轻应用大文件分片上传：断点续传、秒传与完整性校验的落地实现

async function calcHash(file, chunkSize = 5 * 1024 * 1024) {
  const spark = new SparkMD5.ArrayBuffer();
  for (let i = 0; i < file.size; i += chunkSize) {
    spark.append(await file.slice(i, i + chunkSize).arrayBuffer());
  }
  return spark.end();                       // 整文件指纹，用于秒传与合并校验
}

async function uploadChunks(file, fileHash, concurrency = 4) {
  const chunks = createChunks(file);        // 切成 {index, blob} 列表
  const pool = [];
  for (let i = 0; i < chunks.length; i++) {
    const task = uploadOne(fileHash, i, chunks[i]);
    pool.push(task);
    if (pool.length >= concurrency) {       // 控制并发，完成一个再补一个
      await Promise.race(pool);
      pool.splice(pool.findIndex(t => t.settled), 1);
    }
  }
  await Promise.all(pool);
}

async function resumeUpload(file, fileHash) {
  // 查询已上传分片列表
  const { uploaded } = await fetch(`/upload/check?hash=${fileHash}`).then(r => r.json());
  const chunks = createChunks(file);
  const todo = chunks
    .map((c, i) => i)
    .filter(i => !uploaded.includes(i));    // 只挑出缺失分片
  for (const i of todo) {
    await retry(() => uploadOne(fileHash, i, chunks[i]), 3);  // 单片失败只重传该片
  }
  return mergeRequest(fileHash, file.name); // 缺口补齐后请求合并
}

from fastapi import APIRouter
router = APIRouter()

@router.get("/upload/check")
def check(file_hash: str):
    # 已存在完整文件 -> 秒传；否则返回已收到的分片用于续传
    if file_store.is_complete(file_hash):
        return {"instant": True, "url": file_store.get_url(file_hash)}
    return {"instant": False, "uploaded": file_store.list_chunks(file_hash)}

@router.post("/upload/merge")
def merge(file_hash: str, filename: str, total: int):
    have = file_store.list_chunks(file_hash)
    missing = [i for i in range(total) if i not in have]
    if missing:                      # 分片不全不许合并，避免拼出残缺文件
        return {"ok": False, "missing": missing}
    url = file_store.merge_in_order(file_hash, filename, total)
    if file_store.sha256_of(url) != file_store.meta(file_hash)["sha"]:
        file_store.delete(url)       # 总哈希不符，判定损坏并重来
        return {"ok": False, "error": "hash_mismatch"}
    return {"ok": True, "url": url}