Exclude a column using SELECT * [except columnA] FROM tableA?

Question

We all know that to select all columns from a table, we can use

SELECT * FROM tableA

Is there a way to exclude column(s) from a table without specifying all the columns?

SELECT * [except columnA] FROM tableA

The only way that I know is to manually specify all the columns and exclude the unwanted column. This is really time consuming so I'm looking for ways to save time and effort on this, as well as future maintenance should the table has more/less columns.

Answer 1

You can try it this way:

/* Get the data into a temp table */
SELECT * INTO #TempTable
FROM YourTable
/* Drop the columns that are not needed */
ALTER TABLE #TempTable
DROP COLUMN ColumnToDrop
/* Get results and drop temp table */
SELECT * FROM #TempTable
DROP TABLE #TempTable

Answer 2

No.

Maintenance-light best practice is to specify only the required columns.

At least 2 reasons:

• This makes your contract between client and database stable. Same data, every time
• Performance, covering indexes

Edit (July 2011):

If you drag from Object Explorer the Columns node for a table, it puts a CSV list of columns in the Query Window for you which achieves one of your goals

Answer 3

A modern SQL dialect like BigQuery, Databricks proposes an excellent solution.

SELECT * EXCEPT(ColumnNameX, [ColumnNameY, ...])
FROM TableA

This is a very powerful SQL syntax to avoid a long list of columns that must be constantly updated due to table column name changes. And this functionality is missing in the current SQL Server implementation, which is a pity. Hopefully, this will someday become SQL standard and make it more data scientist-friendly.

Data scientists like a quick option to shorten a query and remove some columns (due to duplication or any other reason).

https://cloud.google.com/bigquery/docs/reference/standard-sql/query-syntax#select-modifiers

https://docs.databricks.com/sql/language-manual/sql-ref-syntax-qry-select.html#syntax

DuckDB uses EXCLUDE instead EXCEPT https://duckdb.org/docs/sql/query_syntax/select

Answer 4

If you don't want to write each column name manually you can use Script Table As by right clicking on table or view in SSMS like this:

Then you will get whole select query in New Query Editor Window then remove unwanted column like this:

Done

Answer 5

The automated way to do this in SQL (SQL Server) is:

declare @cols varchar(max), @query varchar(max);
SELECT  @cols = STUFF
    (
        ( 
            SELECT DISTINCT '], [' + name
            FROM sys.columns
            where object_id = (
                select top 1 object_id from sys.objects
                where name = 'MyTable'
            )
            and name not in ('ColumnIDontWant1', 'ColumnIDontWant2')
            FOR XML PATH('')
        ), 1, 2, ''
    ) + ']';

SELECT @query = 'select ' + @cols + ' from MyTable';  
EXEC (@query);

Answer 6

You could create a view that has the columns you wish to select, then you can just select * from the view...

Answer 7

Yes it's possible (but not recommended).

CREATE TABLE contact (contactid int, name varchar(100), dob datetime)
INSERT INTO contact SELECT 1, 'Joe', '1974-01-01'

DECLARE @columns varchar(8000)

SELECT @columns = ISNULL(@columns + ', ','') + QUOTENAME(column_name)
FROM INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME = 'contact' AND COLUMN_NAME <> 'dob'
ORDER BY ORDINAL_POSITION

EXEC ('SELECT ' + @columns + ' FROM contact')

Explanation of the code:

1. Declare a variable to store a comma separated list of column names. This defaults to NULL.
2. Use a system view to determine the names of the columns in our table.
3. Use SELECT @variable = @variable + ... FROM to concatenate the column names. This type of SELECT does not not return a result set. This is perhaps undocumented behaviour but works in every version of SQL Server. As an alternative you could use SET @variable = (SELECT ... FOR XML PATH('')) to concatenate strings.
4. Use the ISNULL function to prepend a comma only if this is not the first column name. Use the QUOTENAME function to support spaces and punctuation in column names.
5. Use the WHERE clause to hide columns we don't want to see.
6. Use EXEC (@variable), also known as dynamic SQL, to resolve the column names at runtime. This is needed because we don't know the column names at compile time.

Answer 8

Like the others have said there is no way to do this, but if you're using Sql Server a trick that I use is to change the output to comma separated, then do

select top 1 * from table

and cut the whole list of columns from the output window. Then you can choose which columns you want without having to type them all in.

Answer 9

Basically, you cannot do what you would like - but you can get the right tools to help you out making things a bit easier.

If you look at Red-Gate's SQL Prompt, you can type "SELECT * FROM MyTable", and then move the cursor back after the "*", and hit <TAB> to expand the list of fields, and remove those few fields you don't need.

It's not a perfect solution - but a darn good one! :-) Too bad MS SQL Server Management Studio's Intellisense still isn't intelligent enough to offer this feature.......

