Fix GitCheckout.prepare to be sane and use git-cache

checkout.py

 # coding=utf8
 # 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.
 """Manages a project checkout.
 
 Includes support for svn, git-svn and git.
 """
 
 import ConfigParser
@@ skipping 553 matching lines @@
     # There is no reason to not hardcode origin.
     self.remote = 'origin'
     # There is no reason to not hardcode master.
     self.master_branch = 'master'
 
   def prepare(self, