mjhjmtu

I'm trying to automate a database refresh but, now I'm stuck at a part trying to extract the dump files filename to use it with the IMPDP command.

The thing is that we use this parameter doing the export:

dumpfile=dbname_env_date_%u.dmp and a 16-degree parallelism, which creates 16 files like this:

dbname_env_date_01.dmp
dbname_env_date_02.dmp
.
.
.
dbname_env_date_16.dmp

What I'm trying to do is to just extract this "dbname_env_date", so I can use that parameter when importing into another env database.

What I've got so far is this:

dump_filename=$(ls -tr *.dmp | tail -1)

which gives me something like this:

dbname_env_date_16.dmp

and I want it to use later like this:

imdp ' / as sysdba ' dumpfile=$dump_filename_u%.dmp

There is no need to use tail, you know that you get all 16 files.

dump_filename=$(ls -t *_01.dmp | head -n 1)
dump_filename=$dump_filename%_01.dmp
dump_filename="$dump_filename.dmp"

The first line will get the name of the newest file *_01.dmp, the second line will remove the _01.dmp from the name. The third line appends .dmp.

You can also use this:

dump_filename=$(ls -t *_01.dmp | head -n 1 | sed -e 's/_01.dmp/.dmp/')