Elasticsearch 使用terms对long类型日期统计按月销售

索引mapping：

{"properties" : {"_class" : {"type" : "text","fields" : {"keyword" : {"type" : "keyword","ignore_above" : 256}}},"createDate" : {"type" : "long"},"goodsCode" : {"type" : "keyword"},"id" : {"type" : "text","fields" : {"keyword" : {"type" : "keyword","ignore_above" : 256}}},"orderNumber" : {"type" : "keyword"},"payDate" : {"type" : "long"},"saleQty" : {"type" : "double"}}}

payDate字段映射为long类型，在使用date_histogram统计时，统计不出想要的效果。

{..."aggs": {"agg_name": {"date_histogram": {"field": "payDate","interval": "1m","format": "yyyy-MM","min_doc_count": 0}}}
}

结果输出的是"yyyy-MM1577836800000"，并不是期待的2023-06.

{"key_as_string": "yyyy-MM1577836800000","key": 1577836800000,"doc_count": 3
}

改用terms+script+子聚合方式统计：

{"query": {"range": {"payDate": {"gte": 1685548800000}}},"size":0, "aggs": {"month_sales":{"terms": {"script": {"source": """DateTimeFormatter df = DateTimeFormatter.ofPattern("yyyy-MM").withZone(ZoneId.systemDefault());return df.format(Instant.ofEpochMilli(doc['payDate'].value));""","lang":"painless"},"order": {"_key": "asc"}},"aggs": {"sale_sum": {"sum": {"field": "saleQty"}}}}}
}

script：使用脚本把pay格式化成yyyy-MM的字符串。
order：根据日期排序
sale_sum：统计每个桶里面商品销售数量

输出结果：

{"month_sales": {"doc_count_error_upper_bound": 0,"sum_other_doc_count": 9723368,"buckets": [{"key": "2023-06","doc_count": 49,"sale_sum": {"value": 77.0}},{"key": "2023-07","doc_count": 217,"sale_sum": {"value": 40012.0}},{"key": "2023-08","doc_count": 2187911,"sale_sum": {"value": 3486633.0}},{"key": "2023-09","doc_count": 2746235,"sale_sum": {"value": 4556942.0}},{"key": "2023-10","doc_count": 2346896,"sale_sum": {"value": 4097690.0}},{"key": "2023-11","doc_count": 2442224,"sale_sum": {"value": 4120708.0}},{"key": "2023-12","doc_count": 2359840,"sale_sum": {"value": 4167825.0}},{"key": "2024-01","doc_count": 3619202,"sale_sum": {"value": 6506941.0}},{"key": "2024-02","doc_count": 2673919,"sale_sum": {"value": 5066871.0}},{"key": "2024-03","doc_count": 3078439,"sale_sum": {"value": 4810836.0}}]}
}