From 16257b2ef663d7dce3b3d9545962e02c61a2b756 Mon Sep 17 00:00:00 2001
From: "Kyle J. McKay" <mackyle@gmail.com>
Date: Sun, 25 Sep 2016 08:41:36 -0700
Subject: [PATCH] clone/update: optimize ref removal

The clone operation needs to remove any pre-existing refs before
restarting a clone and also any garbage git-svn refs after an svn
clone.

The update operation needs to remove any garbage git-svn refs after
an update and also unclean refs when switching to a clean mirror.

Calling update-ref once for each ref to be removed is slow and can
be terribly painful when thousands of refs are involved.

Instead use update-ref --stdin and process all the changes at once.

We do, however, provide a slow backwards compatibiliy function so that a
server that for some incomprehensible reason is not running at least Git
version 1.8.5 (when update-ref --stdin functionality was added) can
still operate properly.

Signed-off-by: Kyle J. McKay <mackyle@gmail.com>
---
 jobd/update.sh | 25 ++++++++++---------------
 shlib.sh       | 21 +++++++++++++++++++++
 taskd/clone.sh | 15 ++++-----------
 3 files changed, 35 insertions(+), 26 deletions(-)

diff --git a/jobd/update.sh b/jobd/update.sh
index 1e692cb..8f4b6ef 100755
--- a/jobd/update.sh
+++ b/jobd/update.sh
@@ -248,16 +248,9 @@ case "$url" in
 		chmod -R ug+rw,o+r svn
 		# git svn also leaves behind ref turds that end with @nnn
 		# We get rid of them now
-		git for-each-ref --format='%(objectname) %(refname)' | \
-		{ while read sha1 ref; do
-			case "$ref" in
-			?*@[1-9]|?*@[1-9][0-9]|?*@[1-9][0-9][0-9]|?*@[1-9][0-9][0-9][0-9]|\
-			?*@[1-9][0-9][0-9][0-9][0-9]|?*@[1-9][0-9][0-9][0-9][0-9][0-9]|\
-			?*@[1-9][0-9][0-9][0-9][0-9][0-9][0-9]|\
-			?*@[1-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9])
-				git update-ref -d "$ref"
-			esac
-		done; }
+		git for-each-ref --format='%(refname)' |
+		LC_ALL=C sed '/^..*@[1-9][0-9]*$/!d; s/^/delete /' |
+		git_updateref_stdin
 		unset GIT_ASKPASS_PASSWORD
 		;;
 	darcs://*)
@@ -318,11 +311,13 @@ case "$url" in
 			if [ -n "$nextisclean" ]; then
 				# We must manually purge the unclean refs now as even prune won't do it
 				git for-each-ref --format='%(refname)' |
-				while read -r aref; do
-					case "$aref" in refs/heads/*|refs/tags/*|refs/notes/*|refs/top-bases/*) :;; *)
-						git update-ref -d "$aref"
-					esac
-				done
+				LC_ALL=C sed \
+					-e '/^refs\/heads\//d' \
+					-e '/^refs\/tags\//d' \
+					-e '/^refs\/notes\//d' \
+					-e '/^refs\/top-bases\//d' \
+					-e 's/^/delete /' |
+				git_updateref_stdin
 			fi
 			git config --bool girocco.lastupdateclean ${nextisclean:-0}
 		fi
diff --git a/shlib.sh b/shlib.sh
index dfd5ae2..329b9d1 100644
--- a/shlib.sh
+++ b/shlib.sh
@@ -259,6 +259,27 @@ git() (
 	exec "$cfg_git_bin" "$@"
 )
 
+# Since we do not yet require at least Git 1.8.5 this is a compatibility function
+# that allows us to use git update-ref --stdin where supported and the slow shell
+# script where not, but only the "delete" operation is currently supported.
+git_updateref_stdin() {
+	if [ -n "$var_have_git_185" ]; then
+		git update-ref --stdin
+	else
+		while read -r _op _ref; do
+			case "$_op" in
+				delete)
+					git update-ref -d "$_ref"
+					;;
+				*)
+					echo "bad git_updateref_stdin op: $_op" >&2
+					exit 1
+					;;
+			esac
+		done
+	fi
+}
+
 # see comments for git() -- callers must explicitly export all variables
 # intended for the commands these functions run before calling them
 perl() { command "${var_perl_bin:-perl}" "$@"; }
diff --git a/taskd/clone.sh b/taskd/clone.sh
index d7ad402..ce7551a 100755
--- a/taskd/clone.sh
+++ b/taskd/clone.sh
@@ -149,7 +149,7 @@ cleanup_failed_clone() {
 
 	# Remove all pre-existing refs
 	rm -f packed-refs
-	eval "$(git for-each-ref --shell --format='git update-ref -d %(refname) || :')" 2>/dev/null || :
+	git for-each-ref --format='delete %(refname)' | git_updateref_stdin 2>/dev/null || :
 
 	# The initial state before a clone starts has HEAD as a symbolic-ref to master
 	git symbolic-ref HEAD refs/heads/master
@@ -280,16 +280,9 @@ case "$url" in
 		chmod -R ug+rw,o+r svn
 		# git svn also leaves behind ref turds that end with @nnn
 		# We get rid of them now
-		git for-each-ref --format='%(objectname) %(refname)' | \
-		{ while read sha1 ref; do
-			case "$ref" in
-			?*@[1-9]|?*@[1-9][0-9]|?*@[1-9][0-9][0-9]|?*@[1-9][0-9][0-9][0-9]|\
-			?*@[1-9][0-9][0-9][0-9][0-9]|?*@[1-9][0-9][0-9][0-9][0-9][0-9]|\
-			?*@[1-9][0-9][0-9][0-9][0-9][0-9][0-9]|\
-			?*@[1-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9])
-				git update-ref -d "$ref"
-			esac
-		done; }
+		git for-each-ref --format='%(refname)' |
+		LC_ALL=C sed '/^..*@[1-9][0-9]*$/!d; s/^/delete /' |
+		git_updateref_stdin
 		unset GIT_ASKPASS_PASSWORD
 		;;
 	darcs://*)
-- 
2.11.4.GIT