Reputation: 20667
I am creating a quick backup script that will dump some databases into a nice/neat directory structure and I realized that I need to test to make sure that the directories exist before I create them. The code I have works, but is there a better way to do it?
[ -d "$BACKUP_DIR" ] || mkdir "$BACKUP_DIR"
[ -d "$BACKUP_DIR/$client" ] || mkdir "$BACKUP_DIR/$client"
[ -d "$BACKUP_DIR/$client/$year" ] || mkdir "$BACKUP_DIR/$client/$year"
[ -d "$BACKUP_DIR/$client/$year/$month" ] || mkdir "$BACKUP_DIR/$client/$year/$month"
[ -d "$BACKUP_DIR/$client/$year/$month/$day" ] || mkdir "$BACKUP_DIR/$client/$year/$month/$day"
Upvotes: 239
Views: 143471
Reputation: 11607
While existing answers definitely solve the purpose, if you’re looking to replicate a nested directory structure under two different subdirectories, then you can do this:
mkdir -p {main,test}/{resources,scala/com/company}
It will create the following directory structure under the directory from where it is invoked:
├── main
│ ├── resources
│ └── scala
│ └── com
│ └── company
└── test
├── resources
└── scala
└── com
└── company
The example was taken from this link for creating an SBT directory structure.
Upvotes: 28
Reputation: 11
mkdir -p newDir/subdir{1..8}
ls newDir/
subdir1 subdir2 subdir3 subdir4 subdir5 subdir6 subdir7 subdir8
Upvotes: 0
Reputation: 100040
You can use the -p
parameter, which is documented as:
-p, --parents
no error if existing, make parent directories as needed
So:
mkdir -p "$BACKUP_DIR/$client/$year/$month/$day"
Upvotes: 533