弹性搜索 - 所有父项和匹配嵌套

Question

我们的用例是有一个产品列表，并在嵌套字段中按客户编号显示最后购买日期。当客户搜索产品时，我们希望返回与其搜索词匹配的所有商品。如果有记录的话，inner_hits 应该只包含该客户的记录。因此，如果客户搜索玉米，他们可能会获得 4 个结果，但他们之前可能只购买过 2 个结果。因此 2 个文档没有内部匹配，另外 2 个文档只有 1 个文档的内部匹配，客户最后一次购买。

大多数商品都会将数百个客户购买加载到嵌套字段中，因此当我们只关心正在搜索的客户的最后购买日期时，我们不希望返回所有这些记录。

我们正在测试的当前查询

{
    "_source": {
        "includes": ["*"],
        "excludes": ["c_purchases"]
    },
    "query":  {
        "bool": {
            "must": [
                {
                    "match": {
                        "i_description": "corn"
                    }
                },
                {
                    "nested" : {
                        "path" : "c_purchases",
                        "inner_hits": {
                            "_source": ["*"]
                        },
                        "query" : {
                            "match": {
                                "c_purchases.customernumber": "1111"
                            }
                        }
                    }   
                }
            ]
        }
    }
}

上述查询的问题是它只返回客户购买的商品。因此，在示例中，它只会返回在 c_purchases 中有嵌套记录的 2 与所有 4 个玉米项目。

当前结果

{
    "hits": {
        "total": {
            "value": 2
        },
        "hits": [
            {
                "_source": {
                    "i_description": "corn 1",
                    "i_code": "111111"
                },
                "inner_hits": {
                    "c_purchases": {
                        "hits": {
                            "total": {
                                "value": 1,
                            },
                            "hits": [
                                {
                                    "_source": {
                                        "customernumber": "100",
                                        "lastordered": "2018-01-30T00:00:00",
                                    }
                                }
                            ]
                        }
                    }
                }
            },
            {
                "_source": {
                    "i_description": "corn 2",
                    "i_code": "222222"
                },
                "inner_hits": {
                    "c_purchases": {
                        "hits": {
                            "total": {
                                "value": 1,
                            },
                            "hits": [
                                {
                                    "_source": {
                                        "customernumber": "100",
                                        "lastordered": "2018-01-30T00:00:00",
                                    }
                                }
                            ]
                        }
                    }
                }
            }
        ]
    }
}

预期结果

{
    "hits": {
        "total": {
            "value": 2
        },
        "hits": [
            {
                "_source": {
                    "i_description": "corn 1",
                    "i_code": "111111"
                },
                "inner_hits": {
                    "c_purchases": {
                        "hits": {
                            "total": {
                                "value": 1,
                            },
                            "hits": [
                                {
                                    "_source": {
                                        "customernumber": "100",
                                        "lastordered": "2018-01-30T00:00:00",
                                    }
                                }
                            ]
                        }
                    }
                }
            },
            {
                "_source": {
                    "i_description": "corn 2",
                    "i_code": "222222"
                },
                "inner_hits": {
                    "c_purchases": {
                        "hits": {
                            "total": {
                                "value": 1,
                            },
                            "hits": [
                                {
                                    "_source": {
                                        "customernumber": "100",
                                        "lastordered": "2018-01-30T00:00:00",
                                    }
                                }
                            ]
                        }
                    }
                }
            },
            {
                "_source": {
                    "i_description": "corn 3",
                    "i_code": "333333"
                }
            },
            {
                "_source": {
                    "i_description": "corn 4",
                    "i_code": "444444"
                }
            }
        ]
    }
}

提前致谢！

Answer 1

通过搜索解决了这个问题：

{
    "_source": {
        "includes": ["*"],
        "excludes": ["c_purchases"]
    },
    "query":  {
        "bool": {
            "must": {
                "match": {
                    "i_description": "corn"
                }
            },
            "should": {
                "nested" : {
                    "path" : "c_purchases",
                    "inner_hits": {
                        "_source": ["*"]
                    },
                    "query" : {
                        "match": {
                            "c_purchases.customernumber": "1111"
                        }
                    }
                }   
            }
        }
    }
}