path: root/git-remote-gcrypt

#!/bin/sh

# git-remote-gcrypt
# Copyright 2013  by Ulrik
# License: GPLv2 or any later version, see http://www.gnu.org/licenses/
#
# See README

#set -x
set -e

Did_find_repo=  # yes for connected, no for no repo
Localdir="${GIT_DIR:=.git}/remote-gcrypt"
export GITCEPTION="${GITCEPTION:-}+" # Reuse $Gref except when stacked
Gref="refs/gcrypt/gitception$GITCEPTION"
Gref_rbranch="refs/heads/master"
Repoid=
Packkey_bytes=33  # 33 random bytes for passphrase, still compatible if changed
Hashtype=SHA224   # incompatible if changed
Packpfx="pack :${Hashtype}:"
Keeppfx="keep :${Hashtype}:"

Branchlist=
Packlist=
Keeplist=
Extension_list=
Repack_limit=25
Packlist_delete=

Recipients=
Signers=
Goodsig=

# compat/utility functions
xecho()
{
	cat <<EOF
$@
EOF
}
xecho_n() { xecho "$@" | tr -d \\n ; } # kill newlines
echo_git() { xecho "$@" ; }  # Code clarity
echo_info() { xecho "gcrypt:" "$@" >&2; }
echo_die() { echo_info "$@" ; exit 1; }

isnull() { case "$1" in "") return 0;; *) return 1;; esac; }
isnonnull() { ! isnull "$1"; }
iseq() { case "$1" in "$2") return 0;; *) return 1;; esac; }
isnoteq() { ! iseq "$@"; }

# Append $2 to $1 with a newline separator
append() { isnull "$1" || xecho "$1" && xecho "$2"; }
isurl() { isnull "${2%%$1://*}"; }
islocalrepo() { isnull "${1##/*}" && [ ! -e "$1/HEAD" ]; }

xgrep() { command grep "$@" || : ; }
sort_C() { LC_ALL=C command sort "$@"; }
sort_stable_k2()
{
	awk '{ printf("%08d\t%s\n", NR, $0) }' | sort_C -k 3,3 -k 1,1 |cut -f 2-
}

tac() { sed '1!G;h;$!d'; }

# Split $1 into $prefix_:$suffix_
splitcolon()
{
	prefix_=${1%%:*}
	suffix_=${1#*:}
}
repoidstr() { xecho "repo :${Hashtype}:$Repoid 1"; }

## gitception part
# Fetch giturl $1, file $2
gitception_get()
{
	# Take care to preserve FETCH_HEAD
	local ret_=: obj_id= f_head="$GIT_DIR/FETCH_HEAD"
	[ -e "$f_head" ] && command mv -f "$f_head" "$f_head.$$~" || :
	git fetch -q -f "$1" "$Gref_rbranch:$Gref" 2>/dev/tty >/dev/null &&
		obj_id="$(git ls-tree "$Gref" | xgrep -E '\b'"$2"'$' | awk '{print $3}')" &&
		isnonnull "$obj_id" && git cat-file blob "$obj_id" && ret_=: ||
		{ ret_=false && : ; }
	[ -e "$f_head.$$~" ] && command mv -f "$f_head.$$~" "$f_head" || :
	$ret_
}

anon_commit()
{
	GIT_AUTHOR_NAME="root" GIT_AUTHOR_EMAIL="root@localhost" \
	GIT_AUTHOR_DATE="1356994801 -0400" GIT_COMMITTER_NAME="root" \
	GIT_COMMITTER_EMAIL="root@localhost" \
	GIT_COMMITTER_DATE="1356994801 -0400" \
		git commit-tree "$@" <<EOF
Initial commit
EOF
}

# Get 'tree' from $1, change file $2 to obj id $3
update_tree()
{
	local tab_="	"
	# $2 is a filename from the repo format
	(git ls-tree "$1" | xgrep -v -E '\b'"$2"'$';
		xecho "100644 blob $3$tab_$2") | git mktree
}

