为什么这种情况部分起作用？

Question

我有一个dict列表，我想删除一些特定键的值小于另一个dict中对应键的所有dict：

这是我的密码：

for d in clean_hosting_dicts:
    if (int(str(d["cores"]) ) < int(self.query["cores"]) ) or ( float(str(d["bandwidth"]).replace(",",".") ) < int(str(self.query["bandwidth"])) ) or ( float(str(d["storage"]).replace(",",".") ) < int(str(self.query["storage"])) ) or (float(str(d["ram"]).replace(",",".") ) < int(str(self.query["ram"])) ):
        del clean_hosting_dicts[clean_hosting_dicts.index(d)]

其中clean_hosting_dicts是我的dict列表，self.query是另一个dict，它有相同的键，但有其他值要对它们进行测试。你知道吗

以下是用于自我查询地址：

{"hosting_type" : "vps","cores": "3", "storage" :"2", "bandwidth" : "1", "ram" : "2", "ip" : "1"}

下面是代码运行前dict列表的json（未过滤列表）：

[
  {
    "support_score": "4,5",
    "name": "Hostgator Snappy 8000",
    "ip": "2",
    "support": "chat,email,docs,phone,ticket",
    "ram": "8",
    "storage": "240",
    "frequency_rate": 3,
    "cpu": "AMD Opteron 6378",
    "reviews": "3",
    "bandwidth": "3",
    "domains": "1",
    "cores": "4",
    "os": "CentOS 66",
    "price": "39,95",
    "guarantee": "45"
  },
  {
    "support_score": "3,5",
    "name": "Bluehost Standard",
    "ip": "1",
    "support": "email,chat,doc,phone,ticket",
    "ram": "2",
    "storage": "30",
    "frequency_rate": 3,
    "cpu": "AMD Opteron 6378",
    "reviews": "3,2",
    "bandwidth": "1",
    "domains": "1",
    "cores": "2",
    "os": "CentOS 66",
    "price": "19,99",
    "guarantee": "30"
  },
  {
    "support_score": "3,5",
    "name": "Bluehost Enhanced",
    "ip": "2",
    "support": "email,chat,doc,phone,ticket",
    "ram": "4",
    "storage": "60",
    "frequency_rate": 3,
    "cpu": "AMD Opteron 6378",
    "reviews": "3,2",
    "bandwidth": "2",
    "domains": "1",
    "cores": "2",
    "os": "CentOS 66",
    "price": "29,99",
    "guarantee": "30"
  },
  {
    "support_score": "3,5",
    "name": "Bluehost Ultimate",
    "ip": "2",
    "support": "email,chat,doc,phone,ticket",
    "ram": "8",
    "storage": "240",
    "frequency_rate": 3,
    "cpu": "AMD Opteron 6378",
    "reviews": "3,2",
    "bandwidth": "4",
    "domains": "1",
    "cores": "4",
    "os": "CentOS 66",
    "price": "59,99",
    "guarantee": "30"
  },
  {
    "support_score": "4",
    "name": "iPage Business",
    "ip": "2",
    "support": "email,chat,doc,phone,ticket",
    "ram": "4",
    "storage": "90",
    "frequency_rate": 3,
    "cpu": "Unknown",
    "reviews": "3,7",
    "bandwidth": "3",
    "domains": "1",
    "cores": "2",
    "os": "CentOS 64",
    "price": "47,99",
    "guarantee": "30"
  },
  {
    "support_score": "4,5",
    "name": "InMotion vps-1000ha-s",
    "ip": "3",
    "support": "email, phone, skype, ticket,chat",
    "ram": "4",
    "storage": "75",
    "frequency_rate": 3,
    "cpu": "Unknown",
    "reviews": "4,5",
    "bandwidth": "4",
    "domains": "1",
    "cores": 5,
    "os": "CentOS 66",
    "price": "29,99",
    "guarantee": "30"
  },
  {
    "support_score": "4,5",
    "name": "InMotion vps-3000ha-s",
    "ip": "5",
    "support": "email, phone, skype, ticket,chat",
    "ram": "8",
    "storage": "260",
    "frequency_rate": 3,
    "cpu": "Unknown",
    "reviews": "4,5",
    "bandwidth": "6",
    "domains": "1",
    "cores": 5,
    "os": "CentOS 66",
    "price": "74,99",
    "guarantee": "30"
  },
  {
    "support_score": "4",
    "name": "Fatcow Business",
    "ip": "2",
    "support": "email,chat,doc,phone,ticket",
    "ram": "4",
    "storage": "90",
    "frequency_rate": 3,
    "cpu": "Unknown",
    "reviews": "3,8",
    "bandwidth": "3",
    "domains": "1",
    "cores": "2",
    "os": "CentOS 64",
    "price": "47,99",
    "guarantee": "30"
  },
  {
    "support_score": "4",
    "name": "1and1 Cloud L",
    "ip": "1",
    "support": "email,doc,phone,ticket",
    "ram": "2",
    "storage": "80",
    "frequency_rate": 3,
    "cpu": "Intel Xeon",
    "reviews": "3",
    "bandwidth": "0",
    "domains": "1",
    "cores": "2",
    "os": "CentOS 7",
    "price": "14,99",
    "guarantee": "30"
  },
  {
    "support_score": "4,5",
    "name": "Hostgator Snappy 2000",
    "ip": "2",
    "support": "chat,email,docs,phone,ticket",
    "ram": "2",
    "storage": "120",
    "frequency_rate": 3,
    "cpu": "AMD Opteron 6376",
    "reviews": "3",
    "bandwidth": "1,5",
    "domains": "1",
    "cores": "2",
    "os": "CentOS 64",
    "price": "19,95",
    "guarantee": "45"
  }
]

在这个代码运行之后，删除所有不符合要求的dict，除了一个“名称为1&1 Cloud L的dict”，这个dict的核心值是“2”，而query["cores"]是大于2的4（我将所有这些值转换成数字）

我不知道为什么它没有被删除，我找不到任何有助于在线红外调试通过pdb。你知道吗

Answer 1

列表理解可用于根据条件对列表进行缩减：

clean_hosting_dicts = [
    d for d in clean_hosting_dicts
    if not ((int(str(d["cores"])) < int(self.query["cores"])) or (
        float(str(d["bandwidth"]).replace(",", ".")) < int(
            str(self.query["bandwidth"]))) or (
        float(str(d["storage"]).replace(",", ".")) < int(
            str(self.query["storage"]))) or (
        float(str(d["ram"]).replace(",", ".")) < int(str(self.query["ram"]))))
]

在OP的代码中，问题出现在从列表中删除迭代时。上面的方法可以工作，因为建立了一个新的列表，然后将其分配回原始名称。你知道吗

注意（从注释中）这将构建一个全新的对象，因此如果有人引用了原始列表，那么该列表将不会被更新。你知道吗

Answer 2

问题是您正在从正在迭代的列表中删除项，因此我建议您在其他列表上迭代，例如使用以下表达式：

for d in clean_hosting_dicts[:]:
    ...

所以你在一个相等的列表上迭代，虽然不是同一个对象，但是只在原始的列表中删除。你知道吗

希望这有帮助！你知道吗