Update sharness

4 files changed, 842 insertions, 556 deletions

diff --git a/test/Makefile b/test/Makefile
index 16da562..d584e11 100644
--- a/test/Makefile
+++ b/test/Makefile
@@ -17,7 +17,7 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see http://www.gnu.org/licenses/ .
 
-SHELL := /bin/bash
+SHELL ?= /bin/bash
 SHELL_PATH ?= $(SHELL)
 SHELL_PATH_SQ = $(subst ','\'',$(SHELL_PATH))
 RM ?= rm -f
diff --git a/test/aggregate-results.sh b/test/aggregate-results.sh
index 958cbea..cfb766e 100755
--- a/test/aggregate-results.sh
+++ b/test/aggregate-results.sh
@@ -22,36 +22,36 @@ failed=0
 broken=0
 total=0
 
-while read file; do
-	while read type value; do
+while read -r file; do
+	while read -r type value; do
 		case $type in
 		'')
 			continue ;;
 		fixed)
-			fixed=$(($fixed + $value)) ;;
+			fixed=$((fixed + value)) ;;
 		success)
-			success=$(($success + $value)) ;;
+			success=$((success + value)) ;;
 		failed)
-			failed=$(($failed + $value))
-			if test $value != 0; then
+			failed=$((failed + value))
+			if test "$value" != 0; then
 				test_name=$(expr "$file" : 'test-results/\(.*\)\.[0-9]*\.counts')
 				failed_tests="$failed_tests $test_name"
 			fi
 			;;
 		broken)
-			broken=$(($broken + $value)) ;;
+			broken=$((broken + value)) ;;
 		total)
-			total=$(($total + $value)) ;;
+			total=$((total + value)) ;;
 		esac
 	done <"$file"
 done
 
 if test -n "$failed_tests"; then
-	printf "\nfailed test(s):$failed_tests\n\n"
+	printf '\nfailed test(s):%s\n\n' "$failed_tests"
 fi
 
-printf "%-8s%d\n" fixed $fixed
-printf "%-8s%d\n" success $success
-printf "%-8s%d\n" failed $failed
-printf "%-8s%d\n" broken $broken
-printf "%-8s%d\n" total $total
+printf '%-8s%d\n' fixed $fixed
+printf '%-8s%d\n' success $success
+printf '%-8s%d\n' failed $failed
+printf '%-8s%d\n' broken $broken
+printf '%-8s%d\n' total $total
diff --git a/test/lib-sharness/functions.sh b/test/lib-sharness/functions.sh
new file mode 100644
index 0000000..56401d8
--- /dev/null
+++ b/test/lib-sharness/functions.sh
@@ -0,0 +1,593 @@
+# Library of functions shared by all tests scripts, included by
+# sharness.sh.
+#
+# Copyright (c) 2005-2019 Junio C Hamano
+# Copyright (c) 2005-2019 Git project
+# Copyright (c) 2011-2019 Mathias Lafeldt
+# Copyright (c) 2015-2019 Christian Couder
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see http://www.gnu.org/licenses/ .
+
+# These assignments are to make shellcheck happy. They should be
+# removed when we can use a new version of shellcheck that contains:
+# https://github.com/koalaman/shellcheck/pull/1553
+: "${debug:=}"
+: "${verbose:=}"
+: "${this_test:=}"
+: "${skip_all:=}"
+: "${EXIT_OK:=}"
+: "${test_failure:=0}"
+: "${test_fixed:=0}"
+: "${test_broken:=0}"
+: "${test_success:=0}"
+
+# Public: Define that a test prerequisite is available.
+#
+# The prerequisite can later be checked explicitly using test_have_prereq or
+# implicitly by specifying the prerequisite name in calls to test_expect_success
+# or test_expect_failure.
+#
+# $1 - Name of prerequisite (a simple word, in all capital letters by convention)
+#
+# Examples
+#
+#   # Set PYTHON prerequisite if interpreter is available.
+#   command -v python >/dev/null && test_set_prereq PYTHON
+#
+#   # Set prerequisite depending on some variable.
+#   test -z "$NO_GETTEXT" && test_set_prereq GETTEXT
+#
+# Returns nothing.
+test_set_prereq() {
+	satisfied_prereq="$satisfied_prereq$1 "
+}
+satisfied_prereq=" "
+
+# Public: Check if one or more test prerequisites are defined.
+#
+# The prerequisites must have previously been set with test_set_prereq.
+# The most common use of this is to skip all the tests if some essential
+# prerequisite is missing.
+#
+# $1 - Comma-separated list of test prerequisites.
+#
+# Examples
+#
+#   # Skip all remaining tests if prerequisite is not set.
+#   if ! test_have_prereq PERL; then
+#       skip_all='skipping perl interface tests, perl not available'
+#       test_done
+#   fi
+#
+# Returns 0 if all prerequisites are defined or 1 otherwise.
+test_have_prereq() {
+	# prerequisites can be concatenated with ','
+	save_IFS=$IFS
+	IFS=,
+	set -- $@
+	IFS=$save_IFS
+
+	total_prereq=0
+	ok_prereq=0
+	missing_prereq=
+
+	for prerequisite; do
+		case "$prerequisite" in
+		!*)
+			negative_prereq=t
+			prerequisite=${prerequisite#!}
+			;;
+		*)
+			negative_prereq=
+		esac
+
+		total_prereq=$((total_prereq + 1))
+		case "$satisfied_prereq" in
+		*" $prerequisite "*)
+			satisfied_this_prereq=t
+			;;
+		*)
+			satisfied_this_prereq=
+		esac
+
+		case "$satisfied_this_prereq,$negative_prereq" in
+		t,|,t)
+			ok_prereq=$((ok_prereq + 1))
+			;;
+		*)
+			# Keep a list of missing prerequisites; restore
+			# the negative marker if necessary.
+			prerequisite=${negative_prereq:+!}$prerequisite
+			if test -z "$missing_prereq"; then
+				missing_prereq=$prerequisite
+			else
+				missing_prereq="$prerequisite,$missing_prereq"
+			fi
+		esac
+	done
+
+	test $total_prereq = $ok_prereq
+}
+
+# Public: Execute commands in debug mode.
+#
+# Takes a single argument and evaluates it only when the test script is started
+# with --debug. This is primarily meant for use during the development of test
+# scripts.
+#
+# $1 - Commands to be executed.
+#
+# Examples
+#
+#   test_debug "cat some_log_file"
+#
+# Returns the exit code of the last command executed in debug mode or 0
+#   otherwise.
+test_debug() {
+	test "$debug" = "" || eval "$1"
+}
+
+# Public: Stop execution and start a shell.
+#
+# This is useful for debugging tests and only makes sense together with "-v".
+# Be sure to remove all invocations of this command before submitting.
+test_pause() {
+	if test "$verbose" = t; then
+		"$SHELL_PATH" <&6 >&3 2>&4
+	else
+		error >&5 "test_pause requires --verbose"
+	fi
+}
+
+# Public: Run test commands and expect them to succeed.
+#
+# When the test passed, an "ok" message is printed and the number of successful
+# tests is incremented. When it failed, a "not ok" message is printed and the
+# number of failed tests is incremented.
+#
+# With --immediate, exit test immediately upon the first failed test.
+#
+# Usually takes two arguments:
+# $1 - Test description
+# $2 - Commands to be executed.
+#
+# With three arguments, the first will be taken to be a prerequisite:
+# $1 - Comma-separated list of test prerequisites. The test will be skipped if
+#      not all of the given prerequisites are set. To negate a prerequisite,
+#      put a "!" in front of it.
+# $2 - Test description
+# $3 - Commands to be executed.
+#
+# Examples
+#
+#   test_expect_success \
+#       'git-write-tree should be able to write an empty tree.' \
+#       'tree=$(git-write-tree)'
+#
+#   # Test depending on one prerequisite.
+#   test_expect_success TTY 'git --paginate rev-list uses a pager' \
+#       ' ... '
+#
+#   # Multiple prerequisites are separated by a comma.
+#   test_expect_success PERL,PYTHON 'yo dawg' \
+#       ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
+#
+# Returns nothing.
+test_expect_success() {
+	test "$#" = 3 && { test_prereq=$1; shift; } || test_prereq=
+	test "$#" = 2 || error "bug in the test script: not 2 or 3 parameters to test_expect_success"
+	export test_prereq
+	if ! test_skip_ "$@"; then
+		say >&3 "expecting success: $2"
+		if test_run_ "$2"; then
+			test_ok_ "$1"
+		else
+			test_failure_ "$@"
+		fi
+	fi
+	echo >&3 ""
+}
+
+# Public: Run test commands and expect them to fail. Used to demonstrate a known
+# breakage.
+#
+# This is NOT the opposite of test_expect_success, but rather used to mark a
+# test that demonstrates a known breakage.
+#
+# When the test passed, an "ok" message is printed and the number of fixed tests
+# is incremented. When it failed, a "not ok" message is printed and the number
+# of tests still broken is incremented.
+#
+# Failures from these tests won't cause --immediate to stop.
+#
+# Usually takes two arguments:
+# $1 - Test description
+# $2 - Commands to be executed.
+#
+# With three arguments, the first will be taken to be a prerequisite:
+# $1 - Comma-separated list of test prerequisites. The test will be skipped if
+#      not all of the given prerequisites are set. To negate a prerequisite,
+#      put a "!" in front of it.
+# $2 - Test description
+# $3 - Commands to be executed.
+#
+# Returns nothing.
+test_expect_failure() {
+	test "$#" = 3 && { test_prereq=$1; shift; } || test_prereq=
+	test "$#" = 2 || error "bug in the test script: not 2 or 3 parameters to test_expect_failure"
+	export test_prereq
+	if ! test_skip_ "$@"; then
+		say >&3 "checking known breakage: $2"
+		if test_run_ "$2" expecting_failure; then
+			test_known_broken_ok_ "$1"
+		else
+			test_known_broken_failure_ "$1"
+		fi
+	fi
+	echo >&3 ""
+}
+
+# Public: Run test commands and expect anything from them. Used when a
+# test is not stable or not finished for some reason.
+#
+# When the test passed, an "ok" message is printed, but the number of
+# fixed tests is not incremented.
+#
+# When it failed, a "not ok ... # TODO known breakage" message is
+# printed, and the number of tests still broken is incremented.
+#
+# Failures from these tests won't cause --immediate to stop.
+#
+# Usually takes two arguments:
+# $1 - Test description
+# $2 - Commands to be executed.
+#
+# With three arguments, the first will be taken to be a prerequisite:
+# $1 - Comma-separated list of test prerequisites. The test will be skipped if
+#      not all of the given prerequisites are set. To negate a prerequisite,
+#      put a "!" in front of it.
+# $2 - Test description
+# $3 - Commands to be executed.
+#
+# Returns nothing.
+test_expect_unstable() {
+	test "$#" = 3 && { test_prereq=$1; shift; } || test_prereq=
+	test "$#" = 2 || error "bug in the test script: not 2 or 3 parameters to test_expect_unstable"
+	export test_prereq
+	if ! test_skip_ "$@"; then
+		say >&3 "checking unstable test: $2"
+		if test_run_ "$2" unstable; then
+			test_ok_ "$1"
+		else
+			test_known_broken_failure_ "$1"
+		fi
+	fi
+	echo >&3 ""
+}
+
+# Public: Run command and ensure that it fails in a controlled way.
+#
+# Use it instead of "! <command>". For example, when <command> dies due to a
+# segfault, test_must_fail diagnoses it as an error, while "! <command>" would
+# mistakenly be treated as just another expected failure.
+#
+# This is one of the prefix functions to be used inside test_expect_success or
+# test_expect_failure.
+#
+# $1.. - Command to be executed.
+#
+# Examples
+#
+#   test_expect_success 'complain and die' '
+#       do something &&
+#       do something else &&
+#       test_must_fail git checkout ../outerspace
+#   '
+#
+# Returns 1 if the command succeeded (exit code 0).
+# Returns 1 if the command died by signal (exit codes 130-192)
+# Returns 1 if the command could not be found (exit code 127).
+# Returns 0 otherwise.
+test_must_fail() {
+	"$@"
+	exit_code=$?
+	if test $exit_code = 0; then
+		echo >&2 "test_must_fail: command succeeded: $*"
+		return 1
+	elif test $exit_code -gt 129 -a $exit_code -le 192; then
+		echo >&2 "test_must_fail: died by signal: $*"
+		return 1
+	elif test $exit_code = 127; then
+		echo >&2 "test_must_fail: command not found: $*"
+		return 1
+	fi
+	return 0
+}
+
+# Public: Run command and ensure that it succeeds or fails in a controlled way.
+#
+# Similar to test_must_fail, but tolerates success too. Use it instead of
+# "<command> || :" to catch failures caused by a segfault, for instance.
+#
+# This is one of the prefix functions to be used inside test_expect_success or
+# test_expect_failure.
+#
+# $1.. - Command to be executed.
+#
+# Examples
+#
+#   test_expect_success 'some command works without configuration' '
+#       test_might_fail git config --unset all.configuration &&
+#       do something
+#   '
+#
+# Returns 1 if the command died by signal (exit codes 130-192)
+# Returns 1 if the command could not be found (exit code 127).
+# Returns 0 otherwise.
+test_might_fail() {
+	"$@"
+	exit_code=$?
+	if test $exit_code -gt 129 -a $exit_code -le 192; then
+		echo >&2 "test_might_fail: died by signal: $*"
+		return 1
+	elif test $exit_code = 127; then
+		echo >&2 "test_might_fail: command not found: $*"
+		return 1
+	fi
+	return 0
+}
+
+# Public: Run command and ensure it exits with a given exit code.
+#
+# This is one of the prefix functions to be used inside test_expect_success or
+# test_expect_failure.
+#
+# $1   - Expected exit code.
+# $2.. - Command to be executed.
+#
+# Examples
+#
+#   test_expect_success 'Merge with d/f conflicts' '
+#       test_expect_code 1 git merge "merge msg" B master
+#   '
+#
+# Returns 0 if the expected exit code is returned or 1 otherwise.
+test_expect_code() {
+	want_code=$1
+	shift
+	"$@"
+	exit_code=$?
+	if test "$exit_code" = "$want_code"; then
+		return 0
+	fi
+
+	echo >&2 "test_expect_code: command exited with $exit_code, we wanted $want_code $*"
+	return 1
+}
+
+# Public: Compare two files to see if expected output matches actual output.
+#
+# The TEST_CMP variable defines the command used for the comparison; it
+# defaults to "diff -u". Only when the test script was started with --verbose,
+# will the command's output, the diff, be printed to the standard output.
+#
+# This is one of the prefix functions to be used inside test_expect_success or
+# test_expect_failure.
+#
+# $1 - Path to file with expected output.
+# $2 - Path to file with actual output.
+#
+# Examples
+#
+#   test_expect_success 'foo works' '
+#       echo expected >expected &&
+#       foo >actual &&
+#       test_cmp expected actual
+#   '
+#
+# Returns the exit code of the command set by TEST_CMP.
+test_cmp() {
+	${TEST_CMP:-diff -u} "$@"
+}
+
+# Public: portably print a sequence of numbers.
+#
+# seq is not in POSIX and GNU seq might not be available everywhere,
+# so it is nice to have a seq implementation, even a very simple one.
+#
+# $1 - Starting number.
+# $2 - Ending number.
+#
+# Examples
+#
+#   test_expect_success 'foo works 10 times' '
+#       for i in $(test_seq 1 10)
+#       do
+#           foo || return
+#       done
+#   '
+#
+# Returns 0 if all the specified numbers can be displayed.
+test_seq() {
+	i="$1"
+	j="$2"
+	while test "$i" -le "$j"
+	do
+		echo "$i" || return
+		i=$(("$i" + 1))
+	done
+}
+
+# Public: Check if the file expected to be empty is indeed empty, and barfs
+# otherwise.
+#
+# $1 - File to check for emptiness.
+#
+# Returns 0 if file is empty, 1 otherwise.
+test_must_be_empty() {
+	if test -s "$1"
+	then
+		echo "'$1' is not empty, it contains:"
+		cat "$1"
+		return 1
+	fi
+}
+
+# debugging-friendly alternatives to "test [-f|-d|-e]"
+# The commands test the existence or non-existence of $1. $2 can be
+# given to provide a more precise diagnosis.
+test_path_is_file () {
+	if ! test -f "$1"
+	then
+		echo "File $1 doesn't exist. $2"
+		false
+	fi
+}
+
+test_path_is_dir () {
+	if ! test -d "$1"
+	then
+		echo "Directory $1 doesn't exist. $2"
+		false
+	fi
+}
+
+# Check if the directory exists and is empty as expected, barf otherwise.
+test_dir_is_empty () {
+	test_path_is_dir "$1" &&
+	if test -n "$(find "$1" -mindepth 1 -maxdepth 1)"
+	then
+		echo "Directory '$1' is not empty, it contains:"
+		ls -la "$1"
+		return 1
+	fi
+}
+
+# Public: Schedule cleanup commands to be run unconditionally at the end of a
+# test.
+#
+# If some cleanup command fails, the test will not pass. With --immediate, no
+# cleanup is done to help diagnose what went wrong.
+#
+# This is one of the prefix functions to be used inside test_expect_success or
+# test_expect_failure.
+#
+# $1.. - Commands to prepend to the list of cleanup commands.
+#
+# Examples
+#
+#   test_expect_success 'test core.capslock' '
+#       git config core.capslock true &&
+#       test_when_finished "git config --unset core.capslock" &&
+#       do_something
+#   '
+#
+# Returns the exit code of the last cleanup command executed.
+test_when_finished() {
+	test_cleanup="{ $*
+		} && (exit \"\$eval_ret\"); eval_ret=\$?; $test_cleanup"
+}
+
+# Public: Schedule cleanup commands to be run unconditionally when all tests
+# have run.
+#
+# This can be used to clean up things like test databases. It is not needed to
+# clean up temporary files, as test_done already does that.
+#
+# Examples:
+#
+#   cleanup mysql -e "DROP DATABASE mytest"
+#
+# Returns the exit code of the last cleanup command executed.
+final_cleanup=
+cleanup() {
+	final_cleanup="{ $*
+		} && (exit \"\$eval_ret\"); eval_ret=\$?; $final_cleanup"
+}
+
+# Public: Summarize test results and exit with an appropriate error code.
+#
+# Must be called at the end of each test script.
+#
+# Can also be used to stop tests early and skip all remaining tests. For this,
+# set skip_all to a string explaining why the tests were skipped before calling
+# test_done.
+#
+# Examples
+#
+#   # Each test script must call test_done at the end.
+#   test_done
+#
+#   # Skip all remaining tests if prerequisite is not set.
+#   if ! test_have_prereq PERL; then
+#       skip_all='skipping perl interface tests, perl not available'
+#       test_done
+#   fi
+#
+# Returns 0 if all tests passed or 1 if there was a failure.
+test_done() {
+	EXIT_OK=t
+
+	if test -z "$HARNESS_ACTIVE"; then
+		test_results_dir="$SHARNESS_TEST_OUTDIR/test-results"
+		mkdir -p "$test_results_dir"
+		test_results_path="$test_results_dir/$this_test.$$.counts"
+
+		cat >>"$test_results_path" <<-EOF
+		total $SHARNESS_TEST_NB
+		success $test_success
+		fixed $test_fixed
+		broken $test_broken
+		failed $test_failure
+
+		EOF
+	fi
+
+	if test "$test_fixed" != 0; then
+		say_color error "# $test_fixed known breakage(s) vanished; please update test(s)"
+	fi
+	if test "$test_broken" != 0; then
+		say_color warn "# still have $test_broken known breakage(s)"
+	fi
+	if test "$test_broken" != 0 || test "$test_fixed" != 0; then
+		test_remaining=$((SHARNESS_TEST_NB - test_broken - test_fixed))
+		msg="remaining $test_remaining test(s)"
+	else
+		test_remaining=$SHARNESS_TEST_NB
+		msg="$SHARNESS_TEST_NB test(s)"
+	fi
+
+	case "$test_failure" in
+	0)
+		# Maybe print SKIP message
+		check_skip_all_
+		if test "$test_remaining" -gt 0; then
+			say_color pass "# passed all $msg"
+		fi
+		say "1..$SHARNESS_TEST_NB$skip_all"
+
+		test_eval_ "$final_cleanup"
+
+		remove_trash_
+
+		exit 0 ;;
+
+	*)
+		say_color error "# failed $test_failure among $msg"
+		say "1..$SHARNESS_TEST_NB"
+
+		exit 1 ;;
+
+	esac
+}
diff --git a/test/sharness.sh b/test/sharness.sh
index 1d57ce9..a3d8815 100644
--- a/test/sharness.sh
+++ b/test/sharness.sh
@@ -1,4 +1,4 @@
-#!/bin/sh
+# Sharness test framework.
 #
 # Copyright (c) 2011-2012 Mathias Lafeldt
 # Copyright (c) 2005-2012 Git project
@@ -18,14 +18,43 @@
 # along with this program.  If not, see http://www.gnu.org/licenses/ .
 
 # Public: Current version of Sharness.
-SHARNESS_VERSION="1.0.0"
+SHARNESS_VERSION="1.1.0"
 export SHARNESS_VERSION
 
+: "${SHARNESS_TEST_EXTENSION:=t}"
 # Public: The file extension for tests.  By default, it is set to "t".
-: ${SHARNESS_TEST_EXTENSION:=t}
 export SHARNESS_TEST_EXTENSION
 
-#  Reset TERM to original terminal if found, otherwise save orignal TERM
+if test -z "$SHARNESS_TEST_DIRECTORY"
+then
+	SHARNESS_TEST_DIRECTORY=$(pwd)
+else
+	# ensure that SHARNESS_TEST_DIRECTORY is an absolute path so that it
+	# is valid even if the current working directory is changed
+	SHARNESS_TEST_DIRECTORY=$(cd "$SHARNESS_TEST_DIRECTORY" && pwd) || exit 1
+fi
+# Public: Root directory containing tests. Tests can override this variable,
+# e.g. for testing Sharness itself.
+export SHARNESS_TEST_DIRECTORY
+
+: "${SHARNESS_TEST_SRCDIR:=$(cd "$(dirname "$0")" && pwd)}"
+# Public: Source directory of test code and sharness library.
+# This directory may be different from the directory in which tests are
+# being run.
+export SHARNESS_TEST_SRCDIR
+
+if test -z "$SHARNESS_TEST_OUTDIR"
+then
+	# Similarly, override this to store the test-results subdir
+	# elsewhere
+	SHARNESS_TEST_OUTDIR=$SHARNESS_TEST_DIRECTORY
+fi
+
+# Public: Directory where the output of the tests should be stored (i.e.
+# trash directories).
+export SHARNESS_TEST_OUTDIR
+
+#  Reset TERM to original terminal if found, otherwise save original TERM
 [ "x" = "x$SHARNESS_ORIG_TERM" ] &&
 		SHARNESS_ORIG_TERM="$TERM" ||
 		TERM="$SHARNESS_ORIG_TERM"
@@ -33,23 +62,45 @@ export SHARNESS_TEST_EXTENSION
 export SHARNESS_ORIG_TERM
 
 # Export SHELL_PATH
-: ${SHELL_PATH:=$SHELL}
+: "${SHELL_PATH:=/bin/sh}"
 export SHELL_PATH
 
+# if --tee was passed, write the output not only to the terminal, but
+# additionally to the file test-results/$BASENAME.out, too.
+case "$SHARNESS_TEST_TEE_STARTED, $* " in
+done,*)
+	# do not redirect again
+	;;
+*' --tee '*|*' --verbose-log '*)
+	mkdir -p "$SHARNESS_TEST_OUTDIR/test-results"
+	BASE="$SHARNESS_TEST_OUTDIR/test-results/$(basename "$0" ".$SHARNESS_TEST_EXTENSION")"
+
+	# Make this filename available to the sub-process in case it is using
+	# --verbose-log.
+	SHARNESS_TEST_TEE_OUTPUT_FILE="$BASE.out"
+	export SHARNESS_TEST_TEE_OUTPUT_FILE
+
+	# Truncate before calling "tee -a" to get rid of the results
+	# from any previous runs.
+	: >"$SHARNESS_TEST_TEE_OUTPUT_FILE"
+
+	(SHARNESS_TEST_TEE_STARTED="done" ${SHELL_PATH} "$0" "$@" 2>&1;
+	 echo $? >"$BASE.exit") | tee -a "$SHARNESS_TEST_TEE_OUTPUT_FILE"
+	test "$(cat "$BASE.exit")" = 0
+	exit
+	;;
+esac
+
 # For repeatability, reset the environment to a known state.
 # TERM is sanitized below, after saving color control sequences.
 LANG=C
 LC_ALL=C
-PAGER=cat
+PAGER="cat"
 TZ=UTC
 EDITOR=:
 export LANG LC_ALL PAGER TZ EDITOR
 unset VISUAL CDPATH GREP_OPTIONS
 
-# Line feed
-LF='
-'
-
 [ "x$TERM" != "xdumb" ] && (
 		[ -t 1 ] &&
 		tput bold >/dev/null 2>&1 &&
@@ -82,9 +133,17 @@ while test "$#" -ne 0; do
 		chain_lint=; shift ;;
 	--no-color)
 		color=; shift ;;
+	--tee)
+		shift ;; # was handled already
 	--root=*)
 		root=$(expr "z$1" : 'z[^=]*=\(.*\)')
 		shift ;;
+	-x)
+		trace=t
+		shift ;;
+	--verbose-log)
+		verbose_log=t
+		shift ;;
 	*)
 		echo "error: unknown test option '$1'" >&2; exit 1 ;;
 	esac
@@ -109,21 +168,62 @@ if test -n "$color"; then
 	say_color_pass=$(tput setaf 2) # green
 	say_color_info=$(tput setaf 6) # cyan
 	say_color_reset=$(tput sgr0)
-	say_color_="" # no formatting for normal text
+	say_color_raw="" # no formatting for normal text
 	say_color() {
 		test -z "$1" && test -n "$quiet" && return
-		eval "say_color_color=\$say_color_$1"
+		case "$1" in
+			error) say_color_color=$say_color_error ;;
+			skip) say_color_color=$say_color_skip ;;
+			warn) say_color_color=$say_color_warn ;;
+			pass) say_color_color=$say_color_pass ;;
+			info) say_color_color=$say_color_info ;;
+			*) say_color_color=$say_color_raw ;;
+		esac
 		shift
-		printf "%s\\n" "$say_color_color$*$say_color_reset"
+		printf '%s%s%s\n' "$say_color_color" "$*" "$say_color_reset"
 	}
 else
 	say_color() {
 		test -z "$1" && test -n "$quiet" && return
 		shift
-		printf "%s\n" "$*"
+		printf '%s\n' "$*"
 	}
 fi
 
