GSUtil.py wrapper script

gsutil.py

Index: gsutil.py
diff --git a/gsutil.py b/gsutil.py
new file mode 100755
index 0000000000000000000000000000000000000000..9fdef76d323f37ee1b260ee35794662c62e8f490
--- /dev/null
+++ b/gsutil.py
@@ -0,0 +1,142 @@
+#!/usr/bin/env python
+# Copyright (c) 2014 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+"""Run a pinned gsutil."""
+
+
+import argparse
+import shutil
+import zipfile
+import hashlib
+import base64
+import os
+import sys
+import json
+import urllib
+import subprocess
+
+
+GSUTIL_URL = 'https://storage.googleapis.com/pub/'
+API_URL = 'https://www.googleapis.com/storage/v1/b/pub/o/'
+
+THIS_DIR = os.path.dirname(os.path.abspath(__file__))
+DEFAULT_BIN_DIR = os.path.join(THIS_DIR, 'bin', 'gsutil')

[dnj, 2014/11/20 23:03:02]

My only concern with 'bin' is that it's part of the LSB, and other people might
start dumping things there. Either dump a 'README' in it or name it something
like 'external_bin' or 'bin.fetch' or something to indicate its transient
nature.

[Ryan Tseng, 2014/11/21 02:14:55]

Good point, I'll use external_bin

+
+
+class SubprocessError(Exception):
+  pass
+class InvalidGsutilError(Exception):
+  pass
+
+
+def call(args, verbose=True, **kwargs):
+  kwargs['stdout'] = subprocess.PIPE
+  kwargs['stderr'] = subprocess.STDOUT
+  proc = subprocess.Popen(args, **kwargs)
+  out = []
+  for line in proc.stdout:
+    out.append(line)
+    if verbose:
+      sys.stdout.write(line)
+  code = proc.wait()
+  if code:
+    raise SubprocessError('%s failed with %s' % (args, code))
+  return ''.join(out)
+
+
+def download_gsutil(version, target_dir):
+  """Downloads gsutil into the target_dir."""
+  filename = 'gsutil_%s.zip' % version
+  target_filename = os.path.join(target_dir, filename)
+
+  # Check if the target exists already.
+  if os.path.exists(target_filename):
+    md5_calc = hashlib.md5()
+    with open(target_filename, 'rb') as f:
+      while True:
+        buf = f.read(4096)
+        if not buf:
+          break
+        md5_calc.update(buf)
+    local_md5 = md5_calc.hexdigest()
+
+    metadata_url = '%s%s' % (API_URL, filename)
+    metadata = json.load(urllib.urlopen(metadata_url))
+    remote_md5 = base64.b64decode(metadata['md5Hash'])
+
+    if local_md5 == remote_md5:
+      return target_filename
+    os.remove(target_filename)
+
+  # Do the download.
+  url = '%s%s' % (GSUTIL_URL, filename)
+  u = urllib.urlopen(url)
+  with open(target_filename, 'wb') as f:
+    while True:
+      buf = u.read(4096)
+      if not buf:
+        break
+      f.write(buf)
+  return target_filename
+
+
+def check_gsutil(gsutil_bin):
+  """Run gsutil version and make sure it runs."""
+  try:
+    call([gsutil_bin, 'version'], verbose=False)
+    return True
+  except SubprocessError:
+    return False
+
+
+def ensure_gsutil(version, target):
+  bin_dir = os.path.join(target, 'gsutil_%s' % version)
+  gsutil_bin = os.path.join(bin_dir, 'gsutil', 'gsutil')
+  if os.path.isfile(gsutil_bin) and check_gsutil(gsutil_bin):
+    # Everything is awesome! we're all done here.
+    return gsutil_bin
+
+  if os.path.isdir(bin_dir):
+    # Clean up if we're redownloading a corrupted gsutil.
+    shutil.rmtree(bin_dir)
+  cache_dir = os.path.join(target, '.cache_dir')
+  if not os.path.isdir(cache_dir):
+    os.makedirs(cache_dir)
+  target_zip_filename = download_gsutil(version, cache_dir)
+  with zipfile.ZipFile(target_zip_filename, 'r') as target_zip:
+    target_zip.extractall(bin_dir)
+  os.chmod(gsutil_bin, 0755)  # python zipfile doesn't set exec bit.

[dnj, 2014/11/20 23:03:02]

Windows will ignore this. Does anything special need to be done, or are all
things executable on Windows? I forget~

[Ryan Tseng, 2014/11/21 02:14:55]

Actually if i used [sys.executable, <path to gsutil>, ...] then this is
unnecessary :D

+
+  # Final check that the gsutil bin is okay.  This should never fail.
+  if not check_gsutil(gsutil_bin):
+    raise InvalidGsutilError()
+
+  return gsutil_bin
+
+
+def run_gsutil(_force_version, fallback, _target, args):
+  # TODO(hinoka): Actually use this
+  # gsutil_bin = ensure_gsutil(force_version, target)
+  cmd = [sys.executable, fallback] + args
+  call(cmd)
+
+
+def parse_args():
+  parser = argparse.ArgumentParser()
+  parser.add_argument('--force_version', default='4.6')
+  parser.add_argument('--fallback')
+  parser.add_argument('--target', default=DEFAULT_BIN_DIR)
+  parser.add_argument('args', nargs='*')

[dnj, 2014/11/20 23:03:02]

Replace '*' with 'argparse.REMAINDER' to make argparse pick up flags and stuff.

[Ryan Tseng, 2014/11/21 02:14:55]

Good catch, done.

+
+  args = parser.parse_args()
+  return args.force_version, args.fallback, args.target, args.args
+
+
+def main():
+  force_version, fallback, target, args = parse_args()
+  run_gsutil(force_version, fallback, target, args)
+
+if __name__ == '__main__':
+  sys.exit(main())