在Golang项目中实现MongoDB聚合regexMatch逻辑的语法求助

我在Studio3T里写了一段能正常运行的MongoDB聚合查询，现在要迁移到正在开发的Golang项目里。但我对bson.A、bson.D、bson.M这些类型的适用场景不太清楚，而且MongoDB查询和Golang代码的语法、括号结构转换总是出错，急需语法上的帮助。

现有MongoDB文档数据 #

{
    "errorMsg" : " Authorization is a duplicate; Tracking Number(s): AAAA-BBBB"
}
{
    "errorMsg" : " Authorization is a duplicate; Tracking Number(s): CCCC-DDDD"
}
{
    "errorMsg" : " Authorization is a duplicate of IP123456789; Tracking Number(s): EEEE-FFFF"
}
{
    "errorMsg" : " authorizationNbr [GGGG-HHHH] INVALID: unknown origin"
}
{
    "errorMsg" : " authorizationNbr [IIII-JJJJ] INVALID: unknown origin"
}
{
    "errorMsg" : " Unable to update, apologies . Please resubmit: AUTH_CRUD_API: %!s(<nil>)"
}
{
    "errorMsg" : " Unable to update, apologies . Please resubmit"
}

期望处理结果 #

{
    "errorMsg" : " Authorization is a duplicate"
}
{
    "errorMsg" : " Authorization is a duplicate"
}
{
    "errorMsg" : " Authorization is a duplicate"
}
{
    "errorMsg" : " authorizationNbr"
}
{
    "errorMsg" : " authorizationNbr"
}
{
    "errorMsg" : " Unable to update"
}
{
    "errorMsg" : " Unable to update"
}

Studio3T中的addFields阶段查询 #

addFields
{
    "errorMsg": { "$cond": [{ "$regexMatch": {"input": "$errorMsg", "regex": "Authorization is a duplicate"}}, "Authorization is a duplicate", 
        {"$cond": [{ "$regexMatch": {"input": "$errorMsg", "regex": "authorizationNbr"}}, "authorizationNbr", 
            {"$cond" : [{"$regexMatch" : {"input" : "$errorMsg", "regex" : "[Uu]nable to update, apologies"}}, "Unable to update", "$errorMsg"]}]}]}
}

我写的Golang代码（需要修正） #

addFields := bson.D{{"$addFields", bson.D{
        {"errorMsg", bson.D{{"$cond", bson.A{bson.D{{"$regexMatch", bson.A{"input": "$errorMsg", "regex": "Authorization is a duplicate"}}}, "Authorization is a duplicate",
            {"$cond", bson.A{bson.D{{"$regexMatch", bson.A{"input": "$errorMsg", "regex": "authorizationNbr"}}}, "authorizationNbr",
                {"$cond", bson.A{bson.D{{"$regexMatch", bson.A{"input": "$errorMsg", "regex": "[Uu]nable to update, apologies"}}}, "Unable to update", "$errorMsg"}}}}
            }}}},
    }}}

先明确bson类型的使用场景 #

• bson.D：有序键值对列表，适合需要严格保持字段顺序的场景（比如聚合阶段定义、$cond的条件判断结构）
• bson.A：数组类型，对应MongoDB中的数组（比如$cond的参数列表，因为$cond是按[条件, 满足值, 不满足值]的顺序传参）
• bson.M：无序map结构，适合不需要顺序的普通文档，写法更简洁

修正后的Golang代码 #

你的代码核心错误是把$regexMatch的参数写成了bson.A，但$regexMatch需要的是键值对文档，不是数组。修正后的代码如下：

addFields := bson.D{
    {"$addFields", bson.D{
        {"errorMsg", bson.D{
            {"$cond", bson.A{
                // 第一个$cond：匹配Authorization is a duplicate
                bson.D{
                    {"$regexMatch", bson.D{
                        {"input", "$errorMsg"},
                        {"regex", "Authorization is a duplicate"},
                    }},
                },
                "Authorization is a duplicate",
                // 嵌套第二个$cond
                bson.D{
                    {"$cond", bson.A{
                        bson.D{
                            {"$regexMatch", bson.D{
                                {"input", "$errorMsg"},
                                {"regex", "authorizationNbr"},
                            }},
                        },
                        "authorizationNbr",
                        // 嵌套第三个$cond
                        bson.D{
                            {"$cond", bson.A{
                                bson.D{
                                    {"$regexMatch", bson.D{
                                        {"input", "$errorMsg"},
                                        {"regex", "[Uu]nable to update, apologies"},
                                    }},
                                },
                                "Unable to update",
                                "$errorMsg",
                            }},
                        },
                    }},
                },
            }},
        }},
    }},
}

简化写法（用bson.M更易读） #

如果不需要严格字段顺序，用bson.M可以大幅简化代码：

addFields := bson.D{
    {"$addFields", bson.M{
        "errorMsg": bson.M{
            "$cond": bson.A{
                bson.M{"$regexMatch": bson.M{"input": "$errorMsg", "regex": "Authorization is a duplicate"}},
                "Authorization is a duplicate",
                bson.M{
                    "$cond": bson.A{
                        bson.M{"$regexMatch": bson.M{"input": "$errorMsg", "regex": "authorizationNbr"}},
                        "authorizationNbr",
                        bson.M{
                            "$cond": bson.A{
                                bson.M{"$regexMatch": bson.M{"input": "$errorMsg", "regex": "[Uu]nable to update, apologies"}},
                                "Unable to update",
                                "$errorMsg",
                            },
                        },
                    },
                },
            },
        },
    }},
}

错误点说明 #

你原来的代码中，$regexMatch的参数用了bson.A{"input": "$errorMsg", "regex": "..."}，这是错误的——$regexMatch需要的是键值对文档（对应MongoDB的{"input": ..., "regex": ...}），而bson.A仅用于表示MongoDB中的数组类型（比如$cond的三个参数组成的数组）。

内容的提问来源于stack exchange，提问作者Centdesk67