# Put giturl $1, file $2
# depends on previous GET to set $Gref and depends on PUT_FINAL later
gitception_put()
{
	local obj_id= tree_id= commit_id=
	obj_id=$(git hash-object -w --stdin) &&
		tree_id=$(update_tree "$Gref" "$2" "$obj_id") &&
		commit_id=$(anon_commit "$tree_id") &&
		git update-ref "$Gref" "$commit_id"
}

# Remove giturl $1, file $2
# depends on previous GET like put
gitception_remove()
{
	local tree_id= commit_id= tab_="	"
	# $2 is a filename from the repo format
	tree_id=$(git ls-tree "$Gref" | xgrep -v -E '\b'"$2"'$' | git mktree) &&
		commit_id=$(anon_commit "$tree_id") &&
		git update-ref "$Gref" "$commit_id"
}

gitception_new_repo()
{
	local empty_tree=4b825dc642cb6eb9a060e54bf8d69288fbee4904
	# get any file to update Gref, and if it's not updated we create empty
	git update-ref -d "$Gref" || :
	gitception_get "$1" "x" || :
	git rev-parse -q --verify "$Gref" >/dev/null && return 0 ||
		commit_id=$(anon_commit "$empty_tree") &&
		git update-ref "$Gref" "$commit_id"
}
## end gitception

# Fetch repo $1, file $2, tmpfile in $3
GET()
{
	if isurl ssh "$1"
	then
		splitcolon "${1#ssh://}"
		(exec 0>&-; ssh "$prefix_" "cat $suffix_/$2") > "$3"
	elif isurl sftp "$1"
	then
		(exec 0>&-; curl -s -S -k "$1/$2") > "$3"
	elif isurl rsync "$1"
	then
		(exec 0>&-; rsync -I -W "${1#rsync://}"/"$2" "$3" >&2)
	elif islocalrepo "$1"
	then
		cat "$1/$2" > "$3"
	else
		gitception_get "${1#gitception://}" "$2" > "$3"
	fi
}

# Put repo $1, file $2 or fail, tmpfile in $3
PUT()
{
	if isurl ssh "$1"
	then
		splitcolon "${1#ssh://}"
		ssh "$prefix_" "cat > $suffix_/$2" < "$3"
	elif isurl sftp "$1"
	then
		curl -s -S -k --ftp-create-dirs -T "$3" "$1/$2"
	elif isurl rsync "$1"
	then
		rsync -I -W "$3" "${1#rsync://}"/"$2" >&2
	elif islocalrepo "$1"
	then
		cat > "$1/$2" < "$3"
	else
		gitception_put "${1#gitception://}" "$2" < "$3"
	fi
}

# Put all PUT changes for repo $1 at once
PUT_FINAL()
{
	if isurl ssh "$1" || isurl sftp "$1" || islocalrepo "$1" || isurl rsync "$1"
	then
		:
	else
		git push --quiet -f "${1#gitception://}" "$Gref:$Gref_rbranch"
	fi
}

# Put directory for repo $1
PUTREPO()
{
	if isurl ssh "$1"
	then
		splitcolon "${1#ssh://}"
		(exec 0>&- ; ssh "$prefix_" "mkdir -p $suffix_")
	elif isurl sftp "$1"
	then
		:
	elif isurl rsync "$1"
	then
		rsync -q -r --exclude='*' "$Localdir/" "${1#rsync://}" >&2
	elif islocalrepo "$1"
	then
		mkdir -p "$1"
	else
		gitception_new_repo "${1#gitception://}"
	fi
}

# For repo $1, delete all newline-separated files in $2
REMOVE()
{
	local fn_=
	if isurl ssh "$1"
	then
		splitcolon "${1#ssh://}"
		(exec 0>&- ; ssh "$prefix_" "cd $suffix_; rm $2")
	elif isurl sftp "$1"
	then
		# FIXME
		echo_info "sftp: Ignore remove request $1/$2"
	elif isurl rsync "$1"
	then
		xecho "$2" | rsync -I -W -v -r --delete --include-from=- \
			--exclude='*' "$Localdir"/ "${1#rsync://}/" >&2
	elif islocalrepo "$1"
	then
		(cd "$1"; rm $2)
	else
		for fn_ in $2; do
			gitception_remove "${1#gitception://}" "$fn_"
		done
	fi
}

