[telemetry] Script for finding all Telemetry dependencies.

Given a list of Python modules, it will find all:
- bootstrap_deps folders, recursively.
- Python dependencies.
- Page sets in those modules' folders.
- Serving dirs of those page sets.
And print a list of all files, or zip them up.

Example bot command: tools/telemetry_tools/find_dependencies -v tools/perf/run_benchmark tools/perf/run_measurement tools/perf/record_wpr content/test/gpu/run_gpu_test -z telemetry.zip


BUG=311380
TEST=Ran it.

Review URL: https://codereview.chromium.org/103433002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@251298 0039d316-1c4b-4281-b951-d872f2087c98

tools/perf/benchmarks/peacekeeper.py

@@ -22,7 +22,7 @@ from telemetry.page import page_measurement
 from telemetry.page import page_set
 from telemetry.value import merge_values
 
-class PeaceKeeperMeasurement(page_measurement.PageMeasurement):
+class _PeaceKeeperMeasurement(page_measurement.PageMeasurement):
 
   def WillNavigateToPage(self, page, tab):
     page.script_to_evaluate_on_commit = """
@@ -68,9 +68,9 @@ class PeaceKeeperMeasurement(page_measurement.PageMeasurement):
     results.AddSummary('Score', 'score', total, 'Total')
 
 
-class PeaceKeeperBenchmark(test.Test):
+class _PeaceKeeperBenchmark(test.Test):
   """A base class for Peackeeper benchmarks."""
-  test = PeaceKeeperMeasurement
+  test = _PeaceKeeperMeasurement
 
   def CreatePageSet(self, options):
     """Makes a PageSet for PeaceKeeper benchmarks."""
@@ -98,7 +98,7 @@ class PeaceKeeperBenchmark(test.Test):
     return page_set.PageSet.FromDict(page_set_dict, os.path.abspath(__file__))
 
 
-class PeaceKeeperRender(PeaceKeeperBenchmark):
+class PeaceKeeperRender(_PeaceKeeperBenchmark):
   """PeaceKeeper rendering benchmark suite.
 
   These tests measure your browser's ability to render and modify specific
@@ -113,7 +113,7 @@ class PeaceKeeperRender(PeaceKeeperBenchmark):
                ]
 
 
-class PeaceKeeperData(PeaceKeeperBenchmark):
+class PeaceKeeperData(_PeaceKeeperBenchmark):
   """PeaceKeeper Data operations benchmark suite.
 
   These tests measure your browser's ability to add, remove and modify data
@@ -132,7 +132,7 @@ class PeaceKeeperData(PeaceKeeperBenchmark):
                ]
 
 
-class PeaceKeeperDom(PeaceKeeperBenchmark):
+class PeaceKeeperDom(_PeaceKeeperBenchmark):
   """PeaceKeeper DOM operations benchmark suite.
 
   These tests emulate the methods used to create typical dynamic webpages.
@@ -172,7 +172,7 @@ class PeaceKeeperDom(PeaceKeeperBenchmark):
                ]
 
 
-class PeaceKeeperTextParsing(PeaceKeeperBenchmark):
+class PeaceKeeperTextParsing(_PeaceKeeperBenchmark):
   """PeaceKeeper Text Parsing benchmark suite.
 
   These tests measure your browser's performance in typical text manipulations
@@ -201,7 +201,7 @@ class PeaceKeeperTextParsing(PeaceKeeperBenchmark):
                ]
 
 
-class PeaceKeeperHTML5Canvas(PeaceKeeperBenchmark):
+class PeaceKeeperHTML5Canvas(_PeaceKeeperBenchmark):
   """PeaceKeeper HTML5 Canvas benchmark suite.
 
   These tests use HTML5 Canvas, which is a web technology for drawing and
@@ -218,7 +218,7 @@ class PeaceKeeperHTML5Canvas(PeaceKeeperBenchmark):
                ]
 
 
-class PeaceKeeperHTML5Capabilities(PeaceKeeperBenchmark):
+class PeaceKeeperHTML5Capabilities(_PeaceKeeperBenchmark):
   """PeaceKeeper HTML5 Capabilities benchmark suite.
 
