luci-py/isolateserver.py: Add support for tar archives when downloading.

client/isolateserver.py

 #!/usr/bin/env python
 # Copyright 2013 The LUCI Authors. All rights reserved.
 # Use of this source code is governed under the Apache License, Version 2.0
 # that can be found in the LICENSE file.
 
 """Archives a set of files or directories to an Isolate Server."""
 
-__version__ = '0.6.0'
+__version__ = '0.7.0'
 
 import base64
 import errno
 import functools
 import io
 import logging
 import optparse
 import os
 import re
 import signal
 import stat
 import sys
+import tarfile
 import tempfile
 import threading
 import time
 import types
 import zlib
 
 from third_party import colorama
 from third_party.depot_tools import fix_encoding
 from third_party.depot_tools import subcommand
 
@@ skipping 131 matching lines @@
   if name is None:
     return
 
   # If the file like object was created using something like open("test.txt")
   # name will end up being a str (such as a function outside our control, like
   # the standard library). We want all our paths to be unicode objects, so we
   # decode it.
   if not isinstance(name, unicode):
     name = name.decode(sys.getfilesystemencoding())
 
+  # fs.exists requires an absolute path, otherwise it will fail with an
+  # assertion error.
+  if not os.path.isabs(name):
+      return
+
   if fs.exists(name):
     return name
 
 
 # TODO(tansell): Replace fileobj_copy with shutil.copyfileobj once proper file
 # wrappers have been created.
 def fileobj_copy(
     dstfileobj, srcfileobj, size=-1,
     chunk_size=isolated_format.DISK_FILE_CHUNK):
   """Copy data from srcfileobj to dstfileobj.
@@ skipping 1918 matching lines @@
 
               if filetype == 'basic':
                 file_mode = props.get('m')
                 if file_mode:
                   # Ignore all bits apart from the user
                   file_mode &= 0700
                 putfile(
                     srcfileobj, fullpath, file_mode,
                     use_symlink=use_symlinks)
 
+              elif filetype == 'tar':
+                basedir = os.path.dirname(fullpath)
+                with tarfile.TarFile(fileobj=srcfileobj) as extractor:
+                  for ti in extractor:
+                    if not ti.isfile():
+                      logging.warning(
+                          'Path(%r) is nonfile (%s), skipped',
+                          ti.name, ti.type)
+                      continue
+                    fp = os.path.normpath(os.path.join(basedir, ti.name))
+                    if not fp.startswith(basedir):
+                      logging.error(
+                          'Path(%r) is outside root directory',
+                          fp)
+                    ifd = extractor.extractfile(ti)
+                    file_path.ensure_tree(os.path.dirname(fp))
+                    putfile(ifd, fp, 0700, ti.size)
+
               elif filetype == 'ar':
                 basedir = os.path.dirname(fullpath)
                 extractor = arfile.ArFileReader(srcfileobj, fullparse=False)
                 for ai, ifd in extractor:
                   fp = os.path.normpath(os.path.join(basedir, ai.name))
+                  if not fp.startswith(basedir):
+                    logging.error(
+                        'Path(%r) is outside root directory',
+                        fp)
                   file_path.ensure_tree(os.path.dirname(fp))
                   putfile(ifd, fp, 0700, ai.size)
 
               else:
                 raise isolated_format.IsolatedError(
                       'Unknown file type %r', filetype)
 
           # Report progress.
           duration = time.time() - last_update
           if duration > DELAY_BETWEEN_UPDATES_IN_SECS:
@@ skipping 331 matching lines @@
   return dispatcher.execute(OptionParserIsolateServer(), args)
 
 
 if __name__ == '__main__':
   subprocess42.inhibit_os_error_reporting()
   fix_encoding.fix_encoding()
   tools.disable_buffering()
   colorama.init()
   file_path.enable_symlink()
   sys.exit(main(sys.argv[1:]))