Scripts to download files from google storage based on sha1 sums

continuation of: https://codereview.chromium.org/11664024
Moved it from chrome/trunk/src/build to depot_tools/

BUG=153360
TEST=two unittests included in tests/
For end-to-end testing, check out a large directory.  Run
find . -name .svn -prune -o -size +1000k -type f -print0 | upload_to_google_storage.py -b chrome-artifacts -0 -
(replacing chrome-artifacts with an upload-able bucket)
to test upload

run "find . -name .svn -prune -o -size +1000k -type f -print0 | xargs -0 rm" to remove the files uploaded.  Check that the large binary files have been removed

run "download_from_google_storage.py -r -d -b chrome-artifacts ." to download the files again.

Review URL: https://chromiumcodereview.appspot.com/12042069

git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@187951 0039d316-1c4b-4281-b951-d872f2087c98

tests/gstools/download_test_data/rootfolder_text.txt

+This is a test file.

tests/gstools/download_test_data/rootfolder_text.txt.sha1

+e6c4fbd4fe7607f3e6ebf68b2ea4ef694da7b4fe

tests/gstools/download_test_data/subfolder/subfolder_text.txt

+This is a test file.
+This file exists in a subfolder

tests/gstools/download_test_data/subfolder/subfolder_text.txt.sha1

+b5415aa0b64006a95c0c409182e628881d6d6463

tests/gstools/download_test_data/uploaded_lorem_ipsum.txt.sha1

+7871c8e24da15bad8b0be2c36edc9dc77e37727f

tests/gstools/lorem_ipsum.txt

+Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod
+tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam,
+quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo
+consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse
+cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non
+proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

tests/gstools/lorem_ipsum.txt.md5

+734d7c1ed3545383837428f031840a1e

tests/gstools/lorem_ipsum2.txt

+Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod
+tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam,
+quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo
+consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse
+cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non
+proident, sunt in culpa qui officia deserunt mollit anim id est laborum.
+This is the second file.

tests/upload_to_google_storage_unittests.py

+#!/usr/bin/env python
+# Copyright (c) 2012 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.
+
+"""Unit tests for upload_to_google_storage.py."""
+
+import optparse
+import os
+import Queue
+import shutil
+import StringIO
+import sys
+import tempfile
+import threading
+import unittest
+
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+import upload_to_google_storage
+from download_from_google_storage_unittests import GsutilMock
+
+# ../third_party/gsutil/gsutil
+GSUTIL_DEFAULT_PATH = os.path.join(
+    os.path.dirname(os.path.dirname(os.path.abspath(__file__))),
+    'third_party', 'gsutil', 'gsutil')
+TEST_DIR = os.path.dirname(os.path.abspath(__file__))
+
+
+class UploadTests(unittest.TestCase):
+  def setUp(self):
+    self.gsutil = GsutilMock(GSUTIL_DEFAULT_PATH)
+    self.temp_dir = tempfile.mkdtemp(prefix='gstools_test')
+    self.base_path = os.path.join(self.temp_dir, 'gstools')
+    shutil.copytree(os.path.join(TEST_DIR, 'gstools'), self.base_path)
+    self.base_url = 'gs://sometesturl'
+    self.parser = optparse.OptionParser()
+    self.ret_codes = Queue.Queue()
+    self.stdout_queue = Queue.Queue()
+    self.lorem_ipsum = os.path.join(self.base_path, 'lorem_ipsum.txt')
+    self.lorem_ipsum_sha1 = '7871c8e24da15bad8b0be2c36edc9dc77e37727f'
+
+  def cleanUp(self):
+    shutil.rmtree(self.temp_dir)
+    sys.stdin = sys.__stdin__
+
+  def test_upload_single_file(self):
+    filenames = [self.lorem_ipsum]
+    output_filename = '%s.sha1'  % self.lorem_ipsum
+    code = upload_to_google_storage.upload_to_google_storage(
+        filenames, self.base_url, self.gsutil, True, False, 1, False)
+    self.assertEqual(
+        self.gsutil.history,
+        [('check_call',
+          ('ls', '%s/%s' % (self.base_url, self.lorem_ipsum_sha1))),
+         ('check_call',
+          ('cp', '-q', filenames[0], '%s/%s' % (self.base_url,
+                                                self.lorem_ipsum_sha1)))])
+    self.assertTrue(os.path.exists(output_filename))
+    self.assertEqual(
+        open(output_filename, 'rb').read(),
+        '7871c8e24da15bad8b0be2c36edc9dc77e37727f')
+    os.remove(output_filename)
+    self.assertEqual(code, 0)
+
+  def test_upload_single_file_remote_exists(self):
+    filenames = [self.lorem_ipsum]
+    output_filename = '%s.sha1'  % self.lorem_ipsum
+    etag_string = 'ETag: 634d7c1ed3545383837428f031840a1e'
+    self.gsutil.add_expected(0, '', '')
+    self.gsutil.add_expected(0, etag_string, '')
+    code = upload_to_google_storage.upload_to_google_storage(
+        filenames, self.base_url, self.gsutil, False, False, 1, False)
+    self.assertEqual(
+        self.gsutil.history,
+        [('check_call',
+          ('ls', '%s/%s' % (self.base_url, self.lorem_ipsum_sha1))),
+         ('check_call',
+          ('ls', '-L', '%s/%s' % (self.base_url, self.lorem_ipsum_sha1)))])
+    self.assertTrue(os.path.exists(output_filename))
+    self.assertEqual(
+        open(output_filename, 'rb').read(),
+        '7871c8e24da15bad8b0be2c36edc9dc77e37727f')
+    os.remove(output_filename)
+    self.assertEqual(code, 0)
+
+  def test_upload_worker_errors(self):
+    work_queue = Queue.Queue()
+    work_queue.put((self.lorem_ipsum, self.lorem_ipsum_sha1))
+    work_queue.put((None, None))
+    self.gsutil.add_expected(1, '', '')  # For the first ls call.
+    self.gsutil.add_expected(20, '', 'Expected error message')
+    # pylint: disable=W0212
+    upload_to_google_storage._upload_worker(
+        0,
+        work_queue,
+        self.base_url,
+        self.gsutil,
+        threading.Lock(),
+        False,
+        False,
+        self.stdout_queue,
+        self.ret_codes)
+    expected_ret_codes = [
+      (20,
+       'Encountered error on uploading %s to %s/%s\nExpected error message' %
+          (self.lorem_ipsum, self.base_url, self.lorem_ipsum_sha1))]
+    self.assertEqual(list(self.ret_codes.queue), expected_ret_codes)
+
+  def test_skip_hashing(self):
+    filenames = [self.lorem_ipsum]
+    output_filename = '%s.sha1' % self.lorem_ipsum
+    fake_hash = '6871c8e24da15bad8b0be2c36edc9dc77e37727f'
+    with open(output_filename, 'wb') as f:
+      f.write(fake_hash)  # Fake hash.
+    code = upload_to_google_storage.upload_to_google_storage(
+        filenames, self.base_url, self.gsutil, False, False, 1, True)
+    self.assertEqual(
+        self.gsutil.history,
+        [('check_call',
+          ('ls', '%s/%s' % (self.base_url, fake_hash))),
+         ('check_call',
+          ('ls', '-L', '%s/%s' % (self.base_url, fake_hash))),
+         ('check_call',
+          ('cp', '-q', filenames[0], '%s/%s' % (self.base_url, fake_hash)))])
+    self.assertEqual(
+        open(output_filename, 'rb').read(), fake_hash)
+    os.remove(output_filename)
+    self.assertEqual(code, 0)
+
+  def test_get_targets_no_args(self):
+    try:
+      upload_to_google_storage.get_targets([], self.parser, False)
+      self.fail()
+    except SystemExit, e:
+      self.assertEqual(e.code, 2)
+
+  def test_get_targets_passthrough(self):
+    result = upload_to_google_storage.get_targets(
+        ['a', 'b', 'c', 'd', 'e'],
+        self.parser,
+        False)
+    self.assertEqual(result, ['a', 'b', 'c', 'd', 'e'])
+
+  def test_get_targets_multiple_stdin(self):
+    inputs = ['a', 'b', 'c', 'd', 'e']
+    sys.stdin = StringIO.StringIO(os.linesep.join(inputs))
+    result = upload_to_google_storage.get_targets(
+        ['-'],
+        self.parser,
+        False)
+    self.assertEqual(result, inputs)
+
+  def test_get_targets_multiple_stdin_null(self):
+    inputs = ['a', 'b', 'c', 'd', 'e']
+    sys.stdin = StringIO.StringIO('\0'.join(inputs))
+    result = upload_to_google_storage.get_targets(
+        ['-'],
+        self.parser,
+        True)
+    self.assertEqual(result, inputs)
+
+
+if __name__ == '__main__':
+  unittest.main()
\ No newline at end of file

upload_to_google_storage.py

+#!/usr/bin/env python
+# Copyright (c) 2012 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.
+
+"""Uploads files to Google Storage content addressed."""
+
+import hashlib
+import optparse
+import os
+import Queue
+import re
+import sys
+import threading
+import time
+
+from download_from_google_storage import check_bucket_permissions
+from download_from_google_storage import get_sha1
+from download_from_google_storage import Gsutil
+from download_from_google_storage import printer_worker
+
+GSUTIL_DEFAULT_PATH = os.path.join(
+    os.path.dirname(os.path.abspath(__file__)),
+    'third_party', 'gsutil', 'gsutil')
+
+USAGE_STRING = """%prog [options] target [target2 ...].
+Target is the file intended to be uploaded to Google Storage.
+If target is "-", then a list of files will be taken from standard input
+
+This script will generate a file (original filename).sha1 containing the
+sha1 sum of the uploaded file.
+It is recommended that the .sha1 file is checked into the repository,
+the original file removed from the repository, and a hook added to the
+DEPS file to call download_from_google_storage.py.
+
+Example usages
+--------------
+
+Scan the current directory and upload all files larger than 1MB:
+find . -name .svn -prune -o -size +1000k -type f -print0 | %prog -0 -b bkt -
+(Replace "bkt" with the name of a writable bucket.)
+"""
+
+
+def get_md5(filename):
+  md5_calculator = hashlib.md5()
+  with open(filename, 'rb') as f:
+    while True:
+      chunk = f.read(1024*1024)
+      if not chunk:
+        break
+      md5_calculator.update(chunk)
+  return md5_calculator.hexdigest()
+
+
+def get_md5_cached(filename):
+  """Don't calculate the MD5 if we can find a .md5 file."""
+  # See if we can find an existing MD5 sum stored in a file.
+  if os.path.exists('%s.md5' % filename):
+    with open('%s.md5' % filename, 'rb') as f:
+      md5_match = re.search('([a-z0-9]{32})', f.read())
+      if md5_match:
+        return md5_match.group(1)
+  else:
+    md5_hash = get_md5(filename)
+    with open('%s.md5' % filename, 'wb') as f:
+      f.write(md5_hash)
+    return md5_hash
+
+
+def _upload_worker(
+    thread_num, upload_queue, base_url, gsutil, md5_lock, force,
+    use_md5, stdout_queue, ret_codes):
+  while True:
+    filename, sha1_sum = upload_queue.get()
+    if not filename:
+      break
+    file_url = '%s/%s' % (base_url, sha1_sum)
+    if gsutil.check_call('ls', file_url)[0] == 0 and not force:
+      # File exists, check MD5 hash.
+      _, out, _ = gsutil.check_call('ls', '-L', file_url)
+      etag_match = re.search('ETag:\s+([a-z0-9]{32})', out)
+      if etag_match:
+        remote_md5 = etag_match.group(1)
+        # Calculate the MD5 checksum to match it to Google Storage's ETag.
+        with md5_lock:
+          if use_md5:
+            local_md5 = get_md5_cached(filename)
+          else:
+            local_md5 = get_md5(filename)
+        if local_md5 == remote_md5:
+          stdout_queue.put(
+              '%d> File %s already exists and MD5 matches, upload skipped' %
+              (thread_num, filename))
+          continue
+    stdout_queue.put('%d> Uploading %s...' % (
+        thread_num, filename))
+    code, _, err = gsutil.check_call('cp', '-q', filename, file_url)
+    if code != 0:
+      ret_codes.put(
+          (code,
+           'Encountered error on uploading %s to %s\n%s' %
+           (filename, file_url, err)))
+      continue
+
+
+def get_targets(args, parser, use_null_terminator):
+  if not args:
+    parser.error('Missing target.')
+
+  if len(args) == 1 and args[0] == '-':
+    # Take stdin as a newline or null seperated list of files.
+    if use_null_terminator:
+      return sys.stdin.read().split('\0')
+    else:
+      return sys.stdin.read().splitlines()
+  else:
+    return args
+
+
+def upload_to_google_storage(
+    input_filenames, base_url, gsutil, force,
+    use_md5, num_threads, skip_hashing):
+  # We only want one MD5 calculation happening at a time to avoid HD thrashing.
+  md5_lock = threading.Lock()
+
+  # Start up all the worker threads plus the printer thread.
+  all_threads = []
+  ret_codes = Queue.Queue()
+  ret_codes.put((0, None))
+  upload_queue = Queue.Queue()
+  upload_timer = time.time()
+  stdout_queue = Queue.Queue()
+  printer_thread = threading.Thread(target=printer_worker, args=[stdout_queue])
+  printer_thread.daemon = True
