powershell - Powershell：滞后函数优于属性

Question

大家好，

我正在寻找是否有可能实现滞后功能 - 无需遍历整个列表，例如，查看以下示例：

$input = '[
    { "user_id":"1", "earned":"100", "valid_at": "20200101"},
    { "user_id":"1", "earned":"200", "valid_at": "20200102"},
    { "user_id":"1", "earned":"10", "valid_at": "20200103"},
    { "user_id":"2", "earned":"10", "valid_at": "20200101"},
    { "user_id":"2", "earned":"40", "valid_at": "20200103"},
    { "user_id":"3", "earned":"400", "valid_at": "20200101"},
    { "user_id":"3", "earned": null, "valid_at": "20200102"},
    { "user_id":"3", "earned": "20", "valid_at": "20200103"}
]'

$expected_output = '[
    { "user_id":"1", "earned":"100", "valid_at": "20200101", "last_earned":null},
    { "user_id":"1", "earned":"200", "valid_at": "20200102", "last_earned":"100"},
    { "user_id":"1", "earned":"10", "valid_at": "20200103", "last_earned":"200"},
    { "user_id":"2", "earned":"10", "valid_at": "20200101", "last_earned":null},
    { "user_id":"2", "earned":"40", "valid_at": "20200103", "last_earned":"10"},
    { "user_id":"3", "earned":"400", "valid_at": "20200101", "last_earned":null},
    { "user_id":"3", "earned": null, "valid_at": "20200102", "last_earned":"400"},
    { "user_id":"3", "earned": "20", "valid_at": "20200103", "last_earned":null}
]'

$json_input = ConvertFrom-Json –InputObject $input

$expected_json_output = ConvertFrom-Json –InputObject $expected_output

这将创建以下输入：

连同以下预期输出：

现在我的问题是：

• 有没有办法以编程方式构建last_earned列，而不必遍历整个列表（即，以某种方式使用 Group-Object 或 Where-Object？
• 目前，我正在遍历整个列表 - 但必须快速意识到，虽然它按预期工作，但如果你有很多记录来计算它，效率会非常低
• 我目前正在为每个单独的user_id单独计算这个（即，
  • 选择它，
  • 排序结果，然后
  • 检查最后赚取的

标签： powershell

Answer 1

假设输入已经排序，您可以通过使用哈希表跟踪每个用户 id 的最后一次看到的值来一次性完成此操作：

$last = @{}
$json_input |Select-Object *,@{Name='earned_day_before';Expression={$last[$_.user_id];$last[$_.user_id]=$_.earned}}

---

如果没有，您可以先按用户 ID 对所有记录进行排序，然后按日期排序：

$last = @{}
$json_input |Sort-Object user_id,valid_at |Select-Object *,@{Name='earned_day_before';Expression={$last[$_.user_id];$last[$_.user_id]=$_.earned}}