Export table to file with column headers (column names) using the bcp utility and SQL Server 2008
This method automatically outputs column names with your row data using BCP.
The script writes one file for the column headers (read from INFORMATION_SCHEMA.COLUMNS table) then appends another file with the table data.
The final output is combined into TableData.csv which has the headers and row data. Just replace the environment variables at the top to specify the Server, Database and Table name.
set BCP_EXPORT_SERVER=put_my_server_name_here
set BCP_EXPORT_DB=put_my_db_name_here
set BCP_EXPORT_TABLE=put_my_table_name_here
BCP "DECLARE @colnames VARCHAR(max);SELECT @colnames = COALESCE(@colnames + ',', '') + column_name from %BCP_EXPORT_DB%.INFORMATION_SCHEMA.COLUMNS where TABLE_NAME='%BCP_EXPORT_TABLE%'; select @colnames;" queryout HeadersOnly.csv -c -T -S%BCP_EXPORT_SERVER%
BCP %BCP_EXPORT_DB%.dbo.%BCP_EXPORT_TABLE% out TableDataWithoutHeaders.csv -c -t, -T -S%BCP_EXPORT_SERVER%
set BCP_EXPORT_SERVER=
set BCP_EXPORT_DB=
set BCP_EXPORT_TABLE=
copy /b HeadersOnly.csv+TableDataWithoutHeaders.csv TableData.csv
del HeadersOnly.csv
del TableDataWithoutHeaders.csv
Note that if you need to supply credentials, replace the -T option with -U my_username -P my_password
This method has the advantage of always having the column names in sync with the table by using INFORMATION_SCHEMA.COLUMNS. The downside is that it creates temporary files. Microsoft should really fix the bcp utility to support this.
This solution uses the SQL row concatenation trick from here combined with bcp ideas from here
The easiest is to use the queryout option and use union all to link a column list with the actual table content
bcp "select 'col1', 'col2',... union all select * from myschema.dbo.myTableout" queryout myTable.csv /SmyServer01 /c /t, -T
An example:
create table Question1355876
(id int, name varchar(10), someinfo numeric)
insert into Question1355876
values (1, 'a', 123.12)
, (2, 'b', 456.78)
, (3, 'c', 901.12)
, (4, 'd', 353.76)
This query will return the information with the headers as first row (note the casts of the numeric values):
select 'col1', 'col2', 'col3'
union all
select cast(id as varchar(10)), name, cast(someinfo as varchar(28))
from Question1355876
The bcp command will be:
bcp "select 'col1', 'col2', 'col3' union all select cast(id as varchar(10)), name, cast(someinfo as varchar(28)) from Question1355876" queryout myTable.csv /SmyServer01 /c /t, -T
A good alternative is SqlCmd, since it does include headers, but it has the downside of adding space padding around the data for human readability. You can combine SqlCmd with the GnuWin32 sed (stream editing) utility to cleanup the results. Here's an example that worked for me, though I can't guarantee that it's bulletproof.
First, export the data:
sqlcmd -S Server -i C:\Temp\Query.sql -o C:\Temp\Results.txt -s" "
The -s" " is a tab character in double quotes. I found that you have to run this command via a batch file, otherwise the Windows command prompt will treat the tab as an automatic completion command and will substitute a filename in place of the tab.
If Query.sql contains:
SELECT name, object_id, type_desc, create_date
FROM MSDB.sys.views
WHERE name LIKE 'sysmail%'
then you'll see something like this in Results.txt
name object_id type_desc create_date ------------------------------------------- ----------- ------------------- ----------------------- sysmail_allitems 2001442204 VIEW 2012-07-20 17:38:27.820 sysmail_sentitems 2017442261 VIEW 2012-07-20 17:38:27.837 sysmail_unsentitems 2033442318 VIEW 2012-07-20 17:38:27.850 sysmail_faileditems 2049442375 VIEW 2012-07-20 17:38:27.860 sysmail_mailattachments 2097442546 VIEW 2012-07-20 17:38:27.933 sysmail_event_log 2129442660 VIEW 2012-07-20 17:38:28.040 (6 rows affected)
Next, parse the text using sed:
sed -r "s/ +\t/\t/g" C:\Temp\Results.txt | sed -r "s/\t +/\t/g" | sed -r "s/(^ +| +$)//g" | sed 2d | sed $d | sed "/^$/d" > C:\Temp\Results_New.txt
Note that the 2d command means to delete the second line, the $d command means to delete the last line, and "/^$/d" deletes any blank lines.
The cleaned up file looks like this (though I replaced the tabs with | so they could be visualized here):
name|object_id|type_desc|create_date sysmail_allitems|2001442204|VIEW|2012-07-20 17:38:27.820 sysmail_sentitems|2017442261|VIEW|2012-07-20 17:38:27.837 sysmail_unsentitems|2033442318|VIEW|2012-07-20 17:38:27.850 sysmail_faileditems|2049442375|VIEW|2012-07-20 17:38:27.860 sysmail_mailattachments|2097442546|VIEW|2012-07-20 17:38:27.933 sysmail_event_log|2129442660|VIEW|2012-07-20 17:38:28.040