#!/usr/bin/env python
# Copyright 2013 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.

"""Downloads and unpacks a toolchain for building on Windows. The contents are
matched by sha1 which will be updated when the toolchain is updated.

Having a toolchain script in depot_tools means that it's not versioned
directly with the source code. That is, if the toolchain is upgraded, but
you're trying to build an historical version of Chromium from before the
toolchain upgrade, this will cause you to build with a newer toolchain than
was available when that code was committed. This is done for a two main
reasons: 1) it would likely be annoying to have the up-to-date toolchain
removed and replaced by one without a service pack applied); 2) it would
require maintaining scripts that can build older not-up-to-date revisions of
the toolchain. This is likely to be a poorly tested code path that probably
won't be properly maintained. See http://crbug.com/323300.

This does not extend to major versions of the toolchain however, on the
assumption that there are more likely to be source incompatibilities between
major revisions. This script calls a subscript (currently, toolchain2013.py)
to do the main work. It is expected that toolchain2013.py will always be able
to acquire/build the most current revision of a VS2013-based toolchain. In the
future when a hypothetical VS2015 is released, the 2013 script will be
maintained, and a new 2015 script would be added.
"""

import hashlib
import json
import optparse
import os
import shutil
import subprocess
import sys
import tempfile
import time
import zipfile


BASEDIR = os.path.dirname(os.path.abspath(__file__))
DEPOT_TOOLS_PATH = os.path.join(BASEDIR, '..')
sys.path.append(DEPOT_TOOLS_PATH)
try:
  import download_from_google_storage
except ImportError:
  # Allow use of utility functions in this script from package_from_installed
  # on bare VM that doesn't have a full depot_tools.
  pass


def GetFileList(root):
  """Gets a normalized list of files under |root|."""
  assert not os.path.isabs(root)
  assert os.path.normpath(root) == root
  file_list = []
  for base, _, files in os.walk(root):
    paths = [os.path.join(base, f) for f in files]
    file_list.extend(x.lower() for x in paths)
  return sorted(file_list)


def MakeTimestampsFileName(root):
  return os.path.join(root, '..', '.timestamps')


def CalculateHash(root):
  """Calculates the sha1 of the paths to all files in the given |root| and the
  contents of those files, and returns as a hex string."""
  file_list = GetFileList(root)

  # Check whether we previously saved timestamps in $root/../.timestamps. If
  # we didn't, or they don't match, then do the full calculation, otherwise
  # return the saved value.
  timestamps_file = MakeTimestampsFileName(root)
  timestamps_data = {'files': [], 'sha1': ''}
  if os.path.exists(timestamps_file):
    with open(timestamps_file, 'rb') as f:
      try:
        timestamps_data = json.load(f)
      except ValueError:
        # json couldn't be loaded, empty data will force a re-hash.
        pass

  matches = len(file_list) == len(timestamps_data['files'])
  if matches:
    for disk, cached in zip(file_list, timestamps_data['files']):
      if disk != cached[0] or os.stat(disk).st_mtime != cached[1]:
        matches = False
        break
  if matches:
    return timestamps_data['sha1']

  digest = hashlib.sha1()
  for path in file_list:
    digest.update(path)
    with open(path, 'rb') as f:
      digest.update(f.read())
  return digest.hexdigest()


def SaveTimestampsAndHash(root, sha1):
  """Saves timestamps and the final hash to be able to early-out more quickly
  next time."""
  file_list = GetFileList(root)
  timestamps_data = {
    'files': [[f, os.stat(f).st_mtime] for f in file_list],
    'sha1': sha1,
  }
  with open(MakeTimestampsFileName(root), 'wb') as f:
    json.dump(timestamps_data, f)


def HaveSrcInternalAccess():
  """Checks whether access to src-internal is available."""
  with open(os.devnull, 'w') as nul:
    if subprocess.call(
        ['svn', 'ls', '--non-interactive',
         'svn://svn.chromium.org/chrome-internal/trunk/src-internal/'],
        shell=True, stdin=nul, stdout=nul, stderr=nul) == 0:
      return True
    return subprocess.call(
        ['git', '-c', 'core.askpass=true', 'remote', 'show',
         'https://chrome-internal.googlesource.com/chrome/src-internal/'],
        shell=True, stdin=nul, stdout=nul, stderr=nul) == 0


def LooksLikeGoogler():
  """Checks for a USERDOMAIN environment variable of 'GOOGLE', which
  probably implies the current user is a Googler."""
  return os.environ.get('USERDOMAIN', '').upper() == 'GOOGLE'


def CanAccessToolchainBucket():
  """Checks whether the user has access to gs://chrome-wintoolchain/."""
  gsutil = download_from_google_storage.Gsutil(
      download_from_google_storage.GSUTIL_DEFAULT_PATH, boto_path=None)
  code, _, _ = gsutil.check_call('ls', 'gs://chrome-wintoolchain/')
  return code == 0


def RequestGsAuthentication():
  """Requests that the user authenticate to be able to access gs:// as a
  Googler. This allows much faster downloads, and pulling (old) toolchains
  that match src/ revisions.
  """
  print 'Access to gs://chrome-wintoolchain/ not configured.'
  print '-----------------------------------------------------------------'
  print
  print 'You appear to be a Googler.'
  print
  print 'I\'m sorry for the hassle, but you need to do a one-time manual'
  print 'authentication. Please run:'
  print
  print '    download_from_google_storage --config'
  print
  print 'and follow the instructions.'
  print
  print 'NOTE 1: Use your google.com credentials, not chromium.org.'
  print 'NOTE 2: Enter 0 when asked for a "project-id".'
  print
  print '-----------------------------------------------------------------'
  print
  sys.stdout.flush()
  sys.exit(1)


def DelayBeforeRemoving(target_dir):
  """A grace period before deleting the out of date toolchain directory."""
  if (os.path.isdir(target_dir) and
      not bool(int(os.environ.get('CHROME_HEADLESS', '0')))):
    for i in range(9, 0, -1):
      sys.stdout.write(
              '\rRemoving old toolchain in %ds... (Ctrl-C to cancel)' % i)
      sys.stdout.flush()
      time.sleep(1)
    print


def DownloadUsingGsutil(filename):
  """Downloads the given file from Google Storage chrome-wintoolchain bucket."""
  temp_dir = tempfile.mkdtemp()
  assert os.path.basename(filename) == filename
  target_path = os.path.join(temp_dir, filename)
  gsutil = download_from_google_storage.Gsutil(
      download_from_google_storage.GSUTIL_DEFAULT_PATH, boto_path=None)
  code = gsutil.call('cp', 'gs://chrome-wintoolchain/' + filename, target_path)
  if code != 0:
    sys.exit('gsutil failed')
  return temp_dir, target_path


def RmDir(path):
  """Deletes path and all the files it contains."""
  if sys.platform != 'win32':
    shutil.rmtree(path, ignore_errors=True)
  else:
    # shutil.rmtree() doesn't delete read-only files on Windows.
    subprocess.check_call('rmdir /s/q "%s"' % path, shell=True)


def DoTreeMirror(target_dir, tree_sha1):
  """In order to save temporary space on bots that do not have enough space to
  download ISOs, unpack them, and copy to the target location, the whole tree
  is uploaded as a zip to internal storage, and then mirrored here."""
  temp_dir, local_zip = DownloadUsingGsutil(tree_sha1 + '.zip')
  sys.stdout.write('Extracting %s...\n' % local_zip)
  sys.stdout.flush()
  with zipfile.ZipFile(local_zip, 'r', zipfile.ZIP_DEFLATED, True) as zf:
    zf.extractall(target_dir)
  if temp_dir:
    RmDir(temp_dir)


def main():
  if not sys.platform.startswith(('cygwin', 'win32')):
    return 0

  parser = optparse.OptionParser(description=sys.modules[__name__].__doc__)
  parser.add_option('--output-json', metavar='FILE',
                    help='write information about toolchain to FILE')
  options, args = parser.parse_args()

  if sys.platform == 'cygwin':
    # This script requires Windows Python, so invoke with depot_tools' Python.
    def winpath(path):
      return subprocess.check_output(['cygpath', '-w', path]).strip()
    python = os.path.join(DEPOT_TOOLS_PATH, 'python.bat')
    cmd = [python, winpath(__file__)]
    if options.output_json:
      cmd.extend(['--output-json', winpath(options.output_json)])
    cmd.extend(args)
    sys.exit(subprocess.call(cmd))
  assert sys.platform != 'cygwin'

  # We assume that the Pro hash is the first one.
  desired_hashes = args
  if len(desired_hashes) == 0:
    sys.exit('Desired hashes are required.')