   These tests checks browser HTML5 capabilities support for WebGL, Video

tools/perf/generate_profile

@@ -5,7 +5,8 @@
 import os
 import sys
 
-sys.path.append(os.path.join(os.path.dirname(__file__), os.pardir, 'telemetry'))
+sys.path.append(os.path.join(
+  os.path.dirname(os.path.realpath(__file__)), os.pardir, 'telemetry'))
 
 from telemetry.page import profile_generator
 

tools/perf/record_android_profile.py

@@ -7,7 +7,8 @@ import os
 import sys
 import tempfile
 
-sys.path.append(os.path.join(os.path.dirname(__file__), os.pardir, 'telemetry'))
+sys.path.append(os.path.join(
+  os.path.dirname(os.path.realpath(__file__)), os.pardir, 'telemetry'))
 
 from telemetry.core import browser_finder
 from telemetry.core import browser_options

tools/perf/record_wpr

@@ -7,7 +7,7 @@ import sys
 
 # Add telemetry to sys.path so that it can be imported below
 # Relative path from this file is ../telemetry
-_perf_dir = os.path.dirname(__file__)
+_perf_dir = os.path.dirname(os.path.realpath(__file__))
 _telemetry_path = os.path.join(_perf_dir, os.pardir, 'telemetry')
 sys.path.append(_telemetry_path)
 

tools/perf/run_benchmark

@@ -6,7 +6,8 @@
 import os
 import sys
 
-sys.path.append(os.path.join(os.path.dirname(__file__), os.pardir, 'telemetry'))
+sys.path.append(os.path.join(
+  os.path.dirname(os.path.realpath(__file__)), os.pardir, 'telemetry'))
 
 from telemetry import test_runner
 

tools/perf/run_measurement

@@ -12,7 +12,7 @@ import urllib2
 BASE_URL = 'http://src.chromium.org/chrome/trunk/'
 DEPS_FILE = 'bootstrap_deps'
 
-SCRIPT_PATH = os.path.dirname(os.path.abspath(__file__))
+SCRIPT_PATH = os.path.dirname(os.path.realpath(__file__))
 # Directory containing src/ in a Chromium checkout.
 CHECKOUT_BASE_PATH = os.path.join(SCRIPT_PATH, os.pardir, os.pardir, os.pardir)
 # Directory in which to save bootstrap files.

tools/perf/run_tests

@@ -11,13 +11,14 @@ This script DOES NOT run benchmarks. run_benchmarks and run_measurement do that.
 import os
 import sys
 
-sys.path.append(os.path.join(os.path.dirname(__file__), os.pardir, 'telemetry'))
+sys.path.append(os.path.join(
+  os.path.dirname(os.path.realpath(__file__)), os.pardir, 'telemetry'))
 from telemetry.unittest import gtest_testrunner
 from telemetry.unittest import run_tests
 
 
 if __name__ == '__main__':
-  top_level_dir = os.path.abspath(os.path.dirname(__file__))
+  top_level_dir = os.path.abspath(os.path.dirname(os.path.realpath(__file__)))
   runner = gtest_testrunner.GTestTestRunner(print_result_after_run=False)
   ret = run_tests.Main(sys.argv[1:], top_level_dir, top_level_dir, runner)
 

tools/telemetry/run_tests

@@ -11,7 +11,7 @@ from telemetry.unittest import run_tests
 
 
 if __name__ == '__main__':
-  top_level_dir = os.path.abspath(os.path.dirname(__file__))
+  top_level_dir = os.path.abspath(os.path.dirname(os.path.realpath(__file__)))
   runner = gtest_testrunner.GTestTestRunner(print_result_after_run=False)
   ret = run_tests.Main(sys.argv[1:], top_level_dir, top_level_dir, runner)
 

tools/telemetry/telemetry/core/util.py

@@ -16,14 +16,14 @@ class TimeoutException(Exception):
 def GetBaseDir():
   main_module = sys.modules['__main__']
   if hasattr(main_module, '__file__'):
-    return os.path.dirname(os.path.abspath(main_module.__file__))
+    return os.path.dirname(os.path.realpath(main_module.__file__))
   else:
     return os.getcwd()
 
 
 def GetTelemetryDir():
   return os.path.normpath(os.path.join(
-      __file__, os.pardir, os.pardir, os.pardir))
+      os.path.realpath(__file__), os.pardir, os.pardir, os.pardir))
 
 
 def GetUnittestDataDir():

tools/telemetry/telemetry/page/cloud_storage.py

@@ -27,7 +27,7 @@ _DOWNLOAD_PATH = os.path.join(util.GetTelemetryDir(), 'third_party', 'gsutil')
 class CloudStorageError(Exception):
   @staticmethod
   def _GetConfigInstructions(gsutil_path):
-    if _SupportsProdaccess(gsutil_path):
+    if SupportsProdaccess(gsutil_path):
       return 'Run prodaccess to authenticate.'
     else:
       return ('To configure your credentials:\n'
@@ -73,7 +73,7 @@ def _DownloadGsutil():
   return os.path.join(_DOWNLOAD_PATH, 'gsutil')
 
