From deb95f4bba848de03b8cd73a3a23f1c6ddf34dce Mon Sep 17 00:00:00 2001
From: Karsten Blees <blees@dcon.de>
Date: Sun, 15 Jan 2012 23:04:58 +0100
Subject: [PATCH] Add recodetree script to convert non-ASCII file names to
 UTF-8

Add a script that allows to check a repository for non-ASCII file names,
and to convert the HEAD commit or the entire history from the current
system encoding to UTF-8.

Signed-off-by: Karsten Blees <blees@dcon.de>
---
 bin/recodetree             | 74 ++++++++++++++++++++++++++++++++++++++++++++++
 share/WinGit/copy-files.sh |  2 +-
 2 files changed, 75 insertions(+), 1 deletion(-)
 create mode 100644 bin/recodetree

diff --git a/bin/recodetree b/bin/recodetree
new file mode 100644
index 00000000..7cc82576
--- /dev/null
+++ b/bin/recodetree
@@ -0,0 +1,74 @@
+#!/bin/sh
+# reencode file names in git tree objects from Windows system encoding to UTF-8
+
+usage()
+{
+	echo "WARNING this script is not intended as a full-fledged and fool-proof
+migration utility, but rather as a proof of concept or collection of examples.
+Use at your own risk.
+
+In particular, the script doesn't check or fix any non-ascii names in config
+files (such as .git/config, .gitignore or .gitmodules). For specific migration
+tasks, some of the commands in this script may require tweaking.
+
+Usage: recodetree <command>
+recodetree check    Checks git repository for non-ascii file names.
+recodetree preview  Same as 'check', but prints file names as if converted
+                    from cp$(getcp) to UTF-8.
+recodetree head     Converts the current HEAD from cp$(getcp) to UTF-8.
+                    Does not commit, use git status to check results first.
+recodetree history  Converts entire repository history from cp$(getcp) to UTF-8.
+                    WARNING: this rewrites the history of the repository.
+                    - Make a BACKUP copy of the repository before using this!
+                    - Read 'git help filter-branch' for implications of
+                      rewriting history."
+}
+
+check()
+{
+	# use fast-export to dump all file names in the history
+	git fast-export --no-data --signed-tags=strip --tag-of-filtered-object=drop --all |
+	# use awk to filter for non-ascii names
+	awk --posix '/^M [0-9]{6} [0-9a-f]{40} .*[\200-\377]/{print substr($0,51);}'
+}
+
+recode()
+{
+	# convert from system encoding to UTF-8
+	iconv -c -f cp$(getcp) -t utf-8
+}
+
+recode_tree()
+{
+	# clear the index
+	git rm -f -r -q --cached --ignore-unmatch \*
+	# list specified commit, reencode and add to index
+	git ls-tree -z -r $1 | recode | git update-index -z --index-info
+}
+
+case $1 in
+	check)
+		check
+		;;
+
+	preview)
+		check | recode
+		;;
+
+	head)
+		recode_tree HEAD
+		;;
+
+	history)
+		git filter-branch --index-filter 'recodetree filter' -- --all
+		;;
+
+	filter)
+		# used internally by recodetree history
+		recode_tree $GIT_COMMIT
+		;;
+
+	*)
+		usage
+		;;
+esac
diff --git a/share/WinGit/copy-files.sh b/share/WinGit/copy-files.sh
index 1b488549..ac7cc626 100755
--- a/share/WinGit/copy-files.sh
+++ b/share/WinGit/copy-files.sh
@@ -45,7 +45,7 @@ dirname.exe,\
 du.exe,echo,egrep,env.exe,expr.exe,false.exe,find.exe,flex.exe,gawk.exe,grep.exe,\
 head.exe,id.exe,kill.exe,less.exe,libW11.dll,ln.exe,\
 ls.exe,m4.exe,md5sum.exe,mkdir.exe,msys-1.0.dll,msysltdl-3.dll,mv.exe,patch.exe,\
-patch.exe.manifest,perl.exe,printf,ps.exe,pwd,rm.exe,rmdir.exe,rxvt.exe,\
+patch.exe.manifest,perl.exe,printf,ps.exe,pwd,recodetree,rm.exe,rmdir.exe,rxvt.exe,\
 scp.exe,sed.exe,sh.exe,sleep.exe,sort.exe,split.exe,\
 ssh-agent.exe,ssh.exe,ssh-add.exe,ssh-keygen.exe,ssh-keyscan.exe,\
 tail.exe,tar.exe,tee.exe,touch.exe,tr.exe,true.exe,uname.exe,uniq.exe,\
-- 
2.11.4.GIT