From 65d53459ee1837d1c8c321a3270e276ef568c846 Mon Sep 17 00:00:00 2001
From: dennisvang <29799340+dennisvang@users.noreply.github.com>
Date: Tue, 21 Nov 2023 18:03:20 +0100
Subject: [PATCH] replace shutil.make_archive by tarfile and gzip

this provides greater control and allows us to create reproducible gzip files
i.e. without filename in the header, and with mtime set to zero
---
 src/tufup/repo/__init__.py | 28 +++++++++++++---------------
 tests/test_repo.py         | 12 ++++++++++--
 2 files changed, 23 insertions(+), 17 deletions(-)

diff --git a/src/tufup/repo/__init__.py b/src/tufup/repo/__init__.py
index 30b8890..f929b4b 100644
--- a/src/tufup/repo/__init__.py
+++ b/src/tufup/repo/__init__.py
@@ -1,9 +1,11 @@
+import tempfile
 from copy import deepcopy
 from datetime import datetime, timedelta
 import inspect
 import json
 import logging
 import pathlib
+import tarfile
 
 try:
     # workaround for PyInstaller issue 6911 (setuptools issue 3089)
@@ -37,7 +39,7 @@
 )
 from tuf.api.serialization.json import JSONSerializer
 
-from tufup.common import Patcher, SUFFIX_ARCHIVE, SUFFIX_PATCH, TargetMeta
+from tufup.common import Patcher, SUFFIX_PATCH, TargetMeta
 from tufup.utils.platform_specific import _patched_resolve
 
 logger = logging.getLogger(__name__)
@@ -79,12 +81,7 @@ def make_gztar_archive(
     dst_dir: Union[pathlib.Path, str],
     app_name: str,
     version: str,
-    **kwargs,  # allowed kwargs are passed on to shutil.make_archive
 ) -> Optional[TargetMeta]:
-    # remove disallowed kwargs
-    for key in ['base_name', 'root_dir', 'format']:
-        if kwargs.pop(key, None):
-            logger.warning(f'{key} ignored: using default')
     # ensure paths
     src_dir = pathlib.Path(src_dir)
     dst_dir = pathlib.Path(dst_dir)
@@ -97,15 +94,16 @@ def make_gztar_archive(
         if input(f'Found existing archive: {archive_path}.\nOverwrite? [n]/y') != 'y':
             print('Using existing archive.')
             return TargetMeta(archive_path)
-    # make archive
-    base_name = str(dst_dir / archive_filename.replace(SUFFIX_ARCHIVE, ''))
-    archive_path_str = shutil.make_archive(
-        base_name=base_name,  # archive file path, no suffix
-        root_dir=str(src_dir),  # paths in archive will be relative to root_dir
-        format='gztar',
-        **kwargs,
-    )
-    return TargetMeta(target_path=archive_path_str)
+    # create archive
+    with tempfile.NamedTemporaryFile(mode='wb') as temp_file:
+        # make temporary tar archive
+        with tarfile.open(fileobj=temp_file, mode='w') as tar:
+            for path in src_dir.iterdir():
+                tar.add(name=path, arcname=path.relative_to(src_dir), recursive=True)
+        temp_file_path = pathlib.Path(temp_file.name)
+        # compress tar archive using gzip (force mtime to zero for reproducibility)
+        Patcher.gzip(src_path=temp_file_path, dst_path=archive_path, mtime=0)
+    return TargetMeta(target_path=archive_path)
 
 
 class RolesDict(TypedDict):
diff --git a/tests/test_repo.py b/tests/test_repo.py
index 4ed6170..48978c3 100644
--- a/tests/test_repo.py
+++ b/tests/test_repo.py
@@ -3,6 +3,8 @@
 import json
 import logging
 import pathlib
+import struct
+import tarfile
 from tempfile import TemporaryDirectory
 from time import sleep
 from unittest.mock import Mock, patch
@@ -96,14 +98,20 @@ def test_make_gztar_archive(self):
                     dst_dir=self.temp_dir_path,
                     app_name=app_name,
                     version=version,
-                    base_dir='.',  # this kwarg is allowed
-                    root_dir='some path',  # this kwarg is removed
                 )
             self.assertIsInstance(archive, TargetMeta)
             self.assertEqual(exists, mock_input_no.called)
             self.assertTrue(archive.path.exists())
             self.assertTrue(app_name in str(archive.path))
             self.assertTrue(version in str(archive.path))
+        # check mtime in archive gzip header (see RFC 1952 and test_common.py)
+        mtime = struct.unpack('<I', archive.path.read_bytes()[4:8])[0]
+        self.assertEqual(0, mtime)
+        # check archive content
+        with tarfile.open(archive.path) as tar:
+            self.assertEqual(
+                {'root.txt', 'sub', 'sub/sub.txt'}, set(item.name for item in tar)
+            )
 
 
 class BaseTests(TempDirTestCase):