 
-def _FindGsutil():
+def FindGsutil():
   """Return the gsutil executable path. If we can't find it, download it."""
   # Look for a depot_tools installation.
   gsutil_path = _FindExecutableInPath(
@@ -90,7 +90,7 @@ def _FindGsutil():
   return _DownloadGsutil()
 
 
-def _SupportsProdaccess(gsutil_path):
+def SupportsProdaccess(gsutil_path):
   def GsutilSupportsProdaccess():
     with open(gsutil_path, 'r') as gsutil:
       return 'prodaccess' in gsutil.read()
@@ -99,8 +99,7 @@ def _SupportsProdaccess(gsutil_path):
 
 
 def _RunCommand(args):
-  gsutil_path = _FindGsutil()
-
+  gsutil_path = FindGsutil()
   gsutil = subprocess.Popen([sys.executable, gsutil_path] + args,
                             stdout=subprocess.PIPE, stderr=subprocess.PIPE)
   stdout, stderr = gsutil.communicate()

tools/telemetry_tools/find_dependencies

+#!/usr/bin/env python
+# Copyright 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.
+
+import collections
+import fnmatch
+import imp
+import logging
+import modulefinder
+import optparse
+import os
+import sys
+import zipfile
+
+sys.path.append(os.path.join(
+    os.path.dirname(os.path.realpath(__file__)), os.pardir, 'telemetry'))
+from telemetry import test
+from telemetry.core import discover
+from telemetry.core import util
+from telemetry.page import cloud_storage
+
+import path_set
+import telemetry_bootstrap
+
+
+DEPS_FILE = 'bootstrap_deps'
+
+
+def _InDirectory(subdirectory, directory):
+  subdirectory = os.path.realpath(subdirectory)
+  directory = os.path.realpath(directory)
+  common_prefix = os.path.commonprefix([subdirectory, directory])
+  return common_prefix == directory
+
+
+def FindBootstrapDependencies(base_dir):
+  deps_file = os.path.join(base_dir, DEPS_FILE)
+  if not os.path.exists(deps_file):
+    return []
+  deps_paths = telemetry_bootstrap.ListAllDepsPaths(deps_file)
+  return set(
+      os.path.realpath(os.path.join(util.GetChromiumSrcDir(), os.pardir, path))
+      for path in deps_paths)
+
+
+def FindPythonDependencies(module_path):
+  logging.info('Finding Python dependencies of %s' % module_path)
+
+  # Load the module to inherit its sys.path modifications.
+  imp.load_source(
+      os.path.splitext(os.path.basename(module_path))[0], module_path)
+
+  # Analyze the module for its imports.
+  finder = modulefinder.ModuleFinder()
+  finder.run_script(module_path)
+
+  # Filter for only imports in Chromium.
+  for module in finder.modules.itervalues():
+    # If it's an __init__.py, module.__path__ gives the package's folder.
+    module_path = module.__path__[0] if module.__path__ else module.__file__
+    if not module_path:
+      continue
+
+    module_path = os.path.realpath(module_path)
+    if not _InDirectory(module_path, util.GetChromiumSrcDir()):
+      continue
+
+    yield module_path
+
+
+def FindPageSetDependencies(base_dir):
+  logging.info('Finding page sets in %s' % base_dir)
+
+  # Add base_dir to path so our imports relative to base_dir will work.
+  sys.path.append(base_dir)
+  tests = discover.DiscoverClasses(base_dir, base_dir, test.Test,
+                                   index_by_class_name=True)
+
+  for test_class in tests.itervalues():
+    test_obj = test_class()
+
+    # Ensure the test's default options are set if needed.
+    parser = optparse.OptionParser()
+    test_obj.AddTestCommandLineOptions(parser)
+    options = optparse.Values()
+    for k, v in parser.get_default_values().__dict__.iteritems():
+      options.ensure_value(k, v)
+
+    # Page set paths are relative to their runner script, not relative to us.
+    util.GetBaseDir = lambda: base_dir
+    # TODO: Loading the page set will automatically download its Cloud Storage
+    # deps. This is really expensive, and we don't want to do this by default.
+    page_set = test_obj.CreatePageSet(options)
+
+    # Add all of its serving_dirs as dependencies.
+    for serving_dir in page_set.serving_dirs:
+      yield serving_dir
+    for page in page_set:
+      if page.is_file:
+        yield page.serving_dir
+
+
+def FindExcludedFiles(files, options):
+  def MatchesConditions(path, conditions):
+    for condition in conditions:
+      if condition(path):
+        return True
+    return False
+
+  # Define some filters for files.
+  def IsHidden(path):
+    for pathname_component in path.split(os.sep):
+      if pathname_component.startswith('.'):
+        return True
+    return False
+  def IsPyc(path):
+    return os.path.splitext(path)[1] == '.pyc'
+  def IsInCloudStorage(path):
+    return os.path.exists(path + '.sha1')
+  def MatchesExcludeOptions(path):
+    for pattern in options.exclude:
+      if (fnmatch.fnmatch(path, pattern) or
+          fnmatch.fnmatch(os.path.basename(path), pattern)):
+        return True
+    return False
+
+  # Collect filters we're going to use to exclude files.
+  exclude_conditions = [
+      IsHidden,
+      IsPyc,
+      IsInCloudStorage,
+      MatchesExcludeOptions,
+  ]
+
+  # Check all the files against the filters.
+  for path in files:
+    if MatchesConditions(path, exclude_conditions):
+      yield path
+
+
+def FindDependencies(paths, options):
+  # Verify arguments.
+  for path in paths:
+    if not os.path.exists(path):
+      raise ValueError('Path does not exist: %s' % path)
+
+  dependencies = path_set.PathSet()
+
+  # Including this file will include Telemetry and its dependencies.
+  # If the user doesn't pass any arguments, we just have Telemetry.
+  dependencies |= FindPythonDependencies(os.path.realpath(__file__))
+
+  # Add dependencies.
+  for path in paths:
+    base_dir = os.path.dirname(os.path.realpath(path))
+
+    dependencies.add(base_dir)
+    dependencies |= FindBootstrapDependencies(base_dir)
+    dependencies |= FindPythonDependencies(path)
+    if options.include_page_set_data:
+      dependencies |= FindPageSetDependencies(base_dir)
+
+  # Remove excluded files.
+  dependencies -= FindExcludedFiles(set(dependencies), options)
+
+  return dependencies
+
+
+def ZipDependencies(paths, dependencies, options):
+  base_dir = os.path.dirname(os.path.realpath(util.GetChromiumSrcDir()))
+
+  with zipfile.ZipFile(options.zip, 'w', zipfile.ZIP_DEFLATED) as zip_file:
+    # Add dependencies to archive.
+    for path in dependencies:
+      path_in_archive = os.path.join(
+          'telemetry', os.path.relpath(path, base_dir))
+      zip_file.write(path, path_in_archive)
+
+    # Add symlinks to executable paths, for ease of use.
+    for path in paths:
+      link_info = zipfile.ZipInfo(
+          os.path.join('telemetry', os.path.basename(path)))
+      link_info.create_system = 3  # Unix attributes.
+      # 010 is regular file, 0111 is the permission bits rwxrwxrwx.
+      link_info.external_attr = 0100777 << 16  # Octal.
+
+      relative_path = os.path.relpath(path, base_dir)
+      link_script = (
+          '#!/usr/bin/env python\n\n'
+          'import os\n'
+          'import sys\n\n\n'
+          'script = os.path.join(os.path.dirname(__file__), \'%s\')\n'
+          'os.execv(sys.executable, [sys.executable, script] + sys.argv[1:])'
+        % relative_path)
+
+      zip_file.writestr(link_info, link_script)
+
+    # Add gsutil to the archive, if it's available. The gsutil in
+    # depot_tools is modified to allow authentication using prodaccess.
+    # TODO: If there's a gsutil in telemetry/third_party/, bootstrap_deps
+    # will include it. Then there will be two copies of gsutil at the same
+    # location in the archive. This can be confusing for users.
+    gsutil_path = cloud_storage.FindGsutil()
+    if cloud_storage.SupportsProdaccess(gsutil_path):
+      gsutil_dependencies = path_set.PathSet()
+      gsutil_dependencies.add(os.path.dirname(gsutil_path))
+      gsutil_dependencies -= FindExcludedFiles(
+          set(gsutil_dependencies), options)
+
+      gsutil_base_dir = os.path.join(os.path.dirname(gsutil_path), os.pardir)
+      for path in gsutil_dependencies:
+        path_in_archive = os.path.join(
+            'telemetry', os.path.relpath(util.GetTelemetryDir(), base_dir),
+            'third_party', os.path.relpath(path, gsutil_base_dir))
+        zip_file.write(path, path_in_archive)
+
+
+def ParseCommandLine():
+  parser = optparse.OptionParser()
+  parser.add_option(
+      '-v', '--verbose', action='count', dest='verbosity',
+      help='Increase verbosity level (repeat as needed).')
+
+  parser.add_option(
+      '-p', '--include-page-set-data', action='store_true', default=False,
+      help='Scan tests for page set data and include them.')
+
+  parser.add_option(
+      '-e', '--exclude', action='append', default=[],
+      help='Exclude paths matching EXCLUDE. Can be used multiple times.')
+
+  parser.add_option(
+      '-z', '--zip',
+      help='Store files in a zip archive at ZIP.')
+
+  options, args = parser.parse_args()
+
+  if options.verbosity >= 2:
+    logging.getLogger().setLevel(logging.DEBUG)
+  elif options.verbosity:
+    logging.getLogger().setLevel(logging.INFO)
+  else:
+    logging.getLogger().setLevel(logging.WARNING)
+
+  return options, args
+
+
+def main():
+  options, paths = ParseCommandLine()
+
+  dependencies = FindDependencies(paths, options)
+
+  if options.zip:
+    ZipDependencies(paths, dependencies, options)
+    print 'Zip archive written to %s.' % options.zip
+  else:
+    print '\n'.join(sorted(dependencies))
+
+
+if __name__ == '__main__':
+  main()

tools/telemetry_tools/path_set.py

+# Copyright 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.
+
+import collections
+import os
+
+
+class PathSet(collections.MutableSet):
+  """A set of paths.
+
+  All mutation methods can take both directories or individual files, but the
+  iterator yields the individual files. All paths are automatically normalized.
+  """
+  def __init__(self, iterable=None):
+    self._paths = set()
+    if iterable:
+      self |= iterable
+
+  def __contains__(self, path):
+    return os.path.realpath(path) in self._paths
+
+  def __iter__(self):
+    return iter(self._paths)
+
+  def __len__(self):
+    return len(self._paths)
+
+  def add(self, path):
+    path = os.path.realpath(path)
+    if os.path.isfile(path):
+      self._paths.add(path)
+    for root, _, files in os.walk(path):
+      for basename in files:
+        file_path = os.path.join(root, basename)
+        if os.path.isfile(file_path):
+          self._paths.add(file_path)
+
+  def discard(self, path):
+    path = os.path.realpath(path)
+    self._paths.discard(path)
+    for root, _, files in os.walk(path):
+      for basename in files:
+        self._paths.discard(os.path.join(root, basename))

tools/telemetry_tools/path_set_unittest.py

+#!/usr/bin/env python
+# Copyright 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.
+
+import os
+import unittest
+
+import path_set
+
+
+class PathSetTest(unittest.TestCase):
+  def testCreate(self):
+    ps = path_set.PathSet()
+    self.assertEqual(len(ps), 0)
+    self.assertFalse(__file__ in ps)
+    for path in ps:
+      self.fail('New set is not empty.')
+
+    ps = path_set.PathSet([__file__])
+    self.assertEqual(len(ps), 1)
+    self.assertTrue(__file__ in ps)
+    self.assertEqual(ps.pop(), os.path.realpath(__file__))
+
+  def testAdd(self):
+    ps = path_set.PathSet()
+    ps.add(__file__)
+    self.assertEqual(len(ps), 1)
+    self.assertTrue(__file__ in ps)
+    self.assertEqual(ps.pop(), os.path.realpath(__file__))
+
+  def testDiscard(self):
+    ps = path_set.PathSet([__file__])
+    ps.discard(__file__)
+    self.assertEqual(len(ps), 0)
+    self.assertFalse(__file__ in ps)
+
+
+if __name__ == '__main__':
+  unittest.main()