CLEAN_FINAL()
{
	if isurl ssh "$1" || isurl sftp "$1" || islocalrepo "$1" || isurl rsync "$1"
	then
		:
	else
		git update-ref -d "$Gref" || :
	fi
}

addsignkeyparam()
{
	if isnull "$Conf_signkey"; then
		"$@"
	else
		"$@" -u "$Conf_signkey"
	fi
}


ENCRYPT()
{
	gpg --batch --force-mdc --compress-algo none --passphrase-fd 3 -c 3<<EOF
$1
EOF
}

DECRYPT()
{
	gpg -q --batch --no-default-keyring --secret-keyring /dev/null \
		--keyring /dev/null --passphrase-fd 3 -d  3<<EOF
$1
EOF
}

# Encrypt to recipients $1
PRIVENCRYPT()
{
	addsignkeyparam gpg --compress-algo none -se $1
}

PRIVDECRYPT()
{
	local status_=
	exec 4>&1 &&
	status_=$(gpg  --status-fd 3 -q -d 3>&1 1>&4) &&
	xecho "$status_" | grep "^\[GNUPG:\] ENC_TO " >/dev/null &&
	(xecho "$status_" | grep -e "$Goodsig" >/dev/null || {
		echo_info "Failed to verify manifest signature!" &&
		echo_info "Only accepting signatories: ${Signers:-(none)}" &&
		return 1
	})
}

# Generate $1 random bytes
genkey()
{
	gpg --armor --gen-rand 1 "$1"
}

pack_hash()
{
	local hash_=
	hash_=$(gpg --with-colons --print-md "$Hashtype" | tr A-F a-f)
	hash_=${hash_#:*:}
	xecho "${hash_%:}"
}

# Pass the branch/ref by pipe to git
safe_git_rev_parse()
{
	git cat-file --batch-check 2>/dev/null |
		xgrep -v "missing" | cut -f 1 -d ' '
}

make_new_repo()
{
	local urlid_= fix_config=
	echo_info "Setting up new repository"
	PUTREPO "$URL"

	# We need a relatively short ID for URL+REPO
	# The manifest will be stored at pack_hash($urlid_)
	# Needed assumption: the same user should have no duplicate urlid_
	# For now, we arbitrarily use 9 random bytes (72 bits)
	urlid_=$(genkey 9 | tr '/+' 'Zz')
	Repoid=$(xecho_n "$urlid_" | pack_hash)
	iseq "${NAME#gcrypt::}" "$URL" || {
		git config "remote.$NAME.url" "gcrypt::$URL#$urlid_"
		fix_config=1
	}
	echo_info "Repository URL is" "gcrypt::$URL#$urlid_"
	Extension_list=$(xecho "extn comment")
	isnull "$fix_config" || echo_info "(configuration for $NAME updated)"
}


read_config()
{
	local recp_= key_line= cap_= conf_keyring= conf_part=
	Conf_signkey=$(git config --path user.signingkey || :)
	conf_keyring=$(git config --path gcrypt.keyring || :)
	conf_part=$(git config --get gcrypt.participants '.+' || :)

	# Figure out which keys we should encrypt to or accept signatures from
	if isnonnull "$conf_keyring" && isnull "$conf_part"
	then
		echo_info "WARNING: Setting gcrypt.keyring is deprecated," \
			  "use gcrypt.participants instead."
		conf_part=$(gpg --no-default-keyring --keyring "$conf_keyring" \
		--with-colons --fast-list -k | grep ^pub | cut -f 5 -d :)
	fi

	for recp_ in $conf_part
	do
		key_line=$(gpg --with-colons --fast-list -k "$recp_" | xgrep ^pub)
		keyid_=$(xecho "$key_line" | cut -f 5 -d :)

		isnonnull "$keyid_" &&
		Signers="$Signers $keyid_" &&
		Goodsig=$(append "$Goodsig" "^\[GNUPG:\] GOODSIG $keyid_") || {
			echo_info "WARNING: Skipping missing key $recp_"
			continue
		}
		# Check 'E'ncrypt capability
		cap_=$(xecho "$key_line" | cut -f 12 -d :)
		iseq "${cap_#*E}" "$cap_" || Recipients="$Recipients -R $keyid_"
	done

