Skip to content

Instantly share code, notes, and snippets.

@chrisjrn
Created June 30, 2018 18:04
Show Gist options
  • Select an option

  • Save chrisjrn/0ab3e1169110b454a5876903e30b5ebd to your computer and use it in GitHub Desktop.

Select an option

Save chrisjrn/0ab3e1169110b454a5876903e30b5ebd to your computer and use it in GitHub Desktop.
'''
wget -r --mirror --page-requisites SOURCE_DOMAIN -D SOURCE_DOMAIN --adjust-extension
'''
import os
import sys
basedir = sys.argv[1]
directories = []
moves = []
for folder, subdirs, files in os.walk(basedir):
for f in files:
basename, dot, ext = f.rpartition(".")
if not basename:
continue
if ext != "html":
continue
if basename == "index":
continue
if basename not in subdirs:
directories.append((folder, basename))
moves.append((folder, basename, f))
for folder, basename in directories:
a = os.path.join(folder, basename)
print("mkdir", folder, basename)
os.mkdir(a)
for folder, dest, filename in moves:
orig = os.path.join(folder, filename)
nd = os.path.join(folder, dest, "index.html")
print("mv : " + orig + " -to- " + nd)
os.rename(orig, nd)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment