2016-05-23 12:23:55 +00:00
|
|
|
#!/usr/bin/env python3
|
2016-05-25 08:58:11 +00:00
|
|
|
"""Test zstd interoperability between versions"""
|
2016-08-30 17:04:33 +00:00
|
|
|
|
2017-08-31 19:20:50 +00:00
|
|
|
# ################################################################
|
2016-08-30 17:04:33 +00:00
|
|
|
# Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
|
|
|
|
# All rights reserved.
|
|
|
|
#
|
2017-08-31 19:20:50 +00:00
|
|
|
# This source code is licensed under both the BSD-style license (found in the
|
|
|
|
# LICENSE file in the root directory of this source tree) and the GPLv2 (found
|
|
|
|
# in the COPYING file in the root directory of this source tree).
|
|
|
|
# ################################################################
|
2016-05-23 12:23:55 +00:00
|
|
|
|
|
|
|
import filecmp
|
2016-05-25 08:12:39 +00:00
|
|
|
import glob
|
|
|
|
import hashlib
|
2016-05-23 12:23:55 +00:00
|
|
|
import os
|
|
|
|
import shutil
|
|
|
|
import sys
|
2016-06-17 11:39:43 +00:00
|
|
|
import subprocess
|
2016-06-17 12:43:24 +00:00
|
|
|
from subprocess import Popen, PIPE
|
2016-05-23 12:23:55 +00:00
|
|
|
|
2016-09-03 05:11:49 +00:00
|
|
|
repo_url = 'https://github.com/facebook/zstd.git'
|
2016-06-09 10:54:06 +00:00
|
|
|
tmp_dir_name = 'tests/versionsTest'
|
2016-05-23 12:23:55 +00:00
|
|
|
make_cmd = 'make'
|
|
|
|
git_cmd = 'git'
|
|
|
|
test_dat_src = 'README.md'
|
|
|
|
test_dat = 'test_dat'
|
|
|
|
head = 'vdevel'
|
2016-06-16 12:15:32 +00:00
|
|
|
dict_source = 'dict_source'
|
|
|
|
dict_files = './zstd/programs/*.c ./zstd/lib/common/*.c ./zstd/lib/compress/*.c ./zstd/lib/decompress/*.c ./zstd/lib/dictBuilder/*.c ./zstd/lib/legacy/*.c '
|
|
|
|
dict_files += './zstd/programs/*.h ./zstd/lib/common/*.h ./zstd/lib/compress/*.h ./zstd/lib/dictBuilder/*.h ./zstd/lib/legacy/*.h'
|
|
|
|
|
|
|
|
|
2016-06-17 12:07:42 +00:00
|
|
|
def execute(command, print_output=False, print_error=True, param_shell=False):
|
|
|
|
popen = Popen(command, stdout=PIPE, stderr=PIPE, shell=param_shell)
|
2016-06-17 11:39:43 +00:00
|
|
|
stdout_lines, stderr_lines = popen.communicate()
|
|
|
|
stderr_lines = stderr_lines.decode("utf-8")
|
|
|
|
stdout_lines = stdout_lines.decode("utf-8")
|
2016-06-16 12:15:32 +00:00
|
|
|
if print_output:
|
|
|
|
print(stdout_lines)
|
|
|
|
print(stderr_lines)
|
|
|
|
if popen.returncode is not None and popen.returncode != 0:
|
|
|
|
if not print_output and print_error:
|
|
|
|
print(stderr_lines)
|
2016-06-17 11:39:43 +00:00
|
|
|
return popen.returncode
|
2016-05-23 12:23:55 +00:00
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
def proc(cmd_args, pipe=True, dummy=False):
|
|
|
|
if dummy:
|
|
|
|
return
|
|
|
|
if pipe:
|
2016-06-17 11:39:43 +00:00
|
|
|
subproc = Popen(cmd_args, stdout=PIPE, stderr=PIPE)
|
2016-05-23 12:23:55 +00:00
|
|
|
else:
|
2016-06-17 11:39:43 +00:00
|
|
|
subproc = Popen(cmd_args)
|
2016-05-23 12:23:55 +00:00
|
|
|
return subproc.communicate()
|
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
def make(args, pipe=True):
|
|
|
|
return proc([make_cmd] + args, pipe)
|
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
def git(args, pipe=True):
|
|
|
|
return proc([git_cmd] + args, pipe)
|
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
def get_git_tags():
|
|
|
|
stdout, stderr = git(['tag', '-l', 'v[0-9].[0-9].[0-9]'])
|
|
|
|
tags = stdout.decode('utf-8').split()
|
|
|
|
return tags
|
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-06-16 17:29:09 +00:00
|
|
|
def create_dict(tag, dict_source_path):
|
|
|
|
dict_name = 'dict.' + tag
|
|
|
|
if not os.path.isfile(dict_name):
|
|
|
|
cFiles = glob.glob(dict_source_path + "/*.c")
|
|
|
|
hFiles = glob.glob(dict_source_path + "/*.h")
|
2016-06-17 12:43:24 +00:00
|
|
|
if tag == 'v0.5.0':
|
|
|
|
result = execute('./dictBuilder.' + tag + ' ' + ' '.join(cFiles) + ' ' + ' '.join(hFiles) + ' -o ' + dict_name, print_output=False, param_shell=True)
|
|
|
|
else:
|
2016-06-17 13:17:35 +00:00
|
|
|
result = execute('./zstd.' + tag + ' -f --train ' + ' '.join(cFiles) + ' ' + ' '.join(hFiles) + ' -o ' + dict_name, print_output=False, param_shell=True)
|
2016-06-17 12:43:24 +00:00
|
|
|
if result == 0:
|
2016-06-17 11:39:43 +00:00
|
|
|
print(dict_name + ' created')
|
|
|
|
else:
|
|
|
|
print('ERROR: creating of ' + dict_name + ' failed')
|
2016-06-16 17:29:09 +00:00
|
|
|
else:
|
|
|
|
print(dict_name + ' already exists')
|
|
|
|
|
|
|
|
|
|
|
|
def dict_compress_sample(tag, sample):
|
|
|
|
dict_name = 'dict.' + tag
|
2016-06-17 12:43:24 +00:00
|
|
|
DEVNULL = open(os.devnull, 'wb')
|
|
|
|
if subprocess.call(['./zstd.' + tag, '-D', dict_name, '-f', sample], stderr=DEVNULL) == 0:
|
|
|
|
os.rename(sample + '.zst', sample + '_01_64_' + tag + '_dictio.zst')
|
|
|
|
if subprocess.call(['./zstd.' + tag, '-D', dict_name, '-5f', sample], stderr=DEVNULL) == 0:
|
|
|
|
os.rename(sample + '.zst', sample + '_05_64_' + tag + '_dictio.zst')
|
|
|
|
if subprocess.call(['./zstd.' + tag, '-D', dict_name, '-9f', sample], stderr=DEVNULL) == 0:
|
|
|
|
os.rename(sample + '.zst', sample + '_09_64_' + tag + '_dictio.zst')
|
|
|
|
if subprocess.call(['./zstd.' + tag, '-D', dict_name, '-15f', sample], stderr=DEVNULL) == 0:
|
|
|
|
os.rename(sample + '.zst', sample + '_15_64_' + tag + '_dictio.zst')
|
|
|
|
if subprocess.call(['./zstd.' + tag, '-D', dict_name, '-18f', sample], stderr=DEVNULL) == 0:
|
|
|
|
os.rename(sample + '.zst', sample + '_18_64_' + tag + '_dictio.zst')
|
2016-06-16 17:29:09 +00:00
|
|
|
# zstdFiles = glob.glob("*.zst*")
|
|
|
|
# print(zstdFiles)
|
|
|
|
print(tag + " : dict compression completed")
|
|
|
|
|
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
def compress_sample(tag, sample):
|
2016-06-17 12:43:24 +00:00
|
|
|
DEVNULL = open(os.devnull, 'wb')
|
2016-05-25 08:12:39 +00:00
|
|
|
if subprocess.call(['./zstd.' + tag, '-f', sample], stderr=DEVNULL) == 0:
|
2016-06-16 18:23:11 +00:00
|
|
|
os.rename(sample + '.zst', sample + '_01_64_' + tag + '_nodict.zst')
|
2016-05-25 08:12:39 +00:00
|
|
|
if subprocess.call(['./zstd.' + tag, '-5f', sample], stderr=DEVNULL) == 0:
|
2016-06-16 18:23:11 +00:00
|
|
|
os.rename(sample + '.zst', sample + '_05_64_' + tag + '_nodict.zst')
|
2016-05-25 08:12:39 +00:00
|
|
|
if subprocess.call(['./zstd.' + tag, '-9f', sample], stderr=DEVNULL) == 0:
|
2016-06-16 18:23:11 +00:00
|
|
|
os.rename(sample + '.zst', sample + '_09_64_' + tag + '_nodict.zst')
|
2016-05-25 08:12:39 +00:00
|
|
|
if subprocess.call(['./zstd.' + tag, '-15f', sample], stderr=DEVNULL) == 0:
|
2016-06-16 18:23:11 +00:00
|
|
|
os.rename(sample + '.zst', sample + '_15_64_' + tag + '_nodict.zst')
|
2016-05-25 08:12:39 +00:00
|
|
|
if subprocess.call(['./zstd.' + tag, '-18f', sample], stderr=DEVNULL) == 0:
|
2016-06-16 18:23:11 +00:00
|
|
|
os.rename(sample + '.zst', sample + '_18_64_' + tag + '_nodict.zst')
|
2016-05-23 12:23:55 +00:00
|
|
|
# zstdFiles = glob.glob("*.zst*")
|
|
|
|
# print(zstdFiles)
|
2016-06-03 13:41:51 +00:00
|
|
|
print(tag + " : compression completed")
|
2016-05-23 12:23:55 +00:00
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
# http://stackoverflow.com/a/19711609/2132223
|
|
|
|
def sha1_of_file(filepath):
|
|
|
|
with open(filepath, 'rb') as f:
|
|
|
|
return hashlib.sha1(f.read()).hexdigest()
|
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
def remove_duplicates():
|
2016-06-16 18:23:11 +00:00
|
|
|
list_of_zst = sorted(glob.glob('*.zst'))
|
2016-05-23 12:23:55 +00:00
|
|
|
for i, ref_zst in enumerate(list_of_zst):
|
|
|
|
if not os.path.isfile(ref_zst):
|
|
|
|
continue
|
2016-05-25 08:58:11 +00:00
|
|
|
for j in range(i + 1, len(list_of_zst)):
|
2016-05-23 12:23:55 +00:00
|
|
|
compared_zst = list_of_zst[j]
|
|
|
|
if not os.path.isfile(compared_zst):
|
|
|
|
continue
|
|
|
|
if filecmp.cmp(ref_zst, compared_zst):
|
|
|
|
os.remove(compared_zst)
|
|
|
|
print('duplicated : {} == {}'.format(ref_zst, compared_zst))
|
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
|
2016-06-17 12:43:24 +00:00
|
|
|
def decompress_zst(tag):
|
2016-05-23 12:23:55 +00:00
|
|
|
dec_error = 0
|
2016-06-16 18:23:11 +00:00
|
|
|
list_zst = sorted(glob.glob('*_nodict.zst'))
|
2016-05-23 12:23:55 +00:00
|
|
|
for file_zst in list_zst:
|
2016-05-25 08:58:11 +00:00
|
|
|
print(file_zst, end=' ')
|
|
|
|
print(tag, end=' ')
|
2016-05-23 12:23:55 +00:00
|
|
|
file_dec = file_zst + '_d64_' + tag + '.dec'
|
2016-06-17 12:43:24 +00:00
|
|
|
if tag <= 'v0.5.0':
|
2016-06-16 12:15:32 +00:00
|
|
|
params = ['./zstd.' + tag, '-df', file_zst, file_dec]
|
|
|
|
else:
|
|
|
|
params = ['./zstd.' + tag, '-df', file_zst, '-o', file_dec]
|
2016-06-17 12:07:42 +00:00
|
|
|
if execute(params) == 0:
|
2016-05-23 12:23:55 +00:00
|
|
|
if not filecmp.cmp(file_dec, test_dat):
|
|
|
|
print('ERR !! ')
|
|
|
|
dec_error = 1
|
|
|
|
else:
|
|
|
|
print('OK ')
|
2016-05-23 13:04:14 +00:00
|
|
|
else:
|
|
|
|
print('command does not work')
|
2016-08-28 23:25:56 +00:00
|
|
|
dec_error = 1
|
2016-05-23 12:23:55 +00:00
|
|
|
return dec_error
|
|
|
|
|
2016-05-23 13:43:17 +00:00
|
|
|
|
2016-06-17 12:43:24 +00:00
|
|
|
def decompress_dict(tag):
|
2016-06-16 17:29:09 +00:00
|
|
|
dec_error = 0
|
2016-06-16 18:23:11 +00:00
|
|
|
list_zst = sorted(glob.glob('*_dictio.zst'))
|
2016-06-16 17:29:09 +00:00
|
|
|
for file_zst in list_zst:
|
2016-06-16 18:23:11 +00:00
|
|
|
dict_tag = file_zst[0:len(file_zst)-11] # remove "_dictio.zst"
|
|
|
|
if head in dict_tag: # find vdevel
|
2016-06-16 17:29:09 +00:00
|
|
|
dict_tag = head
|
2016-06-16 18:23:11 +00:00
|
|
|
else:
|
|
|
|
dict_tag = dict_tag[dict_tag.rfind('v'):]
|
2016-06-17 12:43:24 +00:00
|
|
|
if tag == 'v0.6.0' and dict_tag < 'v0.6.0':
|
|
|
|
continue
|
2016-06-16 17:29:09 +00:00
|
|
|
dict_name = 'dict.' + dict_tag
|
2016-06-16 18:23:11 +00:00
|
|
|
print(file_zst + ' ' + tag + ' dict=' + dict_tag, end=' ')
|
2016-06-16 17:29:09 +00:00
|
|
|
file_dec = file_zst + '_d64_' + tag + '.dec'
|
2016-06-17 12:43:24 +00:00
|
|
|
if tag <= 'v0.5.0':
|
2016-06-16 17:29:09 +00:00
|
|
|
params = ['./zstd.' + tag, '-D', dict_name, '-df', file_zst, file_dec]
|
|
|
|
else:
|
|
|
|
params = ['./zstd.' + tag, '-D', dict_name, '-df', file_zst, '-o', file_dec]
|
2016-06-17 12:07:42 +00:00
|
|
|
if execute(params) == 0:
|
2016-06-16 17:29:09 +00:00
|
|
|
if not filecmp.cmp(file_dec, test_dat):
|
|
|
|
print('ERR !! ')
|
|
|
|
dec_error = 1
|
|
|
|
else:
|
|
|
|
print('OK ')
|
|
|
|
else:
|
|
|
|
print('command does not work')
|
2016-06-16 18:23:11 +00:00
|
|
|
dec_error = 1
|
2016-06-16 17:29:09 +00:00
|
|
|
return dec_error
|
2016-06-16 12:15:32 +00:00
|
|
|
|
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
if __name__ == '__main__':
|
|
|
|
error_code = 0
|
2016-06-16 12:15:32 +00:00
|
|
|
base_dir = os.getcwd() + '/..' # /path/to/zstd
|
|
|
|
tmp_dir = base_dir + '/' + tmp_dir_name # /path/to/zstd/tests/versionsTest
|
|
|
|
clone_dir = tmp_dir + '/' + 'zstd' # /path/to/zstd/tests/versionsTest/zstd
|
|
|
|
dict_source_path = tmp_dir + '/' + dict_source # /path/to/zstd/tests/versionsTest/dict_source
|
|
|
|
programs_dir = base_dir + '/programs' # /path/to/zstd/programs
|
2016-05-23 12:23:55 +00:00
|
|
|
os.makedirs(tmp_dir, exist_ok=True)
|
|
|
|
|
|
|
|
# since Travis clones limited depth, we should clone full repository
|
|
|
|
if not os.path.isdir(clone_dir):
|
|
|
|
git(['clone', repo_url, clone_dir])
|
|
|
|
|
|
|
|
shutil.copy2(base_dir + '/' + test_dat_src, tmp_dir + '/' + test_dat)
|
|
|
|
|
|
|
|
# Retrieve all release tags
|
|
|
|
print('Retrieve all release tags :')
|
|
|
|
os.chdir(clone_dir)
|
2016-08-28 23:56:17 +00:00
|
|
|
alltags = get_git_tags() + [head]
|
|
|
|
tags = [t for t in alltags if t >= 'v0.4.0']
|
2016-05-25 08:12:39 +00:00
|
|
|
print(tags)
|
2016-05-23 12:23:55 +00:00
|
|
|
|
|
|
|
# Build all release zstd
|
|
|
|
for tag in tags:
|
|
|
|
os.chdir(base_dir)
|
2016-06-16 09:32:57 +00:00
|
|
|
dst_zstd = '{}/zstd.{}'.format(tmp_dir, tag) # /path/to/zstd/tests/versionsTest/zstd.<TAG>
|
2016-05-23 12:23:55 +00:00
|
|
|
if not os.path.isfile(dst_zstd) or tag == head:
|
|
|
|
if tag != head:
|
2016-06-09 10:54:06 +00:00
|
|
|
r_dir = '{}/{}'.format(tmp_dir, tag) # /path/to/zstd/tests/versionsTest/<TAG>
|
2016-05-23 12:23:55 +00:00
|
|
|
os.makedirs(r_dir, exist_ok=True)
|
|
|
|
os.chdir(clone_dir)
|
|
|
|
git(['--work-tree=' + r_dir, 'checkout', tag, '--', '.'], False)
|
2016-06-17 12:43:24 +00:00
|
|
|
if tag == 'v0.5.0':
|
|
|
|
os.chdir(r_dir + '/dictBuilder') # /path/to/zstd/tests/versionsTest/v0.5.0/dictBuilder
|
|
|
|
make(['clean', 'dictBuilder'], False)
|
|
|
|
shutil.copy2('dictBuilder', '{}/dictBuilder.{}'.format(tmp_dir, tag))
|
2016-06-09 10:54:06 +00:00
|
|
|
os.chdir(r_dir + '/programs') # /path/to/zstd/tests/versionsTest/<TAG>/programs
|
2016-05-23 12:23:55 +00:00
|
|
|
make(['clean', 'zstd'], False)
|
|
|
|
else:
|
|
|
|
os.chdir(programs_dir)
|
|
|
|
make(['zstd'], False)
|
|
|
|
shutil.copy2('zstd', dst_zstd)
|
|
|
|
|
|
|
|
# remove any remaining *.zst and *.dec from previous test
|
|
|
|
os.chdir(tmp_dir)
|
|
|
|
for compressed in glob.glob("*.zst"):
|
|
|
|
os.remove(compressed)
|
2016-05-25 08:12:39 +00:00
|
|
|
for dec in glob.glob("*.dec"):
|
2016-05-23 12:23:55 +00:00
|
|
|
os.remove(dec)
|
|
|
|
|
2016-06-16 12:15:32 +00:00
|
|
|
# copy *.c and *.h to a temporary directory ("dict_source")
|
|
|
|
if not os.path.isdir(dict_source_path):
|
|
|
|
os.mkdir(dict_source_path)
|
|
|
|
print('cp ' + dict_files + ' ' + dict_source_path)
|
2016-06-17 12:07:42 +00:00
|
|
|
execute('cp ' + dict_files + ' ' + dict_source_path, param_shell=True)
|
2016-06-16 12:15:32 +00:00
|
|
|
|
2016-05-23 12:23:55 +00:00
|
|
|
print('Compress test.dat by all released zstd')
|
|
|
|
|
2016-05-25 08:12:39 +00:00
|
|
|
error_code = 0
|
2016-05-23 12:23:55 +00:00
|
|
|
for tag in tags:
|
|
|
|
print(tag)
|
2016-06-17 12:43:24 +00:00
|
|
|
if tag >= 'v0.5.0':
|
2016-06-16 12:15:32 +00:00
|
|
|
create_dict(tag, dict_source_path)
|
2016-06-16 17:29:09 +00:00
|
|
|
dict_compress_sample(tag, test_dat)
|
2016-06-16 18:23:11 +00:00
|
|
|
remove_duplicates()
|
|
|
|
error_code += decompress_dict(tag)
|
2016-05-23 12:23:55 +00:00
|
|
|
compress_sample(tag, test_dat)
|
|
|
|
remove_duplicates()
|
2016-06-17 12:43:24 +00:00
|
|
|
error_code += decompress_zst(tag)
|
2016-05-23 12:23:55 +00:00
|
|
|
|
|
|
|
print('')
|
|
|
|
print('Enumerate different compressed files')
|
|
|
|
zstds = sorted(glob.glob('*.zst'))
|
|
|
|
for zstd in zstds:
|
|
|
|
print(zstd + ' : ' + repr(os.path.getsize(zstd)) + ', ' + sha1_of_file(zstd))
|
|
|
|
|
|
|
|
if error_code != 0:
|
2016-08-28 23:25:56 +00:00
|
|
|
print('====== ERROR !!! =======')
|
2016-05-23 12:23:55 +00:00
|
|
|
|
|
|
|
sys.exit(error_code)
|