Hi

I am having an unix shell script called call.ksh and we are calling oracle packages from script file ...
having the following code

----------------------------------
sqlplus -s >> $LOGFILE << EOF
${TPE_LOGIN_STRING}
whenever sqlerror exit SQL.SQLCODE
set serverout on size 1000000 ;
execute PKG_MINE.CALLME() ;
commit;
EOF
----------------------------

CAN U PLEASE HELP ME IN UNDERSTANDING THE CODE

Hi,

the red text you refer to is called the construction of a 'here-document' in UNIX. It consists of a command where between tags the input is specified on the shell commandline. In general:Ofcourse this can also be specified in a shellscript. This makes it possible to redirect the output of the command to a logfile.

Regards,
Hans

Hi Hans,

thanks for your reply.

You mean to say that what ever is their between the lines of code <<EOF ...EOF is going to logfile specified.

sqlplus -s >> $LOGFILE << EOF
--
--
--
EOF



you have specified

COMMAND ....<< INPUT

INPUT

so we can use INPUT also ?

thanks !!

Yes, the EOF or INPUT are used to identify where the here-document ends. You can use whichever character sequence you want - as long as the end-marker doesn't occur on a single line within the here-document itself. (It may be part of some other text within the document.) Typically, EOF is used, however.

No, what I meant to say was that the input the sqlplus program expects is put between the <<WHATEVER ... WHATEVER tags. The output of the sqlplus program is redirected to a logfile but unless the sqlplus program can accept it's input through a pipe it's not possible to both redirect input and output without the here-document construction.

Regards,
Hans

It is as though the contents were coming from a separate file (and so called a "here"file).

The same word signifies the start abnd end of the here file. It could as well have
been:
COMMAND ....<< ANYWORD

ANYWORD

instead of

COMMAND ....<< INPUT

INPUT
.

The end marker for the here file must start at column 1 (annot be indented).

End