SQL as Report Writing tool

Question

Hi,
I am using Oracle 11.2.0.1 on windows xp. After google for report writing tool, i think if i am able to get the output by either SQL or PL/SQL; then i think i need not to learn those new report writing tools; i will continue read Oracle PL/SQL and/or SQL. Here it is my required output :
=======================================================================================================================================================================
S.No.  Customer Name & Address                <-------------------Product Sold Description------------------>            Qty               Rate             Amount
=======================================================================================================================================================================
1      Table1.CustomerName                    1.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                  
       Table1.Address1                        2.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                   
       Table1.Address2    <----Wrapable       3.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                   
       Table1.Address3                        4.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate     Table3.Net_Amt
=======================================================================================================================================================================
2      Table1.CustomerName                    1.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                  
       Table1.Address1                        2.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                   
       Table1.Address2    <----Wrapable       3.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                   
       Table1.Address3                        4.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                  
                                              5.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                  
                                              6.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                   
                                              7.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate                   
                                              8.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rate     Table3.Net_Amt
=======================================================================================================================================================================
3      Table1.CustomerName                    1.  Table2.Prod_Id     Table3.Product_Name     <-------Wrapable        Table3.Qty          Table3.Rat      Table3.Net_Amt
       Table1.Address1               
       Table1.Address2    <----Wrapable
       Table1.Address3                
=======================================================================================================================================================================
Wrapable Means If product name / Address2 are bigger than 30 characters, then they should be wrapped to next line.If you says me to provide create table and couple of insert statements, i will provide, i am just mentioning the tablename.column name for what i wish to get from SQL, rather than learning the report writing tool and this will at the later stage will definitely a best option in view of 3rd party report writing tool.
Kindly help me, how do i get above output (if you need create table and/or insert statements) please.
Thanks.

