fetch_code.yml

name: Fetch JTWC Products

# 触发条件：每10分钟自动运行，且允许手动触发
on:
  schedule:
    - cron: '0 */1 * * *'  # 每10分钟运行一次
  workflow_dispatch:  # 允许手动触发

jobs:
  fetch-and-commit:
    runs-on: ubuntu-latest
    
    steps:
      # 检出代码仓库
      - name: Checkout repository
        uses: actions/checkout@v4
        with:
          fetch-depth: 0  # 确保获取所有分支历史
      
      # 切换到jtwc分支，如果不存在则创建
      - name: Switch to jtwc branch or create it
        run: |
          if git ls-remote --heads origin jtwc | grep -q jtwc; then
            if git rev-parse --verify jtwc 2>/dev/null; then
              git checkout jtwc
            else
              git checkout -b jtwc origin/jtwc
            fi
            git pull origin jtwc
          else
            git checkout --orphan jtwc
            git rm -rf . 2>/dev/null || true
          fi
      
      # 从natyphoon获取活跃台风系统，构造JTWC链接并下载
      - name: Fetch active systems and download JTWC products
        run: |
          # 下载natyphoon目录页
          curl -o natyphoon_index.html "https://www.natyphoon.top/atcf/temp/?C=M;O=A"
          
          # 提取 FN (文件名) 和 LM (最后修改时间) 两个JS数组
          FN_RAW=$(sed -n 's/^const FN = \[\(.*\)\];$/\1/p' natyphoon_index.html)
          LM_RAW=$(sed -n 's/^const LM = \[\(.*\)\];$/\1/p' natyphoon_index.html)
          rm -f natyphoon_index.html
          
          # 转为换行分隔的列表（去掉引号）
          echo "$FN_RAW" | tr ',' '\n' | sed 's/"//g' > /tmp/fn_list.txt
          echo "$LM_RAW" | tr ',' '\n' | sed 's/"//g' > /tmp/lm_list.txt
          
          # 计算2天前的日期 (格式: YYYY/MM/DD)
          CUTOFF_DATE=$(date -u -d '2 days ago' +'%Y/%m/%d')
          
          # 配对文件名和日期，筛选最近2天的
          rm -f /tmp/active_typhoons.txt
          touch /tmp/active_typhoons.txt
          
          paste /tmp/fn_list.txt /tmp/lm_list.txt > /tmp/paired_list.txt
          while IFS=$'\t' read -r fname lm_date; do
            CLEAN_DATE=$(echo "$lm_date" | sed 's/^ *//')
            DATE_PART=$(echo "$CLEAN_DATE" | cut -d' ' -f1)
            if [[ -n "$DATE_PART" ]] && [[ "$DATE_PART" > "$CUTOFF_DATE" || "$DATE_PART" == "$CUTOFF_DATE" ]]; then
              echo -e "${fname}\t${CLEAN_DATE}" >> /tmp/active_typhoons.txt
            fi
          done < /tmp/paired_list.txt
          
          ACTIVE_COUNT=$(wc -l < /tmp/active_typhoons.txt | tr -d ' ')
          if [ "$ACTIVE_COUNT" -eq 0 ]; then
            echo "No active systems found in last 2 days."
          else
            # 临时目录存放下载的原始文件
            mkdir -p /tmp/jtwc_downloads
            rm -f /tmp/download_map.txt
            touch /tmp/download_map.txt
            
            # 对每个活跃的台风系统，构造JTWC URL并尝试下载
            while IFS=$'\t' read -r datfile sys_timestamp; do
              REGION=$(echo "$datfile" | sed -E 's/^b([a-z]{2})[0-9]+\.dat$/\1/')
              NUMYEAR=$(echo "$datfile" | sed -E 's/^b[a-z]{2}(.*)\.dat$/\1/')
              YEAR="${NUMYEAR: -4}"
              NUM="${NUMYEAR%${YEAR}}"
              SHORT_YEAR="${YEAR: -2}"
              
              for TYPE in web prog fix; do
                URL="https://www.metoc.navy.mil/jtwc/products/${REGION}${NUM}${SHORT_YEAR}${TYPE}.txt"
                TMPFILE="/tmp/jtwc_downloads/${REGION}${NUM}${SHORT_YEAR}${TYPE}.txt"
                
                HTTP_CODE=$(curl -o "$TMPFILE" -w '%{http_code}' -s "$URL")
                
                if [ "$HTTP_CODE" = "200" ]; then
                  FIRST_CHAR=$(head -c 1 "$TMPFILE")
                  if [ "$FIRST_CHAR" = "<" ]; then
                    echo "ERROR: $URL returned HTML error page"
                    rm -f "$TMPFILE"
                  else
                    PREFIX6=$(head -c 6 "$TMPFILE" | tr '[:upper:]' '[:lower:]')
                    echo -e "${PREFIX6}\t${TMPFILE}\t${sys_timestamp}\t${URL}" >> /tmp/download_map.txt
                  fi
                else
                  if [ "$HTTP_CODE" != "404" ]; then
                    echo "ERROR: $URL returned HTTP $HTTP_CODE"
                  fi
                  rm -f "$TMPFILE"
                fi
              done
            done < /tmp/active_typhoons.txt
            
            # 按 prefix6 分组，同一 prefix6 保留 natyphoon 更新时间最新的
            sort -t$'\t' -k1,1 -k3,3r /tmp/download_map.txt > /tmp/download_map_sorted.txt
            
            PREV_PREFIX=""
            while IFS=$'\t' read -r prefix6 tmpfile sys_ts url; do
              if [ "$prefix6" = "$PREV_PREFIX" ]; then
                echo "CONFLICT: '${prefix6}' keeping newer (ts=$sys_ts), skipped $url"
                rm -f "$tmpfile"
              else
                cp "$tmpfile" "${prefix6}.txt"
                rm -f "$tmpfile"
              fi
              PREV_PREFIX="$prefix6"
            done < /tmp/download_map_sorted.txt
            
            # 清理临时文件
            rm -rf /tmp/jtwc_downloads /tmp/download_map.txt /tmp/download_map_sorted.txt
          fi
          
          # 清理natyphoon临时文件
          rm -rf /tmp/active_typhoons.txt /tmp/fn_list.txt /tmp/lm_list.txt /tmp/paired_list.txt
      
      # 直接下载 ABPW10 公报（不依赖natyphoon）
      - name: Fetch ABPW10 product
        run: |
          TMPFILE="/tmp/abpwweb.txt"
          HTTP_CODE=$(curl -o "$TMPFILE" -w '%{http_code}' -s "https://www.metoc.navy.mil/jtwc/products/abpwweb.txt")
          
          if [ "$HTTP_CODE" = "200" ]; then
            FIRST_CHAR=$(head -c 1 "$TMPFILE")
            if [ "$FIRST_CHAR" = "<" ]; then
              echo "ERROR: abpwweb.txt returned HTML error page"
              rm -f "$TMPFILE"
            else
              PREFIX6=$(head -c 6 "$TMPFILE" | tr '[:upper:]' '[:lower:]')
              FINAL_NAME="${PREFIX6}.txt"
              if [ -f "$FINAL_NAME" ]; then
                echo "CONFLICT: '${PREFIX6}' already exists from natyphoon fetch, overwriting with abpwweb.txt"
              fi
              cp "$TMPFILE" "$FINAL_NAME"
              rm -f "$TMPFILE"
            fi
          else
            if [ "$HTTP_CODE" != "404" ]; then
              echo "ERROR: abpwweb.txt returned HTTP $HTTP_CODE"
            fi
            rm -f "$TMPFILE"
          fi
      
      # 配置Git用户信息
      - name: Configure Git
        run: |
          git config --global user.name "GitHub Actions"
          git config --global user.email "actions@github.com"
      
      # 提交更改
      - name: Commit and push changes
        run: |
          git add *.txt 2>/dev/null || true
          if git diff --cached --quiet; then
            echo "No changes to commit"
          else
            FILE_COUNT=$(git diff --cached --name-only | wc -l | tr -d ' ')
            git commit -m "Update JTWC products ($FILE_COUNT files): $(date -u +'%Y-%m-%d %H:%M:%S UTC')"
            git push origin jtwc
          fi