+: "${test_untraceable:=}"
+# Public: When set to a non-empty value, the current test will not be
+# traced, unless it's run with a Bash version supporting
+# BASH_XTRACEFD, i.e. v4.1 or later.
+export test_untraceable
+
+if test -n "$trace" && test -n "$test_untraceable"
+then
+	# '-x' tracing requested, but this test script can't be reliably
+	# traced, unless it is run with a Bash version supporting
+	# BASH_XTRACEFD (introduced in Bash v4.1).
+	#
+	# Perform this version check _after_ the test script was
+	# potentially re-executed with $TEST_SHELL_PATH for '--tee' or
+	# '--verbose-log', so the right shell is checked and the
+	# warning is issued only once.
+	if test -n "$BASH_VERSION" && eval '
+	     test ${BASH_VERSINFO[0]} -gt 4 || {
+	       test ${BASH_VERSINFO[0]} -eq 4 &&
+	       test ${BASH_VERSINFO[1]} -ge 1
+	     }
+	   '
+	then
+		: Executed by a Bash version supporting BASH_XTRACEFD.  Good.
+	else
+		echo >&2 "warning: ignoring -x; '$0' is untraceable without BASH_XTRACEFD"
+		trace=
+	fi
+fi
+if test -n "$trace" && test -z "$verbose_log"
+then
+	verbose=t
+fi
+
 TERM=dumb
 export TERM
 
@@ -137,7 +237,7 @@ say() {
 	say_color info "$*"
 }
 
-test -n "$test_description" || error "Test script did not set test_description."
+test -n "${test_description:-}" || error "Test script did not set test_description."
 
 if test "$help" = "t"; then
 	echo "$test_description"
@@ -146,17 +246,32 @@ fi
 
 exec 5>&1
 exec 6<&0
-if test "$verbose" = "t"; then
+if test "$verbose_log" = "t"
+then
+	exec 3>>"$SHARNESS_TEST_TEE_OUTPUT_FILE" 4>&3
+elif test "$verbose" = "t"
+then
 	exec 4>&2 3>&1
 else
 	exec 4>/dev/null 3>/dev/null
 fi
 
-test_failure=0
-test_count=0
-test_fixed=0
-test_broken=0
-test_success=0
+# Send any "-x" output directly to stderr to avoid polluting tests
+# which capture stderr. We can do this unconditionally since it
+# has no effect if tracing isn't turned on.
+#
+# Note that this sets up the trace fd as soon as we assign the variable, so it
+# must come after the creation of descriptor 4 above. Likewise, we must never
+# unset this, as it has the side effect of closing descriptor 4, which we
+# use to show verbose tests to the user.
+#
+# Note also that we don't need or want to export it. The tracing is local to
+# this shell, and we would not want to influence any shells we exec.
+BASH_XTRACEFD=4
+
+# Public: The current test number, starting at 0.
+SHARNESS_TEST_NB=0
+export SHARNESS_TEST_NB
 
 die() {
 	code=$?
@@ -171,159 +286,104 @@ die() {
 EXIT_OK=
 trap 'die' EXIT
 
-# Public: Define that a test prerequisite is available.
-#
-# The prerequisite can later be checked explicitly using test_have_prereq or
-# implicitly by specifying the prerequisite name in calls to test_expect_success
-# or test_expect_failure.
-#
-# $1 - Name of prerequiste (a simple word, in all capital letters by convention)
-#
-# Examples
-#
-#   # Set PYTHON prerequisite if interpreter is available.
-#   command -v python >/dev/null && test_set_prereq PYTHON
-#
-#   # Set prerequisite depending on some variable.
-#   test -z "$NO_GETTEXT" && test_set_prereq GETTEXT
-#
-# Returns nothing.
-test_set_prereq() {
-	satisfied_prereq="$satisfied_prereq$1 "
-}
-satisfied_prereq=" "
-
-# Public: Check if one or more test prerequisites are defined.
-#
-# The prerequisites must have previously been set with test_set_prereq.
-# The most common use of this is to skip all the tests if some essential
-# prerequisite is missing.
-#
-# $1 - Comma-separated list of test prerequisites.
-#
-# Examples
-#
-#   # Skip all remaining tests if prerequisite is not set.
-#   if ! test_have_prereq PERL; then
-#       skip_all='skipping perl interface tests, perl not available'
-#       test_done
-#   fi
-#
-# Returns 0 if all prerequisites are defined or 1 otherwise.
-test_have_prereq() {
-	# prerequisites can be concatenated with ','
-	save_IFS=$IFS
-	IFS=,
-	set -- $*
-	IFS=$save_IFS
-
-	total_prereq=0
-	ok_prereq=0
-	missing_prereq=
-
-	for prerequisite; do
-		case "$prerequisite" in
-		!*)
-			negative_prereq=t
-			prerequisite=${prerequisite#!}
-			;;
-		*)
-			negative_prereq=
-		esac
-
-		total_prereq=$(($total_prereq + 1))
-		case "$satisfied_prereq" in
-		*" $prerequisite "*)
-			satisfied_this_prereq=t
-			;;
-		*)
-			satisfied_this_prereq=
-		esac
+test_prereq=
+missing_prereq=
 
-		case "$satisfied_this_prereq,$negative_prereq" in
-		t,|,t)
-			ok_prereq=$(($ok_prereq + 1))
-			;;
-		*)
-			# Keep a list of missing prerequisites; restore
-			# the negative marker if necessary.
-			prerequisite=${negative_prereq:+!}$prerequisite
-			if test -z "$missing_prereq"; then
-				missing_prereq=$prerequisite
-			else
-				missing_prereq="$prerequisite,$missing_prereq"
-			fi
-		esac
-	done
+test_failure=0
+test_fixed=0
+test_broken=0
+test_success=0
 
-	test $total_prereq = $ok_prereq
-}
+. "$SHARNESS_TEST_SRCDIR/lib-sharness/functions.sh"
 
 # You are not expected to call test_ok_ and test_failure_ directly, use
 # the text_expect_* functions instead.
 
 test_ok_() {
-	test_success=$(($test_success + 1))
-	say_color "" "ok $test_count - $@"
+	test_success=$((test_success + 1))
+	say_color "" "ok $SHARNESS_TEST_NB - $*"
 }
 
 test_failure_() {
-	test_failure=$(($test_failure + 1))
-	say_color error "not ok $test_count - $1"
+	test_failure=$((test_failure + 1))
+	say_color error "not ok $SHARNESS_TEST_NB - $1"
 	shift
 	echo "$@" | sed -e 's/^/#	/'
 	test "$immediate" = "" || { EXIT_OK=t; exit 1; }
 }
 
 test_known_broken_ok_() {
-	test_fixed=$(($test_fixed + 1))
-	say_color error "ok $test_count - $@ # TODO known breakage vanished"
+	test_fixed=$((test_fixed + 1))
+	say_color error "ok $SHARNESS_TEST_NB - $* # TODO known breakage vanished"
 }
 
 test_known_broken_failure_() {
-	test_broken=$(($test_broken + 1))
-	say_color warn "not ok $test_count - $@ # TODO known breakage"
+	test_broken=$((test_broken + 1))
+	say_color warn "not ok $SHARNESS_TEST_NB - $* # TODO known breakage"
 }
 