  # Move to depot_tools\win_toolchain where we'll store our files, and where
  # the downloader script is.
  os.chdir(os.path.normpath(os.path.join(BASEDIR)))
  toolchain_dir = '.'
  if os.environ.get('GYP_MSVS_VERSION') == '2015':
    target_dir = os.path.normpath(os.path.join(toolchain_dir, 'vs_files'))
  else:
    target_dir = os.path.normpath(os.path.join(toolchain_dir, 'vs2013_files'))
  abs_target_dir = os.path.abspath(target_dir)

  got_new_toolchain = False

  # If the current hash doesn't match what we want in the file, nuke and pave.
  # Typically this script is only run when the .sha1 one file is updated, but
  # directly calling "gclient runhooks" will also run it, so we cache
  # based on timestamps to make that case fast.
  current_hash = CalculateHash(target_dir)
  if current_hash not in desired_hashes:
    should_use_gs = False
    if (HaveSrcInternalAccess() or
        LooksLikeGoogler() or
        CanAccessToolchainBucket()):
      should_use_gs = True
      if not CanAccessToolchainBucket():
        RequestGsAuthentication()
    if not should_use_gs:
      print('Please follow the instructions at '
            'http://www.chromium.org/developers/how-tos/'
            'build-instructions-windows')
      return 1
    print('Windows toolchain out of date or doesn\'t exist, updating (Pro)...')
    print('  current_hash: %s' % current_hash)
    print('  desired_hashes: %s' % ', '.join(desired_hashes))
    sys.stdout.flush()
    DelayBeforeRemoving(target_dir)
    if sys.platform == 'win32':
      # This stays resident and will make the rmdir below fail.
      with open(os.devnull, 'wb') as nul:
        subprocess.call(['taskkill', '/f', '/im', 'mspdbsrv.exe'],
                        stdin=nul, stdout=nul, stderr=nul)
    if os.path.isdir(target_dir):
      RmDir(target_dir)

    DoTreeMirror(target_dir, desired_hashes[0])

    got_new_toolchain = True

  win_sdk = os.path.join(abs_target_dir, 'win_sdk')
  try:
    with open(os.path.join(target_dir, 'VS_VERSION'), 'rb') as f:
      vs_version = f.read().strip()
  except IOError:
    # Older toolchains didn't have the VS_VERSION file, and used 'win8sdk'
    # instead of just 'win_sdk'.
    vs_version = '2013'
    win_sdk = os.path.join(abs_target_dir, 'win8sdk')

  data = {
      'path': abs_target_dir,
      'version': vs_version,
      'win_sdk': win_sdk,
      # Added for backwards compatibility with old toolchain packages.
      'win8sdk': win_sdk,
      'wdk': os.path.join(abs_target_dir, 'wdk'),
      'runtime_dirs': [
        os.path.join(abs_target_dir, 'sys64'),
        os.path.join(abs_target_dir, 'sys32'),
      ],
  }
  with open(os.path.join(target_dir, '..', 'data.json'), 'w') as f:
    json.dump(data, f)

  if got_new_toolchain:
    current_hash = CalculateHash(target_dir)
    if current_hash not in desired_hashes:
      print >> sys.stderr, (
          'Got wrong hash after pulling a new toolchain. '
          'Wanted one of \'%s\', got \'%s\'.' % (
              ', '.join(desired_hashes), current_hash))
      return 1
    SaveTimestampsAndHash(target_dir, current_hash)

  if options.output_json:
    shutil.copyfile(os.path.join(target_dir, '..', 'data.json'),
                    options.output_json)

  return 0


if __name__ == '__main__':
  sys.exit(main())