How to Unpivot, Crosstab, or Pivot using Oracle 9i with PL/SQL?

Question

How to Unpivot, Crosstab, or Pivot using Oracle 9i with PL/SQL?
Here is a fictional sample layout of the data I have from My_Source_Query:
Customer | VIN | Year | Make | Odometer | ... followed by 350 more columns/fields
123 | 321XYZ | 2012 | Honda | 1900 |
123 | 432ABC | 2012 | Toyota | 2300 |
456 | 999PDQ | 2000 | Ford | 45586 |
876 | 888QWE | 2010 | Mercedes | 38332 |
... followed by up to 25 more rows of data from this query.
The exact number of records returned by My_Source_Query is unknown ahead of time, but should be less than 25 even under extreme situations.
Here is how I would like the data to be:
Column1 |Column2 |Column3 |Column4 |Column5 |
Customer | 123 | 123 | 456 | 876 |
VIN | 321XYZ | 432ABC | 999PDQ | 888QWE |
Year | 2012 | 2012 | 2000 | 2010 |
Make | Honda | Toyota | Ford | Mercedes|
Odometer | 1900 | 2300 | 45586 | 38332 |
... followed by 350 more rows with the names of the columns/fields from the My_Source_Query.
From reading and trying many, many, many of the posting on this topic I understand that the unknown number or rows in My_Source_Query can be a problem and have considered working with one row at a time until each row has been converted to a column.
If possible I'd like to find a way of doing this conversion from rows to columns using a query instead of scripts if that is possible. I am a novice at this so any help is welcome.
This is a repost. I originally posted this question to the wrong forum. Sorry about that.

Answer 1

The permission level that I have in the Oracle environment is 'read only'. This is also be the permission level of the users of the query I am trying to build.
As requested, here is the 'create' SQL to build a simple table that has the type of data I am working with.
My real select query will have more than 350 columns and the rows returned will be 25 rows of less, but for now I am prototyping with just seven columns that have the different data types noted in my sample data.
NOTE: This SQL has been written and tested in MS Access since I do not have permission to create and populate a table in the Oracle environment and ODBC connections are not allowed.
CREATE TABLE tbl_MyDataSource
(Customer char(50),
VIN char(50),
Year char(50),
Make char(50),
Odometer long,
InvDate date,
Amount currency)
Here is the 'insert into' to populate the tbl_MyDataSource table with four sample records.
INSERT INTO tbl_MyDataSource ( Customer, VIN, [Year], Make, Odometer, InvDate, Amount )
SELECT "123", "321XYZ", "2012", "Honda", "1900", "2/15/2012", "987";
INSERT INTO tbl_MyDataSource ( Customer, VIN, [Year], Make, Odometer, InvDate, Amount )
VALUES ("123", "432ABC", "2012", "Toyota", "2300", "1/10/2012", "6546");
INSERT INTO tbl_MyDataSource ( Customer, VIN, [Year], Make, Odometer, InvDate, Amount )
VALUES ("456", "999PDQ", "2000", "Ford", "45586", "4/25/2002", "456");
INSERT INTO tbl_MyDataSource ( Customer, VIN, [Year], Make, Odometer, InvDate, Amount )
VALUES ("876", "888QWE", "2010", "Mercedes", "38332", "10/13/2010", "15973");
Which should produce a table containing these columns with these values:
tbl_MyDataSource:
Customer     VIN     Year     Make     Odometer     InvDate          Amount
123 | 321XYZ | 2012 | Honda      | 1900          | 2/15/2012     | 987.00
123 | 432ABC | 2012 | Toyota | 2300 | 1/10/2012     | 6,546.00
456 | 999PDQ | 2000 | Ford     | 45586          | 4/25/2002     | 456.00
876 | 888QWE | 2010 | Mercedes | 38332          | 10/13/2010     | 15,973.00
The desired result is to use Oracle 9i to convert the columns into rows using sql without using any scripts if possible.
qsel_MyResults:
Column1          Column2          Column3          Column4          Column5
Customer | 123 | 123 | 456 | 876
VIN | 321XYZ | 432ABC | 999PDQ | 888QWE
Year | 2012 | 2012 | 2000 | 2010
Make | Honda | Toyota | Ford | Mercedes
Odometer | 1900 | 2300 | 45586 | 38332
InvDate | 2/15/2012 | 1/10/2012 | 4/25/2002 | 10/13/2010
Amount | 987.00 | 6,546.00 | 456.00 | 15,973.00
The syntax in SQL is something I am not yet sure of.
You said:
>
"Don't use the same name or alias for two different things. if you have a table called t, then don't use t as an alais for an in-line view. Pick a different name, like ordered_t, instead.">
but I'm not clear on which part of the SQL you are suggesting I change. The code I posted is something I pieced together from some of the other postings and is not something I full understand the syntax of.
Here is my latest (failed) attempt at this.
select *
  from (select * from tbl_MyDataSource) t;
with data as
(select rownum rnum, t.* from (select * from t order by c1) ordered_t), -- changed 't' to 'ordered_t'
rows_to_have as
(select level rr from dual connect by level <= 7 -- number of columns in T
select rnum,
       max(decode(rr, 1, c1)),
       max(decode(rr, 2, c2)),
       max(decode(rr, 3, c3)),
       max(decode(rr, 4, c3)),      
       max(decode(rr, 5, c3)),      
       max(decode(rr, 6, c3)),      
       max(decode(rr, 7, c3)),       
  from data, rows_to_have
group by rnumIn the above code the "select * from tbl_MyDataSource" is a place holder for my select query which runs without error and has these exact number of fields and data types as order shown in the tbl_MyDataSource above.
This code produces the error 'ORA-00936: missing expression'. The error appears to be starting with the 'with data as' line if I am reading my PL/Sql window correctly. Everything above that row runs without error.
Thank you for your great patients and for sharing your considerable depth of knowledge. Any help is gratefully welcomed.