	if isnull "$Recipients"
	then
		echo_info "You have not configured any keys to encrypt to for this repository"
		echo_info "Use ::"
		echo_info "  git config gcrypt.participants YOURKEYID"
		exit 1
	fi
}

ensure_connected()
{
	local manifest_= rcv_repoid= url_id=

	if isnonnull "$Did_find_repo"
	then
		return
	fi
	Did_find_repo=no
	read_config

	# split out Repo ID from URL
	url_id=${URL##*"#"}
	isnoteq "$url_id" "$URL" || {
		url_id=${URL##*/"G."}
		isnoteq "$url_id" "$URL" || return 0
		URL=${URL%/"G.$url_id"}
	}
	URL=${URL%"#$url_id"}
	Repoid=$(xecho_n "$url_id" | pack_hash)

	TmpManifest_Enc="$Localdir/manifest.$$"
	trap 'rm -f "$TmpManifest_Enc"' EXIT
	GET "$URL" "$Repoid" "$TmpManifest_Enc" 2>/dev/null ||
		echo_die "Repository not found: $url_id at $URL"

	Did_find_repo=yes
	echo_info "Decrypting manifest"
	manifest_=$(PRIVDECRYPT < "$TmpManifest_Enc") && 
		isnonnull "$manifest_" ||
		echo_die "Failed to decrypt manifest!"
	rm -f "$TmpManifest_Enc"
	trap - EXIT

	Branchlist=$(xecho "$manifest_" | xgrep -E '^[0-9a-f]{40} ')
	Packlist=$(xecho "$manifest_" | xgrep "^$Packpfx")
	Keeplist=$(xecho "$manifest_" | xgrep "^keep")
	Extension_list=$(xecho "$manifest_" | xgrep "^extn ")
	rcv_repoid=$(xecho "$manifest_" | xgrep "^repo ")
	iseq "$(repoidstr)" "$rcv_repoid" || echo_die "Repository id mismatch!"
}

# $1 is new pack id $2 key
# set did_repack=yes if repacked
repack_if_needed()
{
	local pack_= rcv_id= packline_= premote_= key_= pkeep_= n_=

	# $TmpPack_Encrypted set in caller

	did_repack=no
	isnonnull "$Packlist" || return 0

	premote_=$(xecho "$Packlist" | cut -f 1-2 -d ' ')
	pkeep_=$(xecho "$Keeplist" | cut -f 2 -d ' ')

	if isnull "$pkeep_"; then
		n_=$(xecho "$Packlist" | wc -l)
	else
		n_=$(xecho "$Packlist" | grep -v -F -e "$pkeep_" | wc -l)
	fi
	if [ $Repack_limit -gt "$n_" ]; then
		return
	fi
	echo_info "Repacking remote $NAME, ..."

	rm -r -f "$Localdir/pack"
	mkdir -p "$Localdir/pack"
	DECRYPT "$2" < "$TmpPack_Encrypted" |
		git index-pack -v --stdin "$Localdir/pack/${1}.pack" >/dev/null

	xecho "$premote_" | while read packline_
	do
		isnonnull "$packline_" || continue
		if isnonnull "$pkeep_" &&
			xecho "$packline_" | grep -q -F -e "$pkeep_"
		then
			continue
		fi
		pack_=${packline_#"$Packpfx"}
		GET "$URL" "$pack_" "$TmpPack_Encrypted"
		rcv_id=$(pack_hash < "$TmpPack_Encrypted")
		if isnoteq "$rcv_id" "$pack_"
		then
			echo_die "Packfile $pack_ does not match digest!"
		fi
		key_=$(xecho "$Packlist" | grep "$pack_" | cut -f 3 -d ' ')
		DECRYPT "$key_" < "$TmpPack_Encrypted" |
		git index-pack -v --stdin "$Localdir/pack/${pack_}.pack" >/dev/null
	done
	key_=$(genkey "$Packkey_bytes")

	git verify-pack -v "$Localdir"/pack/*.idx | grep -E '^[0-9a-f]{40}' |
		cut -f 1 -d ' ' |
		GIT_ALTERNATE_OBJECT_DIRECTORIES=$Localdir \
		git pack-objects --stdout | ENCRYPT "$key_" > "$TmpPack_Encrypted"