-# Public: Execute commands in debug mode.
-#
-# Takes a single argument and evaluates it only when the test script is started
-# with --debug. This is primarily meant for use during the development of test
-# scripts.
-#
-# $1 - Commands to be executed.
-#
-# Examples
-#
-#   test_debug "cat some_log_file"
-#
-# Returns the exit code of the last command executed in debug mode or 0
-#   otherwise.
-test_debug() {
-	test "$debug" = "" || eval "$1"
+want_trace () {
+	test "$trace" = t && {
+		test "$verbose" = t || test "$verbose_log" = t
+	}
 }
 
-# Public: Stop execution and start a shell.
-#
-# This is useful for debugging tests and only makes sense together with "-v".
-# Be sure to remove all invocations of this command before submitting.
-test_pause() {
-	if test "$verbose" = t; then
-		"$SHELL_PATH" <&6 >&3 2>&4
-	else
-		error >&5 "test_pause requires --verbose"
+# This is a separate function because some tests use
+# "return" to end a test_expect_success block early
+# (and we want to make sure we run any cleanup like
+# "set +x").
+test_eval_inner_ () {
+	# Do not add anything extra (including LF) after '$*'
+	eval "
+		want_trace && set -x
+		$*"
+}
+
+test_eval_x_ () {
+	# If "-x" tracing is in effect, then we want to avoid polluting stderr
+	# with non-test commands. But once in "set -x" mode, we cannot prevent
+	# the shell from printing the "set +x" to turn it off (nor the saving
+	# of $? before that). But we can make sure that the output goes to
+	# /dev/null.
+	#
+	# There are a few subtleties here:
+	#
+	#   - we have to redirect descriptor 4 in addition to 2, to cover
+	#     BASH_XTRACEFD
+	#
+	#   - the actual eval has to come before the redirection block (since
+	#     it needs to see descriptor 4 to set up its stderr)
+	#
+	#   - likewise, any error message we print must be outside the block to
+	#     access descriptor 4
+	#
+	#   - checking $? has to come immediately after the eval, but it must
+	#     be _inside_ the block to avoid polluting the "set -x" output
+	#
+
+	test_eval_inner_ "$@" </dev/null >&3 2>&4
+	{
+		test_eval_ret_=$?
+		if want_trace
+		then
+			set +x
+		fi
+	} 2>/dev/null 4>&2
+
+	if test "$test_eval_ret_" != 0 && want_trace
+	then
+		say_color error >&4 "error: last command exited with \$?=$test_eval_ret_"
 	fi
+	return $test_eval_ret_
 }
 
 test_eval_() {
-	# This is a separate function because some tests use
-	# "return" to end a test_expect_success block early.
 	case ",$test_prereq," in
 	*,INTERACTIVE,*)
 		eval "$*"
 		;;
 	*)
-		eval </dev/null >&3 2>&4 "$*"
+		test_eval_x_ "$@"
 		;;
 	esac
 }
@@ -335,13 +395,22 @@ test_run_() {
 	eval_ret=$?
 
 	if test "$chain_lint" = "t"; then
+		# turn off tracing for this test-eval, as it simply creates
+		# confusing noise in the "-x" output
+		trace_tmp=$trace
+		trace=
+		# 117 is magic because it is unlikely to match the exit
+		# code of other programs
 		test_eval_ "(exit 117) && $1"
 		if test "$?" != 117; then
 			error "bug in the test script: broken &&-chain: $1"
 		fi
+		trace=$trace_tmp
 	fi
 
-	if test -z "$immediate" || test $eval_ret = 0 || test -n "$expecting_failure"; then
+	if test -z "$immediate" || test $eval_ret = 0 ||
+	   test -n "$expecting_failure" && test "$test_cleanup" != ":"
+	then
 		test_eval_ "$test_cleanup"
 	fi
 	if test "$verbose" = "t" && test -n "$HARNESS_ACTIVE"; then
@@ -351,10 +420,10 @@ test_run_() {
 }
 
 test_skip_() {
-	test_count=$(($test_count + 1))
+	SHARNESS_TEST_NB=$((SHARNESS_TEST_NB + 1))
 	to_skip=
 	for skp in $SKIP_TESTS; do
-		case $this_test.$test_count in
+		case $this_test.$SHARNESS_TEST_NB in
 		$skp)
 			to_skip=t
 			break
@@ -370,8 +439,8 @@ test_skip_() {
 			of_prereq=" of $test_prereq"
 		fi
 
-		say_color skip >&3 "skipping test: $@"
-		say_color skip "ok $test_count # skip $1 (missing $missing_prereq${of_prereq})"
+		say_color skip >&3 "skipping test: $*"
+		say_color skip "ok $SHARNESS_TEST_NB # skip $1 (missing $missing_prereq${of_prereq})"
 		: true
 		;;
 	*)
@@ -380,413 +449,30 @@ test_skip_() {
 	esac
 }
 
-# Public: Run test commands and expect them to succeed.
-#
-# When the test passed, an "ok" message is printed and the number of successful
-# tests is incremented. When it failed, a "not ok" message is printed and the
-# number of failed tests is incremented.
-#
-# With --immediate, exit test immediately upon the first failed test.
-#
-# Usually takes two arguments:
-# $1 - Test description
-# $2 - Commands to be executed.
-#
-# With three arguments, the first will be taken to be a prerequisite:
-# $1 - Comma-separated list of test prerequisites. The test will be skipped if
-#      not all of the given prerequisites are set. To negate a prerequisite,
-#      put a "!" in front of it.
-# $2 - Test description
-# $3 - Commands to be executed.
-#
-# Examples
-#
-#   test_expect_success \
-#       'git-write-tree should be able to write an empty tree.' \
-#       'tree=$(git-write-tree)'
-#
-#   # Test depending on one prerequisite.
-#   test_expect_success TTY 'git --paginate rev-list uses a pager' \
-#       ' ... '
-#
-#   # Multiple prerequisites are separated by a comma.
-#   test_expect_success PERL,PYTHON 'yo dawg' \
-#       ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
-#
-# Returns nothing.
-test_expect_success() {
-	test "$#" = 3 && { test_prereq=$1; shift; } || test_prereq=
-	test "$#" = 2 || error "bug in the test script: not 2 or 3 parameters to test_expect_success"
-	export test_prereq
-	if ! test_skip_ "$@"; then
-		say >&3 "expecting success: $2"
-		if test_run_ "$2"; then
-			test_ok_ "$1"
-		else
-			test_failure_ "$@"
-		fi
-	fi
-	echo >&3 ""
-}
-
-# Public: Run test commands and expect them to fail. Used to demonstrate a known
-# breakage.
-#
-# This is NOT the opposite of test_expect_success, but rather used to mark a
-# test that demonstrates a known breakage.
-#
-# When the test passed, an "ok" message is printed and the number of fixed tests
-# is incremented. When it failed, a "not ok" message is printed and the number
-# of tests still broken is incremented.
-#
-# Failures from these tests won't cause --immediate to stop.
-#
-# Usually takes two arguments:
-# $1 - Test description
-# $2 - Commands to be executed.
-#
-# With three arguments, the first will be taken to be a prerequisite:
-# $1 - Comma-separated list of test prerequisites. The test will be skipped if
-#      not all of the given prerequisites are set. To negate a prerequisite,
-#      put a "!" in front of it.
-# $2 - Test description
-# $3 - Commands to be executed.
-#
-# Returns nothing.
-test_expect_failure() {
-	test "$#" = 3 && { test_prereq=$1; shift; } || test_prereq=
-	test "$#" = 2 || error "bug in the test script: not 2 or 3 parameters to test_expect_failure"
-	export test_prereq
-	if ! test_skip_ "$@"; then
-		say >&3 "checking known breakage: $2"
-		if test_run_ "$2" expecting_failure; then
-			test_known_broken_ok_ "$1"
-		else
-			test_known_broken_failure_ "$1"
-		fi
-	fi
-	echo >&3 ""
-}
-
-# Public: Run command and ensure that it fails in a controlled way.
-#
-# Use it instead of "! <command>". For example, when <command> dies due to a
-# segfault, test_must_fail diagnoses it as an error, while "! <command>" would
-# mistakenly be treated as just another expected failure.
-#
-# This is one of the prefix functions to be used inside test_expect_success or
-# test_expect_failure.
-#
-# $1.. - Command to be executed.
-#
-# Examples
-#
-#   test_expect_success 'complain and die' '
-#       do something &&
-#       do something else &&
-#       test_must_fail git checkout ../outerspace
-#   '
-#
-# Returns 1 if the command succeeded (exit code 0).
-# Returns 1 if the command died by signal (exit codes 130-192)
-# Returns 1 if the command could not be found (exit code 127).
-# Returns 0 otherwise.
-test_must_fail() {
-	"$@"
-	exit_code=$?
-	if test $exit_code = 0; then
-		echo >&2 "test_must_fail: command succeeded: $*"
-		return 1
-	elif test $exit_code -gt 129 -a $exit_code -le 192; then
-		echo >&2 "test_must_fail: died by signal: $*"
-		return 1
-	elif test $exit_code = 127; then
-		echo >&2 "test_must_fail: command not found: $*"
-		return 1
-	fi
-	return 0
-}
-
-# Public: Run command and ensure that it succeeds or fails in a controlled way.
-#
-# Similar to test_must_fail, but tolerates success too. Use it instead of
-# "<command> || :" to catch failures caused by a segfault, for instance.
-#
-# This is one of the prefix functions to be used inside test_expect_success or
-# test_expect_failure.
-#
-# $1.. - Command to be executed.
-#
-# Examples
-#
-#   test_expect_success 'some command works without configuration' '
-#       test_might_fail git config --unset all.configuration &&
-#       do something
-#   '
-#
-# Returns 1 if the command died by signal (exit codes 130-192)
-# Returns 1 if the command could not be found (exit code 127).
-# Returns 0 otherwise.
-test_might_fail() {
-	"$@"
-	exit_code=$?
-	if test $exit_code -gt 129 -a $exit_code -le 192; then
-		echo >&2 "test_might_fail: died by signal: $*"
-		return 1
-	elif test $exit_code = 127; then
-		echo >&2 "test_might_fail: command not found: $*"
-		return 1
-	fi
-	return 0
-}
-
-# Public: Run command and ensure it exits with a given exit code.
-#
-# This is one of the prefix functions to be used inside test_expect_success or
-# test_expect_failure.
-#
-# $1   - Expected exit code.
-# $2.. - Command to be executed.
-#
-# Examples
-#
-#   test_expect_success 'Merge with d/f conflicts' '
-#       test_expect_code 1 git merge "merge msg" B master
-#   '
-#
-# Returns 0 if the expected exit code is returned or 1 otherwise.
-test_expect_code() {
-	want_code=$1
-	shift
-	"$@"
-	exit_code=$?
-	if test $exit_code = $want_code; then
-		return 0
-	fi
-
-	echo >&2 "test_expect_code: command exited with $exit_code, we wanted $want_code $*"
-	return 1
-}
-
-# Public: Compare two files to see if expected output matches actual output.
-#
-# The TEST_CMP variable defines the command used for the comparision; it
-# defaults to "diff -u". Only when the test script was started with --verbose,
-# will the command's output, the diff, be printed to the standard output.
-#
-# This is one of the prefix functions to be used inside test_expect_success or
-# test_expect_failure.
-#
-# $1 - Path to file with expected output.
-# $2 - Path to file with actual output.
-#
-# Examples
-#
-#   test_expect_success 'foo works' '
-#       echo expected >expected &&
-#       foo >actual &&
-#       test_cmp expected actual
-#   '
-#
-# Returns the exit code of the command set by TEST_CMP.
-test_cmp() {
-	${TEST_CMP:-diff -u} "$@"
-}
-
-# Public: portably print a sequence of numbers.
-#
-# seq is not in POSIX and GNU seq might not be available everywhere,
-# so it is nice to have a seq implementation, even a very simple one.
-#
-# $1 - Starting number.
-# $2 - Ending number.
-#
-# Examples
-#
-#   test_expect_success 'foo works 10 times' '
-#       for i in $(test_seq 1 10)
-#       do
-#           foo || return
-#       done
-#   '
-#
-# Returns 0 if all the specified numbers can be displayed.
-test_seq() {
-	i="$1"
-	j="$2"
-	while test "$i" -le "$j"
-	do
-		echo "$i" || return
-		i=$(expr "$i" + 1)
-	done
-}
-
-# Public: Check if the file expected to be empty is indeed empty, and barfs
-# otherwise.
-#
-# $1 - File to check for emptyness.
-#
-# Returns 0 if file is empty, 1 otherwise.
-test_must_be_empty() {
-	if test -s "$1"
-	then
-		echo "'$1' is not empty, it contains:"
-		cat "$1"
-		return 1
-	fi
-}
-
-# Public: Schedule cleanup commands to be run unconditionally at the end of a
-# test.
-#
-# If some cleanup command fails, the test will not pass. With --immediate, no
-# cleanup is done to help diagnose what went wrong.
-#
-# This is one of the prefix functions to be used inside test_expect_success or
-# test_expect_failure.
-#
-# $1.. - Commands to prepend to the list of cleanup commands.
-#
-# Examples
-#
-#   test_expect_success 'test core.capslock' '
-#       git config core.capslock true &&
-#       test_when_finished "git config --unset core.capslock" &&
-#       do_something
-#   '
-#
-# Returns the exit code of the last cleanup command executed.
-test_when_finished() {
-	test_cleanup="{ $*
-		} && (exit \"\$eval_ret\"); eval_ret=\$?; $test_cleanup"
-}
-
-# Public: Schedule cleanup commands to be run unconditionally when all tests
-# have run.
-#
-# This can be used to clean up things like test databases. It is not needed to
-# clean up temporary files, as test_done already does that.
-#
-# Examples:
-#
-#   cleanup mysql -e "DROP DATABASE mytest"
-#
-# Returns the exit code of the last cleanup command executed.
-final_cleanup=
-cleanup() {
-	final_cleanup="{ $*
-		} && (exit \"\$eval_ret\"); eval_ret=\$?; $final_cleanup"
-}
-
-# Public: Summarize test results and exit with an appropriate error code.
-#
-# Must be called at the end of each test script.
-#
-# Can also be used to stop tests early and skip all remaining tests. For this,
-# set skip_all to a string explaining why the tests were skipped before calling
-# test_done.
-#
-# Examples
-#
-#   # Each test script must call test_done at the end.
-#   test_done
-#
-#   # Skip all remaining tests if prerequisite is not set.
-#   if ! test_have_prereq PERL; then
-#       skip_all='skipping perl interface tests, perl not available'
-#       test_done
-#   fi
-#
-# Returns 0 if all tests passed or 1 if there was a failure.
-test_done() {
-	EXIT_OK=t
-
-	if test -z "$HARNESS_ACTIVE"; then
-		test_results_dir="$SHARNESS_TEST_DIRECTORY/test-results"
-		mkdir -p "$test_results_dir"
-		test_results_path="$test_results_dir/$this_test.$$.counts"
-
-		cat >>"$test_results_path" <<-EOF
-		total $test_count
-		success $test_success
-		fixed $test_fixed
-		broken $test_broken
-		failed $test_failure
-
-		EOF
-	fi
-
-	if test "$test_fixed" != 0; then
-		say_color error "# $test_fixed known breakage(s) vanished; please update test(s)"
-	fi
-	if test "$test_broken" != 0; then
-		say_color warn "# still have $test_broken known breakage(s)"
-	fi
-	if test "$test_broken" != 0 || test "$test_fixed" != 0; then
-		test_remaining=$(( $test_count - $test_broken - $test_fixed ))
-		msg="remaining $test_remaining test(s)"
-	else
-		test_remaining=$test_count
-		msg="$test_count test(s)"
-	fi
-
-	case "$test_failure" in
-	0)
-		# Maybe print SKIP message
-		if test -n "$skip_all" && test $test_count -gt 0; then
-			error "Can't use skip_all after running some tests"
-		fi
-		[ -z "$skip_all" ] || skip_all=" # SKIP $skip_all"
-
-		if test $test_remaining -gt 0; then
-			say_color pass "# passed all $msg"
-		fi
-		say "1..$test_count$skip_all"
-
-		test_eval_ "$final_cleanup"
-
-		test -d "$remove_trash" &&
-		cd "$(dirname "$remove_trash")" &&
-		rm -rf "$(basename "$remove_trash")"
-
-		exit 0 ;;
-
-	*)
-		say_color error "# failed $test_failure among $msg"
-		say "1..$test_count"
-
-		exit 1 ;;
-
-	esac
-}
-
-# Public: Root directory containing tests. Tests can override this variable,
-# e.g. for testing Sharness itself.
-: ${SHARNESS_TEST_DIRECTORY:=$(pwd)}
-export SHARNESS_TEST_DIRECTORY
-
-# Public: Source directory of test code and sharness library.
-# This directory may be different from the directory in which tests are
-# being run.
-: ${SHARNESS_TEST_SRCDIR:=$(cd $(dirname $0) && pwd)}
-export SHARNESS_TEST_SRCDIR
-
+: "${SHARNESS_BUILD_DIRECTORY:="$SHARNESS_TEST_DIRECTORY/.."}"
 # Public: Build directory that will be added to PATH. By default, it is set to
 # the parent directory of SHARNESS_TEST_DIRECTORY.
-: ${SHARNESS_BUILD_DIRECTORY:="$SHARNESS_TEST_DIRECTORY/.."}
+export SHARNESS_BUILD_DIRECTORY
 PATH="$SHARNESS_BUILD_DIRECTORY:$PATH"
-export PATH SHARNESS_BUILD_DIRECTORY
+export PATH
 
