Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions .gitignore
Original file line number Diff line number Diff line change
@@ -1,3 +1,6 @@
cmd/tar-diff/tar-diff
cmd/tar-patch/tar-patch
.secrets
tar-diff
tar-patch
test/
10 changes: 8 additions & 2 deletions Makefile
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,8 @@ PROJECT := tar-diff
VERSION := $(shell grep -oP 'VERSION\s*=\s*"\K[^"]+' pkg/protocol/version.go)

PROJ_TARBALL := $(PROJECT)_$(VERSION).tar.gz
GOCOVERDIR := $(CURDIR)/test/coverage
TEST_DIR := $(CURDIR)/test
GOCOVERDIR := $(TEST_DIR)/coverage


.PHONY: all build clean fmt install lint test tools dist unit-test integration-test validate .install.golangci-lint
Expand Down Expand Up @@ -61,10 +62,15 @@ tools: .install.golangci-lint
clean:
rm -f tar-diff tar-patch
rm -rf $(PROJECT)_$(VERSION) $(PROJ_TARBALL)
rm -rf $(GOCOVERDIR)
rm -rf $(TEST_DIR)
Comment thread
djach7 marked this conversation as resolved.

integration-test: build
GOCOVERDIR=$(GOCOVERDIR) tests/test.sh
GOCOVERDIR=$(GOCOVERDIR) tests/test-multi-old.sh
GOCOVERDIR=$(GOCOVERDIR) tests/test-source-prefix.sh
GOCOVERDIR=$(GOCOVERDIR) tests/test-delta-paths.sh
GOCOVERDIR=$(GOCOVERDIR) tests/test-tar-errors.sh
GOCOVERDIR=$(GOCOVERDIR) tests/test-fuzzy-abs.sh
go tool covdata percent -i=$(GOCOVERDIR) -o=$(GOCOVERDIR)/integration.out


Expand Down
69 changes: 69 additions & 0 deletions tests/test-delta-paths.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,69 @@
#!/bin/bash

set -e

source tests/utils.sh

use_gnu_tar_if_available

WORKDIR=$(mktemp -d "${TMPDIR:-/tmp}/test-tardiff-delta-paths-XXXXXX")

cleanup () {
rm -rf "$WORKDIR"
}
trap cleanup EXIT

# Part A: rollsum + copyRest
TEST_DIR="$WORKDIR/rollsum"
BIG_BYTES=$((3 * 1024 * 1024))
OTHER_BYTES=$((2 * 1024 * 1024))

mkdir -p "$TEST_DIR/orig/data" "$TEST_DIR/modified/data"

head -c "$BIG_BYTES" /dev/zero >"$TEST_DIR/orig/data/big.bin"
cp -a "$TEST_DIR/orig/data/big.bin" "$TEST_DIR/modified/data/big.bin"
printf 'ROLLSUM_INTEGRATION' | dd of="$TEST_DIR/modified/data/big.bin" bs=1 seek=1500000 conv=notrunc status=none 2>/dev/null || \
printf 'ROLLSUM_INTEGRATION' | dd of="$TEST_DIR/modified/data/big.bin" bs=1 seek=1500000 conv=notrunc 2>/dev/null

head -c "$OTHER_BYTES" /dev/zero >"$TEST_DIR/orig/data/other.bin"
head -c "$OTHER_BYTES" /dev/zero | LC_ALL=C tr '\000' '\377' >"$TEST_DIR/modified/data/other.bin"

create_tar "$TEST_DIR/old.tar" "$TEST_DIR/orig"
create_tar "$TEST_DIR/new.tar" "$TEST_DIR/modified"

echo "Generating tardiff A (-max-bsdiff-size 1: rollsum + copyRest)"
./tar-diff -max-bsdiff-size 1 "$TEST_DIR/old.tar.gz" "$TEST_DIR/new.tar.bz2" "$TEST_DIR/changes.tardiff"

echo "Applying tardiff A"
./tar-patch "$TEST_DIR/changes.tardiff" "$TEST_DIR/orig" "$TEST_DIR/reconstructed.tar"

echo "Verifying A"
cmp "$TEST_DIR/reconstructed.tar" "$TEST_DIR/new.tar"

echo OK rollsum-and-copyRest

# Part B: bsdiff + qsufsort/split (no low max-bsdiff-size)
BS="$WORKDIR/bsdiff"
BSDIFF_BYTES=$((512 * 1024))

mkdir -p "$BS/orig/data" "$BS/modified/data"
if ! head -c "$BSDIFF_BYTES" /dev/urandom >"$BS/orig/data/large.bin" 2>/dev/null; then
head -c "$BSDIFF_BYTES" /dev/zero >"$BS/orig/data/large.bin"
fi
cp -a "$BS/orig/data/large.bin" "$BS/modified/data/large.bin"
printf 'PATCH_BSDIFF_INTEGRATION_0123456789' | dd of="$BS/modified/data/large.bin" bs=1 seek=200000 conv=notrunc status=none 2>/dev/null || \
printf 'PATCH_BSDIFF_INTEGRATION_0123456789' | dd of="$BS/modified/data/large.bin" bs=1 seek=200000 conv=notrunc 2>/dev/null

create_tar "$BS/old.tar" "$BS/orig"
create_tar "$BS/new.tar" "$BS/modified"

echo "Generating tardiff B (default max-bsdiff: medium file → bsdiff / qsufsort)"
./tar-diff "$BS/old.tar.gz" "$BS/new.tar.bz2" "$BS/changes.tardiff"

echo "Applying tardiff B"
./tar-patch "$BS/changes.tardiff" "$BS/orig" "$BS/reconstructed.tar"

echo "Verifying B"
cmp "$BS/reconstructed.tar" "$BS/new.tar"

echo OK bsdiff-large-old
35 changes: 35 additions & 0 deletions tests/test-fuzzy-abs.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,35 @@
#!/bin/bash

set -e

source tests/utils.sh

use_gnu_tar_if_available

TEST_DIR=$(mktemp -d "${TMPDIR:-/tmp}/test-tardiff-fuzzy-abs-XXXXXX")

cleanup () {
rm -rf "$TEST_DIR"
}
trap cleanup EXIT

mkdir -p "$TEST_DIR/orig/data"
printf 'aaaaaaaaaa' >"$TEST_DIR/orig/data/foo.a"
head -c 50 /dev/zero | tr '\0' 'b' >"$TEST_DIR/orig/data/foo.b"

mkdir -p "$TEST_DIR/modified/data"
printf 'cccccccccccccccccccc' >"$TEST_DIR/modified/data/foo.c"

create_tar "$TEST_DIR/old.tar" "$TEST_DIR/orig"
create_tar "$TEST_DIR/new.tar" "$TEST_DIR/modified"

echo "Generating tardiff (fuzzy rename: foo.a + foo.b -> foo.c)"
./tar-diff "$TEST_DIR/old.tar.gz" "$TEST_DIR/new.tar.bz2" "$TEST_DIR/changes.tardiff"

echo "Applying tardiff"
./tar-patch "$TEST_DIR/changes.tardiff" "$TEST_DIR/orig" "$TEST_DIR/reconstructed.tar"

echo "Verifying reconstruction"
cmp "$TEST_DIR/reconstructed.tar" "$TEST_DIR/new.tar"

echo OK fuzzy-abs
83 changes: 83 additions & 0 deletions tests/test-multi-old.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,83 @@
#!/bin/bash

set -e

source tests/utils.sh

use_gnu_tar_if_available

# Usage: tests/test-multi-old.sh [N]
# N defaults to MULTI_OLD_N or 3. Example: MULTI_OLD_N=6 tests/test-multi-old.sh
N="${1:-${MULTI_OLD_N:-3}}"

if [[ "$N" =~ ^[0-9]+$ ]] && ((N >= 2)); then
:
else
echo "usage: $0 [N] (integer N >= 2, or set MULTI_OLD_N)" >&2
exit 1
fi

TEST_DIR=$(mktemp -d "${TMPDIR:-/tmp}/test-tardiff-multi-old-XXXXXX")
PEEL_COUNT=$((N - 1))

cleanup () {
rm -rf "$TEST_DIR"
}
trap cleanup EXIT

# Relative paths under data/ to peel into separate old layers (order matches older N=2 case: bar first).
declare -a order=( "dir1/bar.txt" "dir1/foo.txt" "dir1/move.txt" )
declare -a PEEL=()

create_orig "$TEST_DIR/orig"

for ((i = 0; i < PEEL_COUNT; i++)); do
if ((i < ${#order[@]})); then
PEEL+=("${order[i]}")
else
k=$((i - ${#order[@]} + 1))
mkdir -p "$TEST_DIR/orig/data/dir1"
echo "multi-pad-$k" >"$TEST_DIR/orig/data/dir1/_multi_$k.txt"
PEEL+=("dir1/_multi_$k.txt")
fi
done

create_tar "$TEST_DIR/orig.tar" "$TEST_DIR/orig"
modify_orig "$TEST_DIR/modified" "$TEST_DIR/orig.tar"
create_tar "$TEST_DIR/modified.tar" "$TEST_DIR/modified"

# old1: full tree minus every peeled file
cp -dR --no-preserve=context "$TEST_DIR/orig" "$TEST_DIR/mo-base" 2>/dev/null \
|| cp -dR "$TEST_DIR/orig" "$TEST_DIR/mo-base"
for rel in "${PEEL[@]}"; do
rm -f "$TEST_DIR/mo-base/data/$rel"
done

# old2..oldN: one peeled file each
for ((i = 0; i < PEEL_COUNT; i++)); do
rel="${PEEL[i]}"
layer=$((i + 2))
sdir="$TEST_DIR/mo-slice-$layer"
mkdir -p "$sdir/data/$(dirname "$rel")"
cp "$TEST_DIR/orig/data/$rel" "$sdir/data/$rel"
create_tar "$TEST_DIR/mo-old${layer}.tar" "$sdir"
done

create_tar "$TEST_DIR/mo-old1.tar" "$TEST_DIR/mo-base"

declare -a DIFF_ARGS=()
for ((i = 1; i <= N; i++)); do
DIFF_ARGS+=("$TEST_DIR/mo-old${i}.tar.gz")
done
DIFF_ARGS+=("$TEST_DIR/modified.tar.bz2" "$TEST_DIR/mo-changes.tardiff")

echo "Generating tardiff (N=$N multi-old)"
./tar-diff "${DIFF_ARGS[@]}"

echo "Applying tardiff (N=$N)"
./tar-patch "$TEST_DIR/mo-changes.tardiff" "$TEST_DIR/orig" "$TEST_DIR/mo-reconstructed.tar"

echo "Verifying reconstruction (N=$N)"
cmp "$TEST_DIR/mo-reconstructed.tar" "$TEST_DIR/modified.tar"

echo "OK multi-old (N=$N)"
76 changes: 76 additions & 0 deletions tests/test-source-prefix.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,76 @@
#!/bin/bash


set -e

source tests/utils.sh

use_gnu_tar_if_available

TEST_DIR=$(mktemp -d "${TMPDIR:-/tmp}/test-tardiff-fuzzy-abs-XXXXXX")
TREE="$TEST_DIR/tree"

cleanup () {
rm -rf "$TEST_DIR"
}
trap cleanup EXIT

create_v1 () {
rm -rf "$TREE"
mkdir -p "$TREE/blobs/sha256" "$TREE/config" "$TREE/data"
printf '%s' 'blob-content' > "$TREE/blobs/sha256/abc123"
printf '%s' 'config-v1' > "$TREE/config/app.conf"
printf '%s' 'data-v1' > "$TREE/data/file.txt"
}

create_v2 () {
rm -rf "$TREE"
mkdir -p "$TREE/blobs/sha256" "$TREE/config" "$TREE/data"
printf '%s' 'blob-content-modified' > "$TREE/blobs/sha256/abc123"
printf '%s' 'config-v2' > "$TREE/config/app.conf"
printf '%s' 'data-v2' > "$TREE/data/file.txt"
}

create_tar () {
local file=$1
local root=$2
tar cf "$file" -C "$root" blobs config data
compress_tar "$file"
}

unpack_old_tree () {
local dest=$1
rm -rf "$dest"
mkdir -p "$dest"
tar xf "$TEST_DIR/old.tar" -C "$dest"
}

create_v1
create_tar "$TEST_DIR/old.tar" "$TREE"

create_v2
create_tar "$TEST_DIR/new.tar" "$TREE"

unpack_old_tree "$TEST_DIR/unpack1"

echo "Generating tardiff (--source-prefix=blobs/)"
./tar-diff --source-prefix=blobs/ "$TEST_DIR/old.tar.gz" "$TEST_DIR/new.tar.bz2" "$TEST_DIR/delta1.tardiff"

echo "Applying tardiff (single prefix)"
./tar-patch "$TEST_DIR/delta1.tardiff" "$TEST_DIR/unpack1" "$TEST_DIR/out1.tar"

echo "Verifying (single prefix)"
cmp "$TEST_DIR/out1.tar" "$TEST_DIR/new.tar"

unpack_old_tree "$TEST_DIR/unpack2"

echo "Generating tardiff (--source-prefix=blobs/ --source-prefix=config/)"
./tar-diff --source-prefix=blobs/ --source-prefix=config/ "$TEST_DIR/old.tar.gz" "$TEST_DIR/new.tar.bz2" "$TEST_DIR/delta2.tardiff"

echo "Applying tardiff (dual prefix)"
./tar-patch "$TEST_DIR/delta2.tardiff" "$TEST_DIR/unpack2" "$TEST_DIR/out2.tar"

echo "Verifying (dual prefix)"
cmp "$TEST_DIR/out2.tar" "$TEST_DIR/new.tar"

echo OK source-prefix
87 changes: 87 additions & 0 deletions tests/test-tar-errors.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,87 @@
#!/bin/bash

# Integration: expected failures for tar-diff and tar-patch CLIs (cover binaries).

set -e

source tests/utils.sh

use_gnu_tar_if_available

TEST_DIR=$(mktemp -d "${TMPDIR:-/tmp}/test-tardiff-fuzzy-abs-XXXXXX")

expect_fail () {
local desc="$1"
shift
set +e
"$@" &>/dev/null
local code=$?
set -e
if [[ "$code" -eq 0 ]]; then
echo "expected failure for: $desc" >&2
exit 1
fi
}

cleanup () {
rm -rf "$TEST_DIR"
}
trap cleanup EXIT

mkdir -p "$TEST_DIR/orig/data"

# tar-diff: CLI / I/O errors
expect_fail "tar-diff too few args" ./tar-diff
expect_fail "tar-diff two args only" ./tar-diff one two

dd if=/dev/urandom of="$TEST_DIR/junk.gz" bs=200 count=1 status=none 2>/dev/null || \
dd if=/dev/urandom of="$TEST_DIR/junk.gz" bs=200 count=1 2>/dev/null
expect_fail "tar-diff missing old archive" ./tar-diff "$TEST_DIR/does-not-exist.gz" "$TEST_DIR/junk.gz" "$TEST_DIR/delta-x.tardiff"

mkdir -p "$TEST_DIR/td-a/data" "$TEST_DIR/td-b/data"
echo x >"$TEST_DIR/td-a/data/f.txt"
echo y >"$TEST_DIR/td-b/data/f.txt"
create_tar "$TEST_DIR/td-old.tar" "$TEST_DIR/td-a"
create_tar "$TEST_DIR/td-new.tar" "$TEST_DIR/td-b"
expect_fail "tar-diff missing new archive" ./tar-diff "$TEST_DIR/td-old.tar.gz" "$TEST_DIR/missing-new.tar.gz" "$TEST_DIR/td-out.tardiff"

expect_fail "tar-diff corrupt archive inputs" ./tar-diff "$TEST_DIR/junk.gz" "$TEST_DIR/junk.gz" "$TEST_DIR/td-badio.tardiff"

# tar-patch: invalid / corrupt delta
printf 'not-a-tardiff' >"$TEST_DIR/bad-magic.tardiff"
expect_fail "tar-patch bad magic" ./tar-patch "$TEST_DIR/bad-magic.tardiff" "$TEST_DIR/orig" "$TEST_DIR/out1.tar"

printf '\x74\x61\x72\x64\x66\x31\x0a\x00' >"$TEST_DIR/bad-zstd.tardiff"
printf '\x00\x01\x02\x03' >>"$TEST_DIR/bad-zstd.tardiff"
expect_fail "tar-patch invalid zstd" ./tar-patch "$TEST_DIR/bad-zstd.tardiff" "$TEST_DIR/orig" "$TEST_DIR/out2.tar"

expect_fail "tar-patch missing base dir" ./tar-patch "$TEST_DIR/bad-magic.tardiff" "$TEST_DIR/does-not-exist" "$TEST_DIR/out3.tar"

# tar-patch: valid delta, missing source file
# Force bsdiff-sized payload + -max-bsdiff-size so the delta emits OPEN for data/only.txt;
# otherwise copyRest-only deltas can apply without that file and expect_fail would flake.
mkdir -p "$TEST_DIR/solo/data" "$TEST_DIR/solom/data"
head -c 4096 /dev/zero >"$TEST_DIR/solo/data/only.txt"
cp -a "$TEST_DIR/solo/data/only.txt" "$TEST_DIR/solom/data/only.txt"
printf 'patched' | dd of="$TEST_DIR/solom/data/only.txt" bs=1 seek=2000 conv=notrunc status=none 2>/dev/null || \
printf 'patched' | dd of="$TEST_DIR/solom/data/only.txt" bs=1 seek=2000 conv=notrunc 2>/dev/null
create_tar "$TEST_DIR/solo-old.tar" "$TEST_DIR/solo"
create_tar "$TEST_DIR/solo-new.tar" "$TEST_DIR/solom"
./tar-diff -max-bsdiff-size 64 "$TEST_DIR/solo-old.tar.gz" "$TEST_DIR/solo-new.tar.bz2" "$TEST_DIR/solo.tardiff"
rm -f "$TEST_DIR/solo/data/only.txt"
if [[ -e "$TEST_DIR/solo/data/only.txt" ]]; then
echo "expected solo/data/only.txt removed before tar-patch" >&2
exit 1
fi
expect_fail "tar-patch missing source member" ./tar-patch "$TEST_DIR/solo.tardiff" "$TEST_DIR/solo" "$TEST_DIR/solo-out.tar"

# tar-patch: stdout destination (happy path)
mkdir -p "$TEST_DIR/st/data" "$TEST_DIR/stm/data"
echo hello >"$TEST_DIR/st/data/only.txt"
echo 'hello!' >"$TEST_DIR/stm/data/only.txt"
create_tar "$TEST_DIR/st-old.tar" "$TEST_DIR/st"
create_tar "$TEST_DIR/st-new.tar" "$TEST_DIR/stm"
./tar-diff "$TEST_DIR/st-old.tar.gz" "$TEST_DIR/st-new.tar.bz2" "$TEST_DIR/st.tardiff"
./tar-patch "$TEST_DIR/st.tardiff" "$TEST_DIR/st" - | cmp - "$TEST_DIR/st-new.tar"

echo OK cli-errors-tar-diff-tar-patch
Loading
Loading