	# Truncate packlist to only the kept packs
	if isnull "$pkeep_"; then
		Packlist_delete=$premote_
		Packlist=
	else
		Packlist_delete=$(xecho "$premote_" | xgrep -v -F -e "$pkeep_")
		Packlist=$(xecho "$Packlist" | xgrep -F -e "$pkeep_")
	fi

	pack_id=$(pack_hash < "$TmpPack_Encrypted")
	Packlist=$(append "$Packlist" "$Packpfx$pack_id $key_")
	Keeplist=$(append "$Keeplist" "$Keeppfx$pack_id 1")
	rm -r -f "$Localdir/pack"
	did_repack=yes
}

do_capabilities()
{
	echo_git fetch
	echo_git push
	echo_git
}

do_list()
{
	local obj_id= ref_name= line_=
	ensure_connected

	xecho "$Branchlist" | while read line_
	do
		isnonnull "$line_" || break
		obj_id=${line_%% *}
		ref_name=${line_##* }
		echo_git "$obj_id" "$ref_name"
		if iseq "$ref_name" "refs/heads/master"
		then
			echo_git "@refs/heads/master HEAD"
		fi
	done

	# end with blank line
	echo_git
}

do_fetch()
{
	# Security protocol:
	# The PACK id is the hash of the encrypted git packfile.
	# We only download packs mentioned in the encrypted manifest,
	# and check their digest when received.
	local pack_= rcv_id= packline_= pneed_= pboth_= phave_= premote_= key_=

	ensure_connected

	if isnull "$Packlist"
	then
		echo_git # end with blank line
		return
	fi

	TmpPack_Encrypted="$Localdir/tmp_pack_ENCRYPTED_.$$"
	trap 'rm -f "$TmpPack_Encrypted"' EXIT

	premote_=$(xecho "$Packlist" | cut -f 1-2 -d ' ')
	# Needed packs is  Packlist - (phave & Packlist)
	# The `+` for $GITCEPTION is pointless but we will be safe for stacking
	phave_="$(cat "$Localdir/have_packs+" 2>/dev/null || :)"
	pboth_="$(xecho "$premote_" | xgrep -F  -x -e "$phave_")"
	pneed_="$(xecho "$premote_" | xgrep -F -vx -e "$phave_")"

	xecho "$pneed_" | while read packline_
	do
		isnonnull "$packline_" || continue
		pack_=${packline_#"$Packpfx"}
		GET "$URL" "$pack_" "$TmpPack_Encrypted"
		rcv_id=$(pack_hash < "$TmpPack_Encrypted")
		if isnoteq "$rcv_id" "$pack_"
		then
			echo_die "Packfile $pack_ does not match digest!"
		fi
		key_=$(xecho "$Packlist" | grep "$pack_" | cut -f 3 -d ' ')
		DECRYPT "$key_" < "$TmpPack_Encrypted" |
			git index-pack -v --stdin >/dev/null
		# add to local pack list
		xecho "$Packpfx$pack_" >> "$Localdir/have_packs$GITCEPTION"
	done

	rm -f "$TmpPack_Encrypted"
	trap - EXIT
	echo_git # end with blank line
}

# do_push PUSHARGS (multiple lines like +src:dst, with both + and src opt.)
do_push()
{
	# Security protocol:
	# Each git packfile is encrypted and then named for the encrypted
	# file's hash. The manifest is updated with the pack id.
	# The manifest is encrypted.
	local remote_has= remote_want= prefix_= suffix_= line_= pack_id= key_=
	del_hash=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

