Order of join operations: would these two FROM clauses produce the same results? [closed]

Question

This question is unlikely to help any future visitors; it is only relevant to a small geographic area, a specific moment in time, or an extraordinarily narrow situation that is not generally applicable to the worldwide audience of the internet. For help making this question more broadly applicable, visit the help center.

Closed 9 years ago.

Would these two FROM clauses produce the same results? And if not, is there a way to write the the first one so that no parenthesis are needed?

FROM            SALESTAX
     RIGHT JOIN (            ITEMS
                  RIGHT JOIN (            PINVOICE
                               INNER JOIN PINVDET ON PINVOICE.PNV_INVOICENO = PINVDET.PND_INVOICENO AND PINVOICE.PNV_Site = PINVDET.PND_Site
                             ) ON ITEMS.ITE_INVNO = PINVDET.PND_INVNO
                ) ON SALESTAX.STX_GroupID = PINVDET.PND_TAX1
      FULL JOIN (            CUSTMS
                  RIGHT JOIN CUSMER ON CUSTMS.TMS_CODE = CUSMER.CUS_TERM
                ) ON PINVDET.PND_CUSTID = CUSMER.CUS_CustID

FROM  CUSTMS RIGHT JOIN
      CUSMER ON TMS_CODE = CUS_TERM FULL JOIN
      PINVDET ON PND_CUSTID = CUS_CustID LEFT JOIN
      PINVOICE ON PNV_INVOICENO = PND_INVOICENO AND PNV_Site = PND_Site LEFT JOIN
      SALESTAX on STX_GROUPID = PND_TAX1 left join
      ITEMS on ITE_INVNO = PND_INVNO

EDIT: While I'd like to know the answer to the first question, I'm more immediately interested in just having a more straightforward version of the first FROM clause that doesn't need parenthesis so if you'd rather just rewrite it than compare the two then feel free to just do that.

Answer 1

I have no idea if the 1st is equivalent to the 2nd (first because the queries are unfriendly formatted, to say the least and second because RIGHT joins are kind of confusing, since many are used to write using LEFT joins.) But to answer the question:

Is there a way to write the the first one so that no parenthesis are needed?

Yes, you can simply remove the parentheses from the 1st query.

Keeping the parentheses and formatted with some white space:

FROM 
        SALESTAX
    RIGHT JOIN 
        (   ITEMS
        RIGHT JOIN 
            (   PINVOICE
            INNER JOIN 
                PINVDET 
              ON  PINVOICE.PNV_INVOICENO = PINVDET.PND_INVOICENO 
              AND PINVOICE.PNV_Site = PINVDET.PND_Site
            ) 
          ON ITEMS.ITE_INVNO = PINVDET.PND_INVNO
        ) 
      ON SALESTAX.STX_GroupID = PINVDET.PND_TAX1
    FULL JOIN 
        (   CUSTMS
        RIGHT JOIN
            CUSMER 
          ON CUSTMS.TMS_CODE = CUSMER.CUS_TERM
        ) 
      ON PINVDET.PND_CUSTID = CUSMER.CUS_CustID

Without parentheses and white-space formatted:

FROM    
        SALESTAX
    RIGHT JOIN
            ITEMS
        RIGHT JOIN
                PINVOICE
            INNER JOIN 
                PINVDET 
              ON  PINVOICE.PNV_INVOICENO = PINVDET.PND_INVOICENO
              AND PINVOICE.PNV_Site = PINVDET.PND_Site
          ON ITEMS.ITE_INVNO = PINVDET.PND_INVNO
      ON SALESTAX.STX_GroupID = PINVDET.PND_TAX1
    FULL JOIN
            CUSTMS
        RIGHT JOIN 
            CUSMER 
          ON CUSTMS.TMS_CODE = CUSMER.CUS_TERM
      ON PINVDET.PND_CUSTID = CUSMER.CUS_CustID

---

To answer the other question, about the 2nd query, no it isn't equivalent. You missed the table aliases and changed an inner join to left join. This is equivalent to the 1st:

FROM  CUSMER  
        LEFT JOIN
      CUSTMS     ON  CUSTMS.TMS_CODE = CUSMER.CUS_TERM
   FULL JOIN
      PINVDET
        INNER JOIN                                      -- this is changed
      PINVOICE   ON  PINVOICE.PNV_INVOICENO = PINVDET.PND_INVOICENO
                 AND PINVOICE.PNV_Site = PINVDET.PND_Site
        LEFT JOIN
      ITEMS      ON  ITEMS.ITE_INVNO = PINVDET.PND_INVNO
        LEFT JOIN
      SALESTAX   ON  SALESTAX.STX_GroupID = PINVDET.PND_TAX1
   ON PINVDET.PND_CUSTID = CUSMER.CUS_CustID

Answer 2

I encourage you to write your from clauses by placing all left joins first followed by inner joins. This greatly simplifies trying to figure out what queries are doing. A series of left joins says "keep all the rows in the first table". A series of inner joins says "keep only rows where there are matches between tables." (Occasionally, you might need subqueries as in the first example.)

It is unlikely in this example that the two are the same. In the first, the full join is the "outermost" join. In the second, the full join is embedded in a series of joins. These are interpreted sequentially from the first to the last. One of these is probably converting the full join to an inner join or left join. Of course, the two could produce equivalent results if all the tables match.

The second example could probably be written as:

FROM  CUSMER  LEFT JOIN
      CUSTMS ON TMS_CODE = CUS_TERM LEFT JOIN
      PINVDET ON PND_CUSTID = CUS_CustID LEFT JOIN
      PINVOICE ON PNV_INVOICENO = PND_INVOICENO AND PNV_Site = PND_Site LEFT JOIN
      SALESTAX on STX_GROUPID = PND_TAX1 left join
      ITEMS on ITE_INVNO = PND_INVNO

(assuming that something after the full join is converting it to a left join anyway).