paunch-dir 0.0.4

paunch-dir

An algorithm to generate a path from file name which directory structure could store mass files

You could generated directory structure (from a uuid file name) just like:

├── 02
│   └── 9d
│       └── 5b
│           └── cb7c89776a923d6050ff9cffcf1333284f.jpg
|               # Original file name : 029d5bcb7c89776a923d6050ff9cffcf1333284f.jpg
├── 03
│   ├── 2b
│   │   └── f9
│   │       └── 7cdb028c51ec56ec53cbe34946525a3e37.jpg
│   └── f9
│       └── 4d
│           └── 9458a9ee5396911268b0da6b3552f154e8.jpg
...

It’s ineffectively for file system parse large number files in same directory. And there have limit files that could be sotre in a directory depends on different file system format. You will see it’s very slow if you open a directory with hundreds of thousands files in file browser. Separate them into different directory with specific algorithm that could greatly improve efficiency when parse those files.

from paunchdir import PaunchDir

# Generate path from uuid file name
pdir = PaunchDir(".")
print(pdir.compose('032bf97cdb028c51ec56ec53cbe34946525a3e37.jpg'))
# ==> "./03/2b/f9/7cdb028c51ec56ec53cbe34946525a3e37.jpg"

# Decompose uuid file name from a path
print(pdir.decompose("./03/2b/f9/7cdb028c51ec56ec53cbe34946525a3e37.jpg"))
# ==> "032bf97cdb028c51ec56ec53cbe34946525a3e37.jpg"

# Generate path from uuid file name with one level (Just like what the git
# store it's objects)
pdir = PaunchDir("./objects", levels=1)
print(pdir.compose('032bf97cdb028c51ec56ec53cbe34946525a3e37'))
# ==> "./03/2bf97cdb028c51ec56ec53cbe34946525a3e37"

History

0.0.2 (2018-06-21)

• First release on PyPI.