+  printer_thread.start()
+  for thread_num in range(num_threads):
+    t = threading.Thread(
+        target=_upload_worker,
+        args=[thread_num, upload_queue, base_url, gsutil, md5_lock,
+              force, use_md5, stdout_queue, ret_codes])
+    t.daemon = True
+    t.start()
+    all_threads.append(t)
+
+  # We want to hash everything in a single thread since its faster.
+  # The bottleneck is in disk IO, not CPU.
+  hashing_start = time.time()
+  for filename in input_filenames:
+    if not os.path.exists(filename):
+      stdout_queue.put('Main> Error: %s not found, skipping.' % filename)
+      continue
+    if os.path.exists('%s.sha1' % filename) and skip_hashing:
+      stdout_queue.put(
+          'Main> Found hash for %s, sha1 calculation skipped.' % filename)
+      with open(filename + '.sha1', 'rb') as f:
+        sha1_file = f.read(1024)
+      if not re.match('^([a-z0-9]{40})$', sha1_file):
+        print >> sys.stderr, 'Invalid sha1 hash file %s.sha1' % filename
+        return 1
+      upload_queue.put((filename, sha1_file))
+      continue
+    stdout_queue.put('Main> Calculating hash for %s...' % filename)
+    sha1_sum = get_sha1(filename)
+    with open(filename + '.sha1', 'wb') as f:
+      f.write(sha1_sum)
+    stdout_queue.put('Main> Done calculating hash for %s.' % filename)
+    upload_queue.put((filename, sha1_sum))
+  hashing_duration = time.time() - hashing_start
+
+  # Wait for everything to finish.
+  for _ in all_threads:
+    upload_queue.put((None, None))  # To mark the end of the work queue.
+  for t in all_threads:
+    t.join()
+  stdout_queue.put(None)
+  printer_thread.join()
+
+  # Print timing information.
+  print 'Hashing %s files took %1f seconds' % (
+      len(input_filenames), hashing_duration)
+  print 'Uploading took %1f seconds' % (time.time() - upload_timer)
+
+  # See if we ran into any errors.
+  max_ret_code = 0
+  for ret_code, message in ret_codes.queue:
+    max_ret_code = max(ret_code, max_ret_code)
+    if message:
+      print >> sys.stderr, message
+
+  if not max_ret_code:
+    print 'Success!'
+
+  return max_ret_code
+
+
+def main(args):
+  parser = optparse.OptionParser(USAGE_STRING)
+  parser.add_option('-b', '--bucket',
+                    help='Google Storage bucket to upload to.')
+  parser.add_option('-e', '--boto', help='Specify a custom boto file.')
+  parser.add_option('-f', '--force', action='store_true',
+                    help='Force upload even if remote file exists.')
+  parser.add_option('-g', '--gsutil_path', default=GSUTIL_DEFAULT_PATH,
+                    help='Path to the gsutil script.')
+  parser.add_option('-m', '--use_md5', action='store_true',
+                    help='Generate MD5 files when scanning, and don\'t check '
+                    'the MD5 checksum if a .md5 file is found.')
+  parser.add_option('-t', '--num_threads', default=1, type='int',
+                    help='Number of uploader threads to run.')
+  parser.add_option('-s', '--skip_hashing', action='store_true',
+                    help='Skip hashing if .sha1 file exists.')
+  parser.add_option('-0', '--use_null_terminator', action='store_true',
+                    help='Use \\0 instead of \\n when parsing '
+                    'the file list from stdin.  This is useful if the input '
+                    'is coming from "find ... -print0".')
+  (options, args) = parser.parse_args()
+
+  # Enumerate our inputs.
+  input_filenames = get_targets(args, parser, options.use_null_terminator)
+
+  # Make sure we can find a working instance of gsutil.
+  if os.path.exists(GSUTIL_DEFAULT_PATH):
+    gsutil = Gsutil(GSUTIL_DEFAULT_PATH)
+  else:
+    gsutil = None
+    for path in os.environ["PATH"].split(os.pathsep):
+      if os.path.exists(path) and 'gsutil' in os.listdir(path):
+        gsutil = Gsutil(os.path.join(path, 'gsutil'))
+    if not gsutil:
+      parser.error('gsutil not found in %s, bad depot_tools checkout?' %
+                   GSUTIL_DEFAULT_PATH)
+
+  # Check we have a valid bucket with valid permissions.
+  base_url, code = check_bucket_permissions(options.bucket, gsutil)
+  if code:
+    return code
+
+  return upload_to_google_storage(
+      input_filenames, base_url, gsutil, options.force, options.use_md5,
+      options.num_threads, options.skip_hashing)
+
+
+if __name__ == '__main__':
+  sys.exit(main(sys.argv))
\ No newline at end of file