Rename files based on their parent directory in Bash

Question

Been trying to piece together a couple previous posts for this task. The directory tree looks like this:

TEST
   |ABC_12345678
       3_XYZ
   |ABC_23456789
       3_XYZ
   etc

Each folder within the parent folder named "TEST" always starts with ABC_\d{8} -the 8 digits are always different. Within the folder ABC_\d{8} is always a folder entitled 3_XYZ that always has a file named "MD2_Phd.txt". The goal is to rename each "MD2_PhD.txt" file with the specific 8 digit ID found in the ABC folder name i.e. "\d{8}_PhD.txt"

After several iterations on various bits of code from different posts this is the best I can come up with,

cd /home/etc/Desktop/etc/TEST
find -type d -name 'ABC_(\d{8})' |
find $d -name "*_PhD.txt" -execdir rename 's/MD2$/$d/' "{}" \;
done

Answer 1

find + bash solution:

find -type f -regextype posix-egrep -regex ".*/TEST/ABC_[0-9]{8}/3_XYZ/MD2_Phd\.txt" \
-exec bash -c 'abc="${0%/*/*}"; fp="${0%/*}/";
 mv "$0" "$fp${abc##*_}_PhD.txt" ' {} \;

---

Viewing results:

$ tree TEST/ABC_*
TEST/ABC_12345678
└── 3_XYZ
    └── 12345678_PhD.txt
TEST/ABC_1234ss5678
└── 3_XYZ
    └── MD2_Phd.txt
TEST/ABC_23456789
└── 3_XYZ
    └── 23456789_PhD.txt