mongodb如何对子文档进行分页

> db.test1.find().pretty()
{
    "_id" : ObjectId("55dc145ef754a3342000002d"),
    "games" : [
        {
            "id" : 1128,
            "name_zh" : "追击野兽",
            "customize_update_time" : "2015-05-18 00:00:00"
        },
        {
            "id" : 3276,
            "name_zh" : "dasda",
            "customize_update_time" : "2015-05-15 00:00:00"
        },
        {
            "id" : 3416,
            "name_zh" : "围剿外星客",
            "customize_update_time" : "2015-04-25 00:00:00"
        },
        {
            "id" : 10357,
            "name_zh" : "合金弹头：防御",
            "customize_update_time" : "0000-00-00 00:00:00"
        },
        {
            "id" : 10360,
            "name_zh" : "绿色忍者蛙年",
            "customize_update_time" : "2015-06-30 15:52:10"
        },
        {
            "id" : 10364,
            "name_zh" : "杀死马里奥",
            "customize_update_time" : "2015-06-30 16:08:09"
        },
        {
            "id" : 10366,
            "name_zh" : "通过繁殖征服世界",
            "customize_update_time" : "2015-06-30 16:21:22"
        },
        {
            "id" : 10229,
            "name_zh" : "冰块切割",
            "customize_update_time" : "2014-11-04 00:00:00"
        },
        {
            "id" : 10356,
            "name_zh" : "吃冰淇淋的怪房子",
            "customize_update_time" : "0000-00-00 00:00:00"
        },
        {
            "id" : 10358,
            "name_zh" : "来杯果汁",
            "customize_update_time" : "0000-00-00 00:00:00"
        },
        {
            "id" : 10362,
            "name_zh" : "梦游先生",
            "customize_update_time" : "2015-06-30 15:58:13"
        },
        {
            "id" : 10363,
            "name_zh" : "清凉方冰冰",
            "customize_update_time" : "2015-06-30 16:04:56"
        },
        {
            "id" : 10365,
            "name_zh" : "嗜血狂鲨2",
            "customize_update_time" : "2015-06-30 16:15:58"
        },
        {
            "id" : 10367,
            "name_zh" : "外星人爱牛奶",
            "customize_update_time" : "2015-06-30 16:28:17"
        }
    ]
}

取games的前3条数据：

> db.test1.find({"_id" : ObjectId("55dc145ef754a3342000002d")},{"games":{ "$slice":[0,3]}}).pretty()
{
    "_id" : ObjectId("55dc145ef754a3342000002d"),
    "games" : [
        {
            "id" : 1128,
            "name_zh" : "追击野兽",
            "customize_update_time" : "2015-05-18 00:00:00"
        },
        {
            "id" : 3276,
            "name_zh" : "dasda",
            "customize_update_time" : "2015-05-15 00:00:00"
        },
        {
            "id" : 3416,
            "name_zh" : "围剿外星客",
            "customize_update_time" : "2015-04-25 00:00:00"
        }
    ]
}

取第四条到第6条数据：

> db.test1.find({"_id" : ObjectId("55dc145ef754a3342000002d")},{"games":{ "$slice":[3,3]}}).pretty()
{
    "_id" : ObjectId("55dc145ef754a3342000002d"),
    "games" : [
        {
            "id" : 10357,
            "name_zh" : "合金弹头：防御",
            "customize_update_time" : "0000-00-00 00:00:00"
        },
        {
            "id" : 10360,
            "name_zh" : "绿色忍者蛙年",
            "customize_update_time" : "2015-06-30 15:52:10"
        },
        {
            "id" : 10364,
            "name_zh" : "杀死马里奥",
            "customize_update_time" : "2015-06-30 16:08:09"
        }
    ]
}

依次类推，即可。 "$slice":[3,3] 第一个3表示查询数组下标的起始位置，第二个3表示取的数据条数。建议games不要过多，不然会超出文档限制16M。不过这个可能是设计问题，我多想了。

第二种方法：就是用代码从数据库中取出来，将games里面的每一个子文档封装成model，放在缓存中做分页，而不是数据库级别的分页也可实现

第三种方法：从数据可看出games里面的子文档是按照id进行排序的，那么也就是说子文档是可比较的，那么就可以使用$gt和$lt，接合$size取数据，我没有试。你可以尝试一下。不过$slice获取数组子集更方便一点。