From 1b2c8eb3b653ecc05e332be16a79e740cd679a9f Mon Sep 17 00:00:00 2001
From: Aaron LI <aly@aaronly.me>
Date: Tue, 27 Aug 2019 15:54:43 +0800
Subject: cli/unzip-gbk: Rewrite in shell and use bsdtar/7z

* Rewrite in shell instead of using the deprecated Python 2.
* Use bsdtar/7z to extract zip to preserve the filename encoding;
  these tools also support encrypted zip archives.
* Use iconv to convert the filename encoding.
---
 cli/unzip-gbk.py | 26 --------------------------
 cli/unzip-gbk.sh | 55 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 55 insertions(+), 26 deletions(-)
 delete mode 100755 cli/unzip-gbk.py
 create mode 100755 cli/unzip-gbk.sh

(limited to 'cli')

diff --git a/cli/unzip-gbk.py b/cli/unzip-gbk.py
deleted file mode 100755
index 423e10f..0000000
--- a/cli/unzip-gbk.py
+++ /dev/null
@@ -1,26 +0,0 @@
-#!/usr/bin/env python
-# -*- coding: utf-8 -*-
-# unzip-gbk.py
-#
-# http://note.ninehills.info/linux-gbk.html
-#
-
-import os
-import sys
-import zipfile
-
-print "Processing File " + sys.argv[1]
-
-file=zipfile.ZipFile(sys.argv[1],"r");
-for name in file.namelist():
-    utf8name=name.decode('gbk')
-    print "Extracting " + utf8name
-    pathname = os.path.dirname(utf8name)
-    if not os.path.exists(pathname) and pathname!= "":
-        os.makedirs(pathname)
-    data = file.read(name)
-    if not os.path.exists(utf8name):
-        fo = open(utf8name, "w")
-        fo.write(data)
-        fo.close
-file.close()
diff --git a/cli/unzip-gbk.sh b/cli/unzip-gbk.sh
new file mode 100755
index 0000000..53c7b57
--- /dev/null
+++ b/cli/unzip-gbk.sh
@@ -0,0 +1,55 @@
+#!/bin/sh
+#
+# Extract a zip archive and fix Chinese filenames.
+#
+# Credit: https://superuser.com/a/872616
+#
+
+has() {
+	type "$1" >/dev/null 2>&1
+}
+
+extract() {
+	if has bsdtar; then
+		# bsdtar provided by libarchive
+		bsdtar -xvf "$1"
+	elif has 7z; then
+		# 7z provided by p7zip
+		env LC_ALL=C 7z x "$1"
+	else
+		echo "ERROR: Neither bsdtar nor 7z found" >&2
+		exit 1
+	fi
+}
+
+fixnames() {
+	find . -depth | while read -r p; do
+		dn=$(dirname "${p}")
+		fn=$(basename "${p}")
+		fn2=$(echo "${fn}" | iconv -f gbk -t utf-8)
+		if [ "${fn}" != "${fn2}" ]; then
+			mv -v "${dn}/${fn}" "${dn}/${fn2}"
+		fi
+	done
+}
+
+case $1 in
+'' | -h | --help)
+	echo "usage: ${0##*/} <zip>"
+	exit 1
+	;;
+esac
+
+zipfile=$(realpath "$1")
+curdir=$(pwd)
+tmpdir=$(mktemp -d)
+cd "${tmpdir}"
+
+echo "Extracting archive '${zipfile}' ..."
+extract "${zipfile}"
+echo "Fixing filenames ..."
+fixnames
+
+cd "${curdir}"
+mv "${tmpdir}"/* .
+rmdir "${tmpdir}"
-- 
cgit v1.2.2