Marc

Answer 10

If you want to exclude a sensitive case column like the password for example, I do this to hide the value :

SELECT * , "" as password FROM tableName;

Answer 11

DECLARE @SQL VARCHAR(max), @TableName sysname = 'YourTableName'

SELECT @SQL = COALESCE(@SQL + ', ', '') + Name 
FROM sys.columns
WHERE OBJECT_ID = OBJECT_ID(@TableName)
AND name NOT IN ('Not This', 'Or that');

SELECT @SQL = 'SELECT ' + @SQL + ' FROM ' + @TableName

EXEC (@SQL)

UPDATE:

You can also create a stored procedure to take care of this task if you use it more often. In this example I have used the built in STRING_SPLIT() which is available on SQL Server 2016+, but if you need there are pleanty of examples of how to create it manually on SO.

CREATE PROCEDURE [usp_select_without]
@schema_name sysname = N'dbo',
@table_name sysname,
@list_of_columns_excluded nvarchar(max),
@separator nchar(1) = N','
AS
BEGIN
 DECLARE 
 @SQL nvarchar(max),
 @full_table_name nvarchar(max) = CONCAT(@schema_name, N'.', @table_name);

 SELECT @SQL = COALESCE(@SQL + ', ', '') + QUOTENAME([Name])
 FROM sys.columns sc
 LEFT JOIN STRING_SPLIT(@list_of_columns_excluded, @separator) ss ON sc.[name] = ss.[value]
 WHERE sc.OBJECT_ID = OBJECT_ID(@full_table_name, N'u')
 AND ss.[value] IS NULL;

 SELECT @SQL = N'SELECT ' + @SQL + N' FROM ' + @full_table_name;
 EXEC(@SQL)
END

And then just:

EXEC [usp_select_without] 
@table_name = N'Test_Table',
@list_of_columns_excluded = N'ID, Date, Name';

Answer 12

no there is no way to do this. maybe you can create custom views if that's feasible in your situation

EDIT May be if your DB supports execution of dynamic sql u could write an SP and pass the columns u don't want to see to it and let it create the query dynamically and return the result to you. I think this is doable in SQL Server atleast

Answer 13

If you are using SQL Server Management Studio then do as follows:

1. Type in your desired tables name and select it
2. Press Alt+F1
3. o/p shows the columns in table.
4. Select the desired columns
5. Copy & paste those in your select query
6. Fire the query.

Enjoy.

Answer 14

Postgres sql has a way of doing it

pls refer: http://www.postgresonline.com/journal/archives/41-How-to-SELECT-ALL-EXCEPT-some-columns-in-a-table.html

The Information Schema Hack Way

SELECT 'SELECT ' || array_to_string(ARRAY(SELECT 'o' || '.' || c.column_name
        FROM information_schema.columns As c
            WHERE table_name = 'officepark' 
            AND  c.column_name NOT IN('officeparkid', 'contractor')
    ), ',') || ' FROM officepark As o' As sqlstmt

The above for my particular example table - generates an sql statement that looks like this

SELECT o.officepark,o.owner,o.squarefootage FROM officepark As o

Answer 15

Actually snowflake just released exclude so now you'd just:

SELECT * EXCLUDE [columnA,columnB,...] FROM tableA

Answer 16

In SQL Management Studio you can expand the columns in Object Explorer, then drag the Columns tree item into a query window to get a comma separated list of columns.

Answer 17

In summary you cannot do it, but I disagree with all of the comment above, there "are" scenarios where you can legitimately use a * When you create a nested query in order to select a specific range out of a whole list (such as paging) why in the world would want to specify each column on the outer select statement when you have done it in the inner?

Answer 18

If we are talking of Procedures, it works with this trick to generate a new query and EXECUTE IMMEDIATE it:

SELECT LISTAGG((column_name), ', ') WITHIN GROUP (ORDER BY column_id)
INTO var_list_of_columns
FROM ALL_TAB_COLUMNS
WHERE table_name = 'PUT_HERE_YOUR_TABLE'
AND column_name NOT IN ('dont_want_this_column','neither_this_one','etc_column');

Answer 19

Is there a way to exclude column(s) from a table without specifying all the columns?

Using declarative SQL in the usual way, no.

I think your proposed syntax is worthy and good. In fact, the relational database language 'Tutorial D' has a very similar syntax where the keywords ALL BUT are followed by a set of attributes (columns).

However, SQL's SELECT * already gets a lot a flak (@Guffa's answer here is a typical objection), so I don't think SELECT ALL BUT will get into the SQL Standard anytime soon.

I think the best 'work around' is to create a VIEW with only the columns you desire then SELECT * FROM ThatView.

Answer 20

The best way to solve this is using view you can create view with required columns and retrieve data form it

example

mysql> SELECT * FROM calls;
+----+------------+---------+
| id | date       | user_id |
+----+------------+---------+
|  1 | 2016-06-22 |       1 |
|  2 | 2016-06-22 |    NULL |
|  3 | 2016-06-22 |    NULL |
|  4 | 2016-06-23 |       2 |
|  5 | 2016-06-23 |       1 |
|  6 | 2016-06-23 |       1 |
|  7 | 2016-06-23 |    NULL |
+----+------------+---------+
7 rows in set (0.06 sec)

mysql> CREATE VIEW C_VIEW AS
    ->     SELECT id,date from calls;
Query OK, 0 rows affected (0.20 sec)

mysql> select * from C_VIEW;
+----+------------+
| id | date       |
+----+------------+
|  1 | 2016-06-22 |
|  2 | 2016-06-22 |
|  3 | 2016-06-22 |
|  4 | 2016-06-23 |
|  5 | 2016-06-23 |
|  6 | 2016-06-23 |
|  7 | 2016-06-23 |
+----+------------+
7 rows in set (0.00 sec)

Answer 21

I do not know of any database that supports this (SQL Server, MySQL, Oracle, PostgreSQL). It is definitely not part of the SQL standards so I think you have to specify only the columns you want.

You could of course build your SQL statement dynamically and have the server execute it. But this opens up the possibility for SQL injection..

Answer 22

Right click table in Object Explorer, Select top 1000 rows

It'll list all columns and not *. Then remove the unwanted column(s). Should be much faster than typing it yourself.

Then when you feel this is a bit too much work, get Red Gate's SQL Prompt, and type ssf from tbl, go to the * and click tab again.

Answer 23

I know this is a little old, but I had just run into the same issue and was looking for an answer. Then I had a senior developer show me a very simple trick.

If you are using the management studio query editor, expand the database, then expand the table that you are selecting from so that you can see the columns folder.

In your select statement, just highlight the referenced columns folder above and drag and drop it into the query window. It will paste all of the columns of the table, then just simply remove the identity column from the list of columns...

Answer 24

A colleage advised a good alternative:

• Do SELECT INTO in your preceding query (where you generate or get the data from) into a table (which you will delete when done). This will create the structure for you.
• Do a script as CREATE to new query window.
• Remove the unwanted columns. Format the remaining columns into a 1 liner and paste as your column list.
• Delete the table you created.

Done...

This helped us a lot.

Answer 25

Well, it is a common best practice to specify which columns you want, instead of just specifying *. So you should just state which fields you want your select to return.

Answer 26

That what I use often for this case:

declare @colnames varchar(max)=''

select @colnames=@colnames+','+name from syscolumns where object_id(tablename)=id and name not in (column3,column4)

SET @colnames=RIGHT(@colnames,LEN(@colnames)-1)

@colnames looks like column1,column2,column5

Answer 27

I did it like this and it works just fine (version 5.5.41):

# prepare column list using info from a table of choice
SET @dyn_colums = (SELECT REPLACE(
GROUP_CONCAT(`COLUMN_NAME`), ',column_name_to_remove','') 
FROM `INFORMATION_SCHEMA`.`COLUMNS` WHERE 
`TABLE_SCHEMA`='database_name' AND `TABLE_NAME`='table_name');

# set sql command using prepared columns
SET @sql = CONCAT("SELECT ", @dyn_colums, " FROM table_name");

# prepare and execute
PREPARE statement FROM @sql;
EXECUTE statement;

Answer 28

Sometimes the same program must handle different database stuctures. So I could not use a column list in the program to avoid errors in select statements.

* gives me all the optional fields. I check if the fields exist in the data table before use. This is my reason for using * in select.

This is how I handle excluded fields:

Dim da As New SqlDataAdapter("select * from table", cn)
da.FillSchema(dt, SchemaType.Source)
Dim fieldlist As String = ""
For Each DC As DataColumn In DT.Columns
   If DC.ColumnName.ToLower <> excludefield Then
    fieldlist = fieldlist &  DC.Columnname & ","
   End If
  Next

Answer 29

In Hive Sql you can do this:

set hive.support.quoted.identifiers=none;
select 
    `(unwanted_col1|unwanted_col2|unwanted_col3)?+.+`
from database.table

this gives you the rest cols

Answer 30

The proposed answer (stored procedure) from BartoszX didn't work for me when using a view instead of a real table.

Credit for the idea and the code below (except for my fix) belongs to BartoszX.

In order that this works for tables as well as for views, use the following code:

SET ANSI_NULLS ON
GO

SET QUOTED_IDENTIFIER ON
GO

CREATE PROCEDURE [dbo].[select_without]
@schema_name sysname = N'dbo',
@table_name sysname,
@list_of_columns_excluded nvarchar(max),
@separator nchar(1) = N','
AS
BEGIN
 DECLARE 
 @SQL nvarchar(max),
 @full_table_name nvarchar(max) = CONCAT(@schema_name, N'.', @table_name);

 SELECT @SQL = COALESCE(@SQL + ', ', '') + QUOTENAME([Name])
 FROM sys.columns sc
 LEFT JOIN STRING_SPLIT(@list_of_columns_excluded, @separator) ss ON sc.[name] = ss.[value]
 WHERE sc.OBJECT_ID = OBJECT_ID(@full_table_name)
 AND ss.[value] IS NULL;

 SELECT @SQL = N'SELECT ' + @SQL + N' FROM ' + @full_table_name;
 EXEC(@SQL)
END
GO