fromis_9/backend/scrape_search.cjs

const https = require("https");
const http = require("http");
const mysql = require("mysql2/promise");

// 설정
const NITTER_URL = "http://nitter:8080";
const USERNAME = "realfromis_9";
const DELAY_MS = 1500;

// 검색 기간 (X 계정 이관일 ~ 기존 스크래핑 시작점)
const SEARCH_SINCE = "2025-04-24";
const SEARCH_UNTIL = "2025-06-16";

// DB 연결
const dbConfig = {
  host: process.env.DB_HOST || "mariadb",
  user: process.env.DB_USER || "fromis9_user",
  password: process.env.DB_PASSWORD || "fromis9_password",
  database: process.env.DB_NAME || "fromis9",
};

async function fetchPage(url) {
  return new Promise((resolve, reject) => {
    const client = url.startsWith("https") ? https : http;
    client
      .get(url, (res) => {
        let data = "";
        res.on("data", (chunk) => (data += chunk));
        res.on("end", () => resolve(data));
      })
      .on("error", reject);
  });
}

function parseDateTime(timeStr) {
  if (!timeStr) return null;
  try {
    const cleaned = timeStr.replace(" · ", " ").replace(" UTC", "");
    const date = new Date(cleaned + " UTC");
    if (isNaN(date.getTime())) return null;
    return date.toISOString().slice(0, 19).replace("T", " ");
  } catch (e) {
    return null;
  }
}

function extractSearchTweets(html) {
  const tweets = [];
  const tweetContainers = html.split('class="timeline-item ');

  for (let i = 1; i < tweetContainers.length; i++) {
    const container = tweetContainers[i];
    const tweet = {};

    tweet.isPinned = false;
    tweet.isRetweet = container.includes('class="retweet-header"');

    const linkMatch = container.match(/href="\/[^\/]+\/status\/(\d+)/);
    tweet.id = linkMatch ? linkMatch[1] : null;

    const timeMatch = container.match(
      /<span class="tweet-date"[^>]*><a[^>]*title="([^"]+)"/
    );
    tweet.time = timeMatch ? parseDateTime(timeMatch[1]) : null;

    const contentMatch = container.match(
      /<div class="tweet-content[^"]*"[^>]*>([\s\S]*?)<\/div>/
    );
    if (contentMatch) {
      tweet.text = contentMatch[1]
        .replace(/<br\s*\/?>/g, "\n")
        .replace(/<a[^>]*>([^<]*)<\/a>/g, "$1")
        .replace(/<[^>]+>/g, "")
        .trim();
    }

    const imageMatches = container.match(/href="\/pic\/([^"]+)"/g);
    tweet.images = [];
    if (imageMatches) {
      imageMatches.forEach((match) => {
        const urlMatch = match.match(/href="\/pic\/([^"]+)"/);
        if (urlMatch) {
          const decoded = decodeURIComponent(urlMatch[1]);
          tweet.images.push("https://pbs.twimg.com/" + decoded);
        }
      });
    }

    tweet.hasVideo =
      container.includes("gallery-video") ||
      container.includes("video-container");

    tweet.url = tweet.id
      ? `https://x.com/${USERNAME}/status/${tweet.id}`
      : null;

    if (tweet.id) {
      tweets.push(tweet);
    }
  }

  return tweets;
}

function extractNextCursor(html) {
  const cursorMatch = html.match(
    /class="show-more"[^>]*>\s*<a href="[^"]*cursor=([^"&]+)/
  );
  return cursorMatch ? cursorMatch[1] : null;
}

async function saveTweets(pool, tweets) {
  let saved = 0;
  for (const tweet of tweets) {
    try {
      const [result] = await pool.query(
        `INSERT IGNORE INTO x_tweets (id, username, text, created_at, is_retweet, is_pinned, images, has_video, url)
         VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)`,
        [
          tweet.id,
          USERNAME,
          tweet.text,
          tweet.time,
          tweet.isRetweet,
          tweet.isPinned,
          JSON.stringify(tweet.images),
          tweet.hasVideo,
          tweet.url,
        ]
      );
      if (result.affectedRows > 0) saved++;
    } catch (e) {
      console.error(`저장 오류 (ID: ${tweet.id}):`, e.message);
    }
  }
  return saved;
}

async function main() {
  console.log("=".repeat(60));
  console.log("X 트윗 검색 스크래핑 (누락 기간)");
  console.log("=".repeat(60));
  console.log(`대상: @${USERNAME}`);
  console.log(`기간: ${SEARCH_SINCE} ~ ${SEARCH_UNTIL}`);
  console.log("");

  const pool = await mysql.createPool(dbConfig);

  const searchQuery = encodeURIComponent(
    `from:${USERNAME} since:${SEARCH_SINCE} until:${SEARCH_UNTIL}`
  );
  let cursor = null;
  let pageNum = 1;
  let totalSaved = 0;
  let consecutiveEmpty = 0;

  while (true) {
    const url = cursor
      ? `${NITTER_URL}/search?f=tweets&q=${searchQuery}&cursor=${cursor}`
      : `${NITTER_URL}/search?f=tweets&q=${searchQuery}`;

    console.log(`[페이지 ${pageNum}] 검색 중...`);

    try {
      const html = await fetchPage(url);
      const tweets = extractSearchTweets(html);

      if (tweets.length === 0) {
        consecutiveEmpty++;
        console.log(`  -> 트윗 없음 (연속 ${consecutiveEmpty}회)`);
        if (consecutiveEmpty >= 3) {
          console.log("\n연속 3페이지 트윗 없음. 스크래핑 완료.");
          break;
        }
      } else {
        consecutiveEmpty = 0;
        const saved = await saveTweets(pool, tweets);
        totalSaved += saved;
        console.log(
          `  -> ${tweets.length}개 추출, ${saved}개 저장 (누적: ${totalSaved})`
        );
      }

      const nextCursor = extractNextCursor(html);
      if (!nextCursor) {
        console.log("\n다음 페이지 없음. 스크래핑 완료.");
        break;
      }

      cursor = nextCursor;
      pageNum++;

      await new Promise((r) => setTimeout(r, DELAY_MS));
    } catch (error) {
      console.error(`  -> 오류: ${error.message}`);
      consecutiveEmpty++;
      if (consecutiveEmpty >= 5) {
        console.log("\n연속 오류. 스크래핑 중단.");
        break;
      }
      await new Promise((r) => setTimeout(r, DELAY_MS * 3));
    }
  }

  console.log("\n" + "=".repeat(60));
  console.log("검색 스크래핑 완료");
  console.log(`추가 저장: ${totalSaved}개`);
  console.log("=".repeat(60));

  const [stats] = await pool.query(`
    SELECT 
      COUNT(*) as total,
      SUM(is_retweet) as retweets,
      SUM(NOT is_retweet) as original,
      MIN(created_at) as oldest,
      MAX(created_at) as newest
    FROM x_tweets
  `);
  console.log("\n[전체 통계]");
  console.log(stats[0]);

  await pool.end();
  process.exit(0);
}

main().catch((err) => {
  console.error("치명적 오류:", err);
  process.exit(1);
});