 # Public: Path to test script currently executed.
 SHARNESS_TEST_FILE="$0"
 export SHARNESS_TEST_FILE
 
+remove_trash_() {
+	test -d "$remove_trash" && (
+		cd "$(dirname "$remove_trash")" &&
+			rm -rf "$(basename "$remove_trash")"
+	)
+}
+
 # Prepare test area.
 SHARNESS_TRASH_DIRECTORY="trash directory.$(basename "$SHARNESS_TEST_FILE" ".$SHARNESS_TEST_EXTENSION")"
 test -n "$root" && SHARNESS_TRASH_DIRECTORY="$root/$SHARNESS_TRASH_DIRECTORY"
 case "$SHARNESS_TRASH_DIRECTORY" in
 /*) ;; # absolute path is good
- *) SHARNESS_TRASH_DIRECTORY="$SHARNESS_TEST_DIRECTORY/$SHARNESS_TRASH_DIRECTORY" ;;
+ *) SHARNESS_TRASH_DIRECTORY="$SHARNESS_TEST_OUTDIR/$SHARNESS_TRASH_DIRECTORY" ;;
 esac
 test "$debug" = "t" || remove_trash="$SHARNESS_TRASH_DIRECTORY"
 rm -rf "$SHARNESS_TRASH_DIRECTORY" || {
@@ -831,6 +517,13 @@ mkdir -p "$SHARNESS_TRASH_DIRECTORY" || exit 1
 # in subprocesses like git equals our $PWD (for pathname comparisons).
 cd -P "$SHARNESS_TRASH_DIRECTORY" || exit 1
 
+check_skip_all_() {
+	if test -n "$skip_all" && test $SHARNESS_TEST_NB -gt 0; then
+		error "Can't use skip_all after running some tests"
+	fi
+	[ -z "$skip_all" ] || skip_all=" # SKIP $skip_all"
+}
+
 this_test=${SHARNESS_TEST_FILE##*/}
 this_test=${this_test%.$SHARNESS_TEST_EXTENSION}
 for skp in $SKIP_TESTS; do