Compare Directories and Remove Duplicates
This Bash script compares subdirectories in two locations, groups them based on fuzzy name similarity, and automatically removes duplicates. It supports both undesirable-word–based removal and an automatic pruning mechanism (keeping the first directory in each group).

Features
Normalization & Cleaning:
Converts directory names to lowercase, removes punctuation, and strips out any undesirable words specified in a words file.

Fuzzy Matching:
Uses Python's difflib to compute a similarity ratio between cleaned directory names. A configurable threshold determines if two names are “similar.”

Automatic Removal:

Undesirable Words: Automatically removes directories whose original name contains one of the undesirable words if another similar directory does not.
Duplicate Pruning: After the initial pass, automatically removes all duplicates in a group while keeping the first entry.
Dry-Run Mode:
Preview actions without deleting any directories by using the --dry-run flag.

Requirements
Bash (version 4+ recommended)
Python 3
bc for floating point comparisons
Installation
Clone the repository:

git clone https://github.com/yourusername/compare-dirs.git
cd compare-dirs
Make the script executable:

bash
Kopiér
chmod +x compare_dirs.sh

Usage
./compare_dirs.sh [--dry-run] [--threshold <threshold>] <dir1> <dir2> <words_file>
Arguments
<dir1>: The first directory containing subdirectories to compare.
<dir2>: The second directory containing subdirectories to compare.
<words_file>: A text file with one undesirable word per line. These words are removed from directory names during the cleaning process.

Options
--dry-run

Print the actions without actually removing any directories.

--threshold <threshold>
Set the fuzzy similarity threshold (default is 0.8). A lower threshold (e.g., 0.7) will group more directories as duplicates.

Examples
Dry Run with Default Threshold (0.8):


./compare_dirs.sh --dry-run /mnt/dsnas /mnt/dsnas1 ./words
Dry Run with a Custom Threshold (0.7):


./compare_dirs.sh --dry-run --threshold 0.7 /mnt/dsnas /mnt/dsnas1 ./words
Actual Run (without dry-run):


./compare_dirs.sh /mnt/dsnas /mnt/dsnas1 ./words
How It Works
Scanning:
The script scans for immediate subdirectories in the two specified directories.

Normalization & Cleaning:
Each subdirectory name is normalized (converted to lowercase, punctuation removed) and then “cleaned” by stripping out undesirable words (one per line from the words file).

Grouping:
Using a Python helper with difflib.SequenceMatcher, directories are grouped by comparing their cleaned names. If the similarity ratio meets or exceeds the threshold, they are considered duplicates.

Removal:
Automatic Removal Based on Undesirable Words:
Within duplicate groups, if one directory’s original name contains an undesirable word while an alternative does not, that directory is flagged for removal.

Duplicate Pruning:
After the undesirable-word check, any remaining duplicate groups are pruned by keeping the first directory in each group and removing the rest.

Dry-Run:
When run with the --dry-run flag, the script will print what it would remove without actually deleting any directories.