	ensure_connected

	if iseq "$Did_find_repo" "no"
	then
		make_new_repo
	fi

	trap 'rm -f "$TmpPack_Encrypted" "$TmpObjlist" "$TmpManifest_Enc"' EXIT
	if isnonnull "$Branchlist"
	then
		# filter through batch-check to mark only the commits we have
		remote_has=$(xecho "$Branchlist" | cut -f 1 -d ' ' |
			safe_git_rev_parse | sed -e 's/^\(.\)/^&/')
	fi

	while read line_ # from <<
	do
		# +src:dst -- remove leading + then split at :
		splitcolon "${line_#+}"
		if isnonnull "$prefix_"
		then
			remote_want=$(append "$remote_want" "$prefix_")
			Branchlist=$(append "$Branchlist" \
			"$(xecho "$prefix_" | safe_git_rev_parse) $suffix_")
		else
			# Mark branch for deletion
			Branchlist=$(append "$Branchlist" "$del_hash $suffix_")
		fi
	done <<EOF
$1
EOF

	Branchlist=$(xecho "$Branchlist" | sort_stable_k2 | tac | uniq -s 40 |
		xgrep -v "^$del_hash")

	TmpPack_Encrypted="$Localdir/tmp_pack_ENCRYPTED_.$$"
	TmpObjlist="$Localdir/tmp_packrevlist.$$"
	key_=$(genkey "$Packkey_bytes")

	append "$remote_has" "$remote_want" |
		git rev-list --objects --stdin -- |
		tee "$TmpObjlist" |
		git pack-objects --stdout | ENCRYPT "$key_">"$TmpPack_Encrypted"
	# Only send pack if we have any objects to send
	if [ -s "$TmpObjlist" ]
	then
		pack_id=$(pack_hash < "$TmpPack_Encrypted")
		did_repack=
		repack_if_needed "$pack_id" "$key_"

		if isnoteq "$did_repack" yes
		then
			Packlist=$(append "$Packlist" "$Packpfx$pack_id $key_")
		fi
		# else, repack rewrote Packlist

	fi

	# Generate manifest
	echo_info "Participants are: $Signers"
	echo_info "Requesting manifest signature"

	TmpManifest_Enc="$Localdir/manifest.$$"

	(xecho "$Branchlist"; xecho "$Packlist"; xecho "$Keeplist";
	repoidstr; xecho "$Extension_list") |
		PRIVENCRYPT "$Recipients" > "$TmpManifest_Enc"

	# Upload pack
	if [ -s "$TmpObjlist" ]
	then
		PUT "$URL" "$pack_id" "$TmpPack_Encrypted"
	fi
	rm -f "$TmpPack_Encrypted"
	rm -f "$TmpObjlist"

	# Upload manifest
	PUT "$URL" "$Repoid" "$TmpManifest_Enc"

	# Delete packs
	if isnonnull "$Packlist_delete"; then
		REMOVE "$URL" "$(xecho "$Packlist_delete" | while read packline_
		do
			isnonnull "$packline_" || continue
			pack_=${packline_#"$Packpfx"}
			xecho "$pack_"
		done)"
	fi

	PUT_FINAL "$URL"

	rm -f "$TmpManifest_Enc"
	trap - EXIT

	# ok all updates
	xecho "$1" | while read line_
	do
		# +src:dst -- remove leading + then split at :
		splitcolon "${line_#+}"
		echo_git "ok $suffix_"
	done
	
	echo_git 
}

NAME=$1  # Remote name
URL=$2   # Remote URL

mkdir -p "$Localdir"

while read Input
do
	case "$Input" in
	capabilities)
		do_capabilities
		;;
	list|list\ for-push)
		do_list
		;;
	fetch\ *)
		args_="${Input##fetch }"
		while read InputX
		do
			case "$InputX" in
			fetch*)
			args_= #ignored
			;;
			*)
			break
			;;
			esac
		done
		do_fetch "$args_"
		;;
	push\ *)
		args_="${Input##push }"
		while read InputX
		do
			case "$InputX" in
			push\ *)
			args_=$(append "$args_" "${InputX#push }")
			;;
			*)
			break
			;;
			esac
		done
		do_push "$args_"
		;;
	?*)
		echo_die "Unknown input!"
		;;
	*)
		CLEAN_FINAL "$URL"
		exit 0
		;;
	esac
done