Answer 1

Hi,
user12050217 wrote:
create table table2
lineno number(4),
item_no varchar2(2),
eno varchar2(4),
product_id varchar2(7),
cust_id varchar2(5),
rate number(6,2),
qty number(5),
amount number(10,2),
remarks varchar2(10),
bill_amt number(10,2),
bill_date date,
bill_no varchar2(10));
insert into table2 values(1,'1',2145,'2145219','Z0105',17.43,143,2492.49,'no remark',2492.49,sysdate,'RF-K-00412');
insert into table2 values(2,'2',2145,'2145400','Z0105',8,50,400,'',400,sysdate-1,'RF-K-00378');There's no need to post all the columns in table2; just post the ones that play some role in this problem.
using SYSDATE in the sample data is asking for trouble. If people from different time zones are trying to help you, or if it takes more than one day to get the problem solved, then it's hard to compare results. In case bill_date matters in this problem, let's use this sample data for table2:
insert into table2 values(1,'1',2145,'2145219','Z0105',17.43,143,2492.49,'no remark',2492.49,DATE '2011-12-03' ,'RF-K-00412');
insert into table2 values(2,'2',2145,'2145400','Z0105',8,50,400,'',400,DATE '2011-12-03' -1,'RF-K-00378');
insert into table2 values(3,'1',2146,'2145100','K3125',10,50,500,'',500,DATE '2011-12-03' -2,'RF-K-00215');
insert into table2 values(4,'2',2146,'2145112','K3125',6.42,478,3068.76,'',3068.76,DATE '2011-12-03' -2,'RF-K-00215');
insert into table2 values(5,'3',2146,'2145115','K3125',6.42,478,3068.76,'',3068.76,DATE '2011-12-03' -2,'RF-K-00215');
insert into table2 values(6,'4',2146,'2145612','K3125',1340,10,13400,'',13400,DATE '2011-12-03' -2,'RF-K-00215');
insert into table2 values(7,'5',2146,'2145725','K3125',100,1,100,'',100,DATE '2011-12-03' -2,'RF-K-00215');I just replaced SYSDATE with DATE '2011-12-03'.
And here it is required Output :
====================================================================...
S.No.  Customer Name & Address                <-------------------Product Sold Description------------------>            Qty ...
======================================================================...
2145   Shri My Cust Name2                     1 2145219 Product Name                                                     143   ...
This is address2                       2 2145400 This may be a long product                                         8   ...
Address2                                         name which may wrap in
Address3                                         report
=======================================================================...
2146   My My Cust Name1                       1 2145100 Another long name in product, so please                           10   ...
This is address1                                 take care
This is long line of address which     2 2145112 Short Name                                                       478       ...
should be wrap                         3 2145115 Sulphar Mixer                                                    478           ...
4 2145612 One costly product name, may name not fit                         10     ...
in the report
5 2145725 Silver Coating                                                     1              100.00            1502.84
=======================================================================================================================================================================The above data is just for example and i have created required output manually, so there may be some typo.How would you feel if someone posted a solution full of bugs and said "this solution is just for example ... so there may be some typo."?
Why is the product description shorter for s.no.=2145? In s.no., there's room for this much of the descripotion on one line
One costly product name, may name not fit In s.no=2145, why can't you have:
This may be a long product name which may Why do you have to start a new line after "product"? Is this one of the typos you mentioned?
... I think to get the output, i will need function which will take address or product name as an input paramter and will return 2 or 3 lines in the select statement, but what should be code of that function, how do i call in select statement, etc. i wish to learn all these things.I agree. A function to split the text into lines of the right length would be very handy. Here's a pipelined function to do that:
CREATE OR REPLACE PACKAGE     word_wrap
AS
TYPE     ww_row     IS RECORD ( r_num     NUMBER
                    , r_txt     VARCHAR2 (4000)
TYPE     ww_tbl  IS TABLE OF ww_row;
FUNCTION     wrap
(     in_txt  IN      VARCHAR2     -- Text to be wrapped
,     in_len     IN     PLS_INTEGER     -- Maximum line length
RETURN  ww_tbl
PIPELINED;
END     word_wrap;
SHOW ERRORS
CREATE OR REPLACE PACKAGE BODY     word_wrap
AS
--          **   w r a p   **
--          wrap splits in_txt into several row2 consisting of r_num (1, 2, 3, ...),
--          and r_txt, up to in_len characters from in_txt.  if possible, the last
--          character in r_txt will be a "nonword" character, such as whitespace or
--          punctuation.  If words are longer than in_len, however, it may have
--          to break a word in the middle.
--          Example:
--              wrap ('I am the monarch or the sea', 10)
--          returns these 3 rows:
--              1     I am the
--              2     monarch
--              3     of the sea
--          On rows 1 and 2, r_txt ends with a space.
--          If we change the last argument to 6:
--              wrap ('I am the monarch or the sea', 6)
--          it returns these 6 rows:
--              1     I am
--              2     the
--              3       monarc
--              4       h of
--              5     the
--              6       sea
FUNCTION     wrap
(     in_txt  IN      VARCHAR2     -- Text to be wrapped
,     in_len     IN     PLS_INTEGER     -- Maximum line length
RETURN  ww_tbl
PIPELINED
IS
    break_pos          PLS_INTEGER;               -- Position of next break character in remaining_txt
    line_len          PLS_INTEGER     := GREATEST ( NVL ( in_len
                                                  , 1
                                  , 1
                                  );     -- Actual max line length (in case a bad in_len was passed)
    remaining_txt     VARCHAR2 (4000) := in_txt;     -- Part of in_txt that hasn't been output yet
    return_row          ww_row;                        -- Row to be returned
    r_num          PLS_INTEGER     := 1;          -- Number of next row
BEGIN
    WHILE  remaining_txt     IS NOT NULL
    AND        r_num          >= 1
    AND        r_num          < 10
    LOOP
        return_row.r_num := r_num;
        IF  remaining_txt          IS NULL
     OR  LENGTH (remaining_txt)     <= line_len
     THEN     -- Everything fits on this row
         return_row.r_txt := remaining_txt;
         remaining_txt := NULL;
     ELSE     -- Not everything fits, so we need to split
         break_pos := REGEXP_INSTR ( SUBSTR (remaining_txt, 1, line_len)
                                     , '\W\w*$'     -- non-word, follewd by any number of word characters, then end
         IF  break_pos = 0
         THEN    --  No good place to break before line_len
             break_pos := line_len;
         END IF;
         return_row.r_txt := SUBSTR (remaining_txt, 1, break_pos);
         remaining_txt := SUBSTR (remaining_txt, 1 + break_pos);
     END IF;
        PIPE ROW (return_row);
     r_num := r_num + 1;
    END LOOP;
    RETURN;
END     wrap;
END     word_wrap;
SHOW ERRORSYou can use the fucntion with your sample data like this:
SELECT       t3.product_id
,       ww.r_txt
,       ww.r_num
FROM          table3     t3
CROSS JOIN   TABLE ( word_wrap.wrap ( t3.description
                                        , 42
             )     ww
ORDER BY  t3.product_id
,            ww.r_num
;Output:
PRODUCT R_TXT                                         R_NUM
2145100 Another long name in product, so please           1
2145100 take care                                         2
2145112 Short Name                                        1
2145115 Sulphar Mixer                                     1
2145219 Product Name                                      1
2145400 This may be a long product name which may         1
2145400 wrap in report                                    2
2145612 One costly product name, may name not fit         1
2145612 in the report                                     2
2145725 Silver Coating                                    1For the report you want, I suggest you start with the products. Modify the query above so that it produces something like this:
D_NUM ENO  PRODUCT R_TXT
    1 2145 2145219 Product Name
    2 2145 2145400 This may be a long product name which may
    3 2145         wrap in report
    1 2146 2145100 Another long name in product, so please
    2 2146         take care
    3 2146 2145112 Short Name
    4 2146 2145115 Sulphar Mixer
    5 2146 2145612 One costly product name, may name not fit
    6 2146         in the report
    7 2146 2145725 Silver CoatingNotice the d_num column at the beginning. It starts with 1 for each eno, and numbers each line in order. You can use the analytic ROW_NUMBER fucntion to produce this column. You won't need to display d_num in the final output, but you will need it for the next step in the solution, which is:
Do something similar for the addresses. Since you have Oracle 11, you can use the SELECT ... UNPIVOT feature to get all the address lines into one column, then use the wrap function to break down the long lines, and number them. Again, use ROW_NUMBER to produce d_num. (You might consider permanently storing the addresses in the unpivoted form, with just one address line per row. Since this will create a one-to-many relationship between customers and addresses, you'll need a separte table for the addresses.)
Finally, do a full outer join to combine the result sets of the two steps above, using the customer id and d_num to match rows. This is an example of a Prix Fixe Query . The page I refereneced earlier has a more detailed description and example.
Edited by: Frank Kulash on Dec 5, 2011 7:04 AM
Added pacakge spec.