如何合并字典列表？

Question

我在这里扩展我原来的问题：How to append/merge list of dictionaries?。

我正在尝试在包含列表的单个字典列表之间合并一些数据。如果它们匹配，则合并将基于“object”和“semver”键进行。如果匹配相同的值，也会添加到他们给定的“部分”。给定以下数据：

data = [
        {
         "semver":"1.0.0",
         "sections":[
            {
               "name":"Add",
               "messages":[
                  "add: comment here"
               ]
            }
         ],
         "object":"files.sh"
      },
      {
         "semver":"1.0.0",
         "sections":[
            {
               "name":"Add",
               "messages":[
                  "add: Second comment here"
               ]
            }
         ],
         "object":"files.sh"
      },
      {
         "semver":"1.0.0",
         "sections":[
            {
               "name":"Fix",
               "messages":[
                  "Comment here"
               ]
            }
         ],
         "object":"files.sh"
      },
      {
         "semver":"2.0.0",
         "sections":[
            {
               "name":"Fix",
               "messages":[
                  "2.0.0 Fix Comment here"
               ]
            }
         ],
         "object":"files.sh"
      },
      {
         "semver":"2.0.0",
         "sections":[
            {
               "name":"Add",
               "messages":[
                  "2.0.0 Add Comment here"
               ]
            }
         ],
         "object":"files.sh"
      },
      {
         "semver":"2.0.0",
         "sections":[
            {
               "name":"Add",
               "messages":[
                  "2.0.0 comment for the NewFile"
               ]
            }
         ],
         "object":"NewFile.sh"
      },
]

我希望最终实现这一目标

data = [
        {
         "semver":"1.0.0",
         "sections":[
            {
               "name":"Add",
               "messages":[
                  "add: comment here",
                  "add: Second comment here"
               ]
            },
            {
               "name":"Fix",
               "messages":[
                  "Fix: comment here"
               ]
            }
         ],
         "object":"files.sh"
      },
        {
         "semver":"2.0.0",
         "sections":[
            {
               "name":"Add",
               "messages":[
                  "2.0.0 Add comment here",
               ]
            },
            {
               "name":"Fix",
               "messages":[
                  "2.0.0 Fix Comment here"
               ]
            }
         ],
         "object":"files.sh"
      },
      {
         "semver":"2.0.0",
         "sections":[
            {
               "name":"Add",
               "messages":[
                  "2.0.0 comment for the NewFile"
               ]
            }
         ],
         "object":"NewFile.sh"
      },

]

代码块

objects = {}  # mapping for object: object_data with sections
sections = defaultdict(list)  # mapping for object: all sections
for d in data:
    print(d["semver"])
    for k, v in list(d.items()):
        if v == d["semver"]:
            try:
                section = d.pop("sections")
                sections[d["object"]].extend(section)
                objects[d["object"]] = d  # populate with object data without sections
            except Exception as e:
                print(e)
                pass

output = []
for object_name, object_data in objects.items():
    object_data["sections"] = sections[object_name]
    output.append(object_data)

到目前为止，我正在循环遍历 dict 中的每个 k,v 对，但我无法理解这两个版本之间的匹配并在循环中附加到特定的 dict。

Answer 1

有 2 个改动需要做：

1. 将objects 和sections 中的键更改为基于由元组表示的object 和semver 的组合。
2. 添加一个辅助功能来合并部分中的消息

试试这个：

import json  # just for pretty print, you don't have to use it
from collections import defaultdict


def merge_messages(sections):
    d = defaultdict(list)
    for m in sections:
        d[m["name"]].extend(m["messages"])
    return [{"name": k, "messages": v} for k, v in d.items()]


objects = {}  # mapping for data with sections for (object, semver) combinations
sections = defaultdict(list)  # mapping for sections data for (object, semver) combinations
for d in data:
    section = d.pop("sections")
    sections[(d["object"], d["semver"])].extend(section)  # extends the sections for the object
    objects[(d["object"], d["semver"])] = d  # # populate with object data without sections

# merge between sections and objects by object key
output = []
for comb, object_data in objects.items():
    object_data["sections"] = merge_messages(sections[comb])
    output.append(object_data)
print(json.dumps(output, indent=4))  # just for pretty print