如何用PyMongo查询MongoDB中含指定关键词的评论商品

现有MongoDB商品数据结构如下（reviews为嵌套JSON对象），需要找出评论内容包含特定单词（如“good”）的商品，但使用PyMongo执行查询后返回空结果，寻求正确实现方法：

{
  "_id": {
    "$oid": "64175a237438834b2397d676"
  },
  "productid": "B002PD61Y4",
  "product_name": "D-Link DWA-131 300 Mbps Wireless Nano USB Adapter (Black)",
  "actual_price": 19.64,
  "discount_percentage": 0.58,
  "product_desc": "Connects your computer to a high-speed wireless network",
  "rating_count": 8131,
  "rating": 4.1,
  "category": [
    "Accessories",
    "Computers",
    "NetworkAdapters",
    "NetworkingDevices",
    "WirelessUSBAdapters"
  ],
  "reviews": {
    "AGMG74N6WQGI376W7GEJJ4XD3ARQ": "Its not plug an play u need to install the driver and will support till 2.4 Ghz not above that",
    "AEHQQTEDMSXRGSBDDEIH3JF4AOMQ": "Good",
    "AGA2PZGWMQIRA46VYOTICFE7KCBA": "good tool to use for",
    "AHI2QJ4CLTCQWACDIC2LDFJPDAPA": "Brand is always good",
    "AETRIARSUFSMNG5LFJZMW6CBJMMQ": "Works well with cpplus dvr",
    "AEXAFY7V2ZRZI2GD2J6KDOWBZUBQ": "Overall good and a better experience",
    "AG7QMBEFFY2LJJKKEVWMJU2BMNRQ": "Surveillance Camera In My House",
    "AHAEBXTXQDY355AGFMFX3Z2VAAUQ": "It is useful to me."
  }
}

尝试的查询代码：

# Find reviews with the value "great"
query = {"reviews": {"$elemMatch": {"$regex": "great", "$options": "i"}}}
results = collection.find(query)

# Print the matching documents
for result in results:
    print(result)

$elemMatch是用于查询数组元素的操作符，但你的reviews是一个嵌套对象（键为随机ID，值为评论内容），不是数组，所以这个操作符完全不适用，导致返回空结果。

方法1：使用$where遍历对象值 #

通过JavaScript函数遍历reviews对象的所有值，检查是否匹配目标正则：

# 查询包含"good"的评论（不区分大小写）
query = {
    "$where": """
        for (let key in this.reviews) {
            if (/good/i.test(this.reviews[key])) {
                return true;
            }
        }
        return false;
    """
}
results = collection.find(query)

for result in results:
    print(result)

注意：$where性能相对较低，不适合超大规模数据集，因为它需要逐文档执行JavaScript代码。

方法2：使用聚合管道转换为数组查询 #

通过$objectToArray将reviews对象转换为键值对数组，再用$match和$regex匹配：

pipeline = [
    {
        "$addFields": {
            "review_entries": {"$objectToArray": "$reviews"}
        }
    },
    {
        "$match": {
            "review_entries.v": {"$regex": "good", "$options": "i"}
        }
    },
    # 可选：移除临时生成的review_entries字段
    {
        "$project": {
            "review_entries": 0
        }
    }
]

results = collection.aggregate(pipeline)

for result in results:
    print(result)

这种方法性能优于$where，适合大多数场景，尤其是数据集较大时。

方法3：重构数据结构（推荐长期方案） #

如果频繁需要查询评论内容，建议将reviews改为数组结构，比如：

"reviews": [
    {"id": "AGMG74N6WQGI376W7GEJJ4XD3ARQ", "content": "Its not plug an play..."},
    {"id": "AEHQQTEDMSXRGSBDDEIH3JF4AOMQ", "content": "Good"}
]

之后就可以用你最初尝试的$elemMatch查询：

query = {"reviews": {"$elemMatch": {"content": {"$regex": "good", "$options": "i"}}}}
results = collection.find(query)

数组结构更符合MongoDB的设计范式，查询和索引优化都更方便。

内容的提问来源于stack exchange，提问作者Exx