MongoDB在某些方面確實比關係型數據庫更強 (比如對追加型日誌數據的吞吐能力),但結構化計算能力方面較弱。比如:MongoDB不支持子查詢,碰到這些複雜的運算就只能先將數據讀出後再計算,而用Java等語言編寫這類計算也不是很簡單。
比如要處理這麼個場景:查出訂單信息,要求訂單中的SELLERID必須是employee集合中STATE= California的員工id。如果寫成sql就是:
Select * from orders where orders.sellerid in (select eid from employee where employee.state=’California’)。
orders的數據量較大無法一次取出,employee的數據量較小,最終結果的數據量也較小。部分數據如下:
MongoDB Collection orders:
… { "_id" : ObjectId("5434f88dd00ab5276493e270"), "ORDERID" : 1, "CLIENT" : "UJRNP ","SELLERID" : 17, "AMOUNT" : 392, "ORDERDATE" : "2008/11/2 15:28" } { "_id" : ObjectId("5434f88dd00ab5276493e271"), "ORDERID" : 2, "CLIENT" : "SJCH" , "SELLERID" : 6, "AMOUNT" : 4802, "ORDERDATE" : "2008/11/9 15:28" } { "_id" : ObjectId("5434f88dd00ab5276493e272"), "ORDERID" : 3, "CLIENT" : "UJRNP ","SELLERID" : 16, "AMOUNT" : 13500, "ORDERDATE" : "2008/11/5 15:28" } { "_id" : ObjectId("5434f88dd00ab5276493e273"), "ORDERID" : 4, "CLIENT" : "PWQ", "SELLERID" : 9, "AMOUNT" : 26100, "ORDERDATE" : "2008/11/8 15:28" } … |
MongoDB Collection employee:
… { "_id" : ObjectId("5437413513bdf2a4048f3480"), "EID" : 1, "NAME" : "Rebecca", " SURNAME" : "Moore", "GENDER" : "F", "STATE" : "California", "BIRTHDAY" : "1974-1 1-20","HIREDATE" : "2005-03-11", "DEPT" : "R&D", "SALARY" : 7000 } { "_id" : ObjectId("5437413513bdf2a4048f3481"), "EID" : 2, "NAME" : "Ashley", "S URNAME" : "Wilson", "GENDER" : "F", "STATE" : "New York", "BIRTHDAY" : "1980-07- 19","HIREDATE" : "2008-03-16", "DEPT" : "Finance", "SALARY" : 11000 } { "_id" : ObjectId("5437413513bdf2a4048f3482"), "EID" : 3, "NAME" : "Rachel", "S URNAME" : "Johnson", "GENDER" : "F", "STATE" : "New Mexico", "BIRTHDAY" : "1970- 12-17","HIREDATE" : "2010-12-01", "DEPT" : "Sales", "SALARY" : 9000 } … |
計算結果:
如果有集算器的協助就簡單多了,它是專門爲結構化、半結構化數據計算設計的編程語言,查詢、子查詢、分組後過濾等等都有函數類庫可以用,能夠很好的彌補MongoDB在結構化方面計算能力的不足。比如上面問題,4行就搞定了:
|
A |
1 |
=mongo_open("mongodb://localhost:27017/test?user=test&password=test") |
2 |
=mongo_shell(A1,"orders.find(,{_id:0})") |
3 |
=mongo_shell@x(A1,"employee.find({STATE:'California'},{_id:0})").fetch() |
4 |
=A2.select(A3.(EID).sort().pos@b(SELLERID)).fetch() |
其實還有很多情況用MongoDB處理結構化計算不太方便,但有集算器SPL的輔助卻很簡單,感興趣可以參考:玩轉Mongo計算、簡化MongoDB關聯運算、輔助MongoDB計算
集算器還很容易嵌入到Java應用程序中,Java如何調用SPL腳本有使用和獲得它的方法。
關於集算器安裝使用、獲得免費授權和相關技術資料,可以參見如何使用集算器。