {
            return "匿名";
        }
    }

    /* ===================== 时间/字符串/XML ===================== */

    private static String toRfc1123Gmt(ZonedDateTime zdt) {
        return zdt.withZoneSameInstant(ZoneId.of("GMT")).format(RFC1123);
    }

    private static String cdata(String s) {
        if (s == null) return "<![CDATA[]]>";
        // 防止出现 "]]>" 终止标记破坏 CDATA
        return "<![CDATA[" + s.replace("]]>", "]]]]><![CDATA[>") + "]]>";
    }

    private static void elem(StringBuilder sb, String name, String value) {
        sb.append('<').append(name).append('>').append(value).append("</").append(name).append('>');
    }

    private static String escapeXml(String s) {
        if (s == null) return "";
        return s.replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;")
                .replace("\"", "&quot;").replace("'", "&apos;");
    }

    private static String trimTrailingSlash(String s) {
        if (s == null) return "";
        return s.endsWith("/") ? s.substring(0, s.length() - 1) : s;
    }

    private static String ensureTrailingSlash(String s) {
        if (s == null || s.isEmpty()) return "/";
        return s.endsWith("/") ? s : s + "/";
    }

    private static String nullSafe(String s) { return s == null ? "" : s; }
}