Extract original filename for GZIP (.gz) archives.

doc/changelog.txt

@@ -19,7 +19,9 @@
   module in Python >= 3.12 to detect unsafe archive entries.
 * Added new command "patool version" to print version information.
 * Do not use maximum compression anymore, the default settings should
-  be good enough for a standard use case.
+  be good enough for our standard use case.
+* Restore original file name for GZIP (.gz) archives.
+  Closes: GH bug #67
 
 
 1.15.0 (released 28.10.2023)

patoolib/programs/gzip.py

@@ -14,12 +14,42 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 """Archive commands for the gzip program."""
-from . import extract_singlefile_standard, test_singlefile_standard
-from .. import util
+from . import test_singlefile_standard
+from .. import util, fileutil, log
+import os
 
-extract_gzip = extract_compress = extract_singlefile_standard
 test_gzip = test_compress = test_singlefile_standard
 
+def extract_gzip(archive, compression, cmd, verbosity, interactive, outdir):
+    cmdlist = [util.shell_quote(cmd)]
+    if verbosity > 1:
+        cmdlist.append('-v')
+    outfile = get_original_filename(cmd, outdir, archive)
+    cmdlist.extend(['-c', '-d', '--', util.shell_quote(archive), '>',
+        util.shell_quote(outfile)])
+    return (cmdlist, {'shell': True})
+
+extract_compress = extract_gzip
+
+
+def get_original_filename(cmd, outdir, archive):
+    outfile = fileutil.get_single_outfile(outdir, archive)
+    cmdlist = [cmd, "--name", "--list", archive]
+    try:
+        output = util.backtick(cmdlist).strip()
+        lines = output.splitlines()
+        values = lines[1].split()
+        baseoutfile = os.path.basename(outfile)
+        basefilename = os.path.basename(values[-1])
+        if os.path.normcase(baseoutfile) != os.path.normcase(basefilename):
+            basearchive = os.path.basename(archive)
+            # avoid overwriting the original archive with the same name
+            if os.path.normcase(basefilename) != os.path.normcase(basearchive):
+                outfile = os.path.join(outdir, basefilename)
+    except OSError as err:
+        log.log_error(f"could not run {cmd}: {err}")
+    return outfile
+
 
 def create_gzip(archive, compression, cmd, verbosity, interactive, filenames):
     """Create a GZIP archive."""

tests/archives/__init__.py

@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-# Copyright (C) 2010-2015 Bastian Kleineidam
+# Copyright (C) 2010-2023 Bastian Kleineidam
 #
 # 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
@@ -72,9 +72,9 @@ def archive_extract(self, filename, check=Content.Recursive):
         relarchive = os.path.join("..", archive[len(basedir)+1:])
         self._archive_extract(relarchive, check, verbosity=1)
 
-    def _archive_extract (self, archive, check, verbosity=0):
+    def _archive_extract(self, archive, check, verbosity=0):
         # create a temporary directory for extraction
-        tmpdir = patoolib.fileutil.tmpdir(dir=basedir)
+        tmpdir = patoolib.fileutil.tmpdir(dir=basedir, prefix="test_")
         try:
             olddir = patoolib.fileutil.chdir(tmpdir)
             try:
@@ -87,7 +87,7 @@ def _archive_extract (self, archive, check, verbosity=0):
         finally:
             patoolib.fileutil.rmtree(tmpdir)
 
-    def check_extracted_archive (self, archive, output, check):
+    def check_extracted_archive(self, archive, output, check):
         if check == Content.Recursive:
             # outdir is the 't' directory of the archive
             self.assertEqual(output, 't')
@@ -96,15 +96,20 @@ def check_extracted_archive (self, archive, output, check):
             self.check_textfile(txtfile, 't.txt')
         elif check == Content.Singlefile:
             # a non-existing directory to ensure files do not exist in it
-            ned = get_nonexisting_directory(os.getcwd())
-            expected_output = os.path.basename(patoolib.fileutil.get_single_outfile(ned, archive))
+            expected_output = self.get_expected_singlefile_output(archive)
             self.check_textfile(output, expected_output)
         elif check == Content.Multifile:
             txtfile = os.path.join(output, 't.txt')
             self.check_textfile(txtfile, 't.txt')
             txtfile2 = os.path.join(output, 't2.txt')
             self.check_textfile(txtfile2, 't2.txt')
 
+    def get_expected_singlefile_output(self, archive):
+        """Return the expected output file name of a singlefile archive."""
+        if archive.endswith(".foo"):
+            return patoolib.fileutil.stripext(os.path.basename(archive))
+        return "t.txt"
+
     def check_directory (self, dirname, expectedname):
         """Check that directory exists."""
         self.assertTrue(os.path.isdir(dirname), dirname)

tests/archives/test_gzip.py

@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-# Copyright (C) 2010-2015 Bastian Kleineidam
+# Copyright (C) 2010-2023 Bastian Kleineidam
 #
 # 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
@@ -28,6 +28,11 @@ def test_gzip (self):
     @needs_program('file')
     @needs_program(program)
     def test_gzip_file(self):
-        self.archive_commands('t.txt.gz.foo', skip_create=True, check=None)
+        self.archive_commands('t.txt.gz.foo', skip_create=True, check=Content.Singlefile)
         self.archive_extract('t.txt.Z.foo', check=Content.Singlefile)
 
+    def get_expected_singlefile_output(self, archive):
+        """Gzip restores the original filename for .gz files"""
+        if archive.endswith(".Z.foo"):
+            return "t.txt.Z"
+        return "t.txt"

tests/archives/test_pigz.py

@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-# Copyright (C) 2010-2015 Bastian Kleineidam
+# Copyright (C) 2010-2023 Bastian Kleineidam
 #
 # 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
@@ -29,3 +29,7 @@ def test_pigz (self):
     def test_pigz_file (self):
         self.archive_commands('t.txt.gz.foo', check=Content.Singlefile,
           skip_create=True, skip_test=True)
+
+    def get_expected_singlefile_output(self, archive):
+        """pigz restores the original filename for .gz files"""
+        return "t.txt"