Question

我有以下 JSON 上下文，我需要聚合 conversion_token 组下的不同关键字及其重复次数；例如：

"conversion_token": [
                    {
                        "keyword": "DBMS",
                        "count":4,
                        "classify":2
                            }

关键字 DBMS 在提供的 json 中的不同结构下多次使用，聚合应显示

"conversion_token": [
                    {
                        "keyword": "DBMS",
                        "count":6,
                        "classify":2
                            }
                {
                        "keyword": "NVL",
                        "count":2,`enter code here`
                        "classify":2
                            }

等等。

我该怎么做？

{
    "select_emp": {
        "specification": {
            "input": [
                "p_empno"
            ],
            "declare_stmt": {
                "anchorvariable": [
                    "V_ENAME",
                    "V_HIREDATE",
                    "V_TITLE",
                    "V_REPORTSTO",
                    "V_DISP_DATE",
                    "V_INS_COUNT",
                    "CITY_FROM"
                ],
                "tablename_variable": [
                    "EMPLOYEE.V_ENAME",
                    "EMPLOYEE.V_HIREDATE",
                    "EMPLOYEE.V_TITLE",
                    "EMPLOYEE.V_REPORTSTO",
                    "EMPLOYEE.V_DISP_DATE",
                    "EMPLOYEE.V_INS_COUNT",
                    "EMPLOYEE.CITY_FROM"
                ]
            }
        },
        "body": {
            "select_stmt1": {
                "columns": [
                    "FIRSNAME",
                    "HIREDATE",
                    "TITLE",
                    "REPORTSTO"
                ],
                "tablename": [
                    "EMPLOYEE"
                ],
                "conversion_token": [
                    {
                        "keyword": "NVL",
                        "count": 1,
                        "classify": 2
                    }
                ]
            },
            "select_stmt2": {
                "columns": [
                    "CITY"
                ],
                "tablename": [
                    "EMPLOYEE"
                ],
                "conversion_token": [
                    {
                        "keyword": "DECODE",
                        "count": 1,
                        "classify": 3
                    }
                ]
            },
            "dbms_stmt1": {
                "dbms_putline": [
                    "P_EMPNO",
                    "V_ENAME",
                    "V_DISP_DATE",
                    "V_REPORTSTO"
                ],
                "conversion_token": [
                    {
                        "keyword": "DBMS",
                        "count": 1,
                        "classify": 2
                    }
                ]
            },
            "forloop1": {
                "select_stmt": {
                    "columns": [
                        "EMPLOYEEID",
                        "ROWID"
                    ],
                    "tablename": [
                        "EMPLOYEE"
                    ],
                    "conversion_token": [
                        {
                            "keyword": "DBMS",
                            "count": 1,
                            "classify": 2
                        }
                    ]
                }
            },
            "merge_stmt1": {
                "merge_into": "EMPLOYEE",
                "merge_using": {
                    "columns": [
                        "EMPLOYEEID",
                        "LASTNAME",
                        "TITLE",
                        "BIRTHDATE",
                        "HIREDATE",
                        "ADDRESS",
                        "CITY",
                        "STATE",
                        "COUNTRY",
                        "POSTALCODE",
                        "PHONE",
                        "FAX",
                        "EMAIL",
                        "BONUS"
                    ],
                    "tablename": [
                        "EMPLOYEE"
                    ]
                },
                "merge_update": {
                    "columns": [
                        "BONUS"
                    ],
                    "tablename": [
                        "EMPLOYEE"
                    ]
                },
                "merge_delete": {
                    "columns": [
                        "BONUS"
                    ],
                    "tablename": [
                        "EMPLOYEE"
                    ]
                },
                "merge_insert": {
                    "columns": [
                        "EMPLOYEEID",
                        "LASTNAME",
                        "FIRSTNAME",
                        "TITLE",
                        "BIRTHDATE",
                        "HIREDATE",
                        "ADDRESS",
                        "CITY",
                        "STATE",
                        "COUNTRY",
                        "POSTALCODE",
                        "PHONE",
                        "FAX",
                        "EMAIL",
                        "BONUS"
                    ],
                    "tablename": [
                        "EMPLOYEE"
                    ]
                },
                "conversion_token": [
                    {
                        "keyword": "Merge",
                        "count": 1,
                        "classify": 4
                    }
                ]
            },
            "exception_handling1": {
                "dbms_putline": [
                    "P_EMPNO"
                ],
                "conversion_token": [
                    {
                        "keyword": "DBMS",
                        "count": 1,
                        "classify": 2
                    }
                ]
            }
        }
    }
}

Answer 1

首先，您需要使用 conversion_token 和 project 将 $concatArrays 数组合并为一个。然后使用 $unwind，您可以分隔每个 conversion_token 对象，使它们准备好进行聚合。在 $unwind 之后，您可以使用 keyword 按 $group 对它们进行分组，并且您可以取 count 的总和。

db.collection.aggregate([
{
    $project: {
        conversion_tokens: {
            $concatArrays:
                [
                    "$select_emp.body.select_stmt1.conversion_token",
                    "$select_emp.body.select_stmt2.conversion_token",
                    "$select_emp.body.dbms_stmt1.conversion_token",
                    "$select_emp.body.forloop1.conversion_token",
                    "$select_emp.body.merge_stmt1.conversion_token",
                    "$select_emp.body.exception_handling1.conversion_token"
                ]
        }
    }
}, {
    $unwind: "$conversion_tokens"
}, {
    $group: {
        _id: "$conversion_tokens.keyword",
        count: {
            $sum: "$conversion_tokens.count"
        }
    }
}
])

这将产生一个数组，如：

{
  "_id": "DBMS",
  "count": 3
}, {
  "_id": "NVL",
  "count": 1
}, {
  "_id": "DECODE",
  "count": 1
}

如果您想将 _id 键更改为 keyword，您可以再次使用 $project。

如何编写聚合查询来计算 Mongodb 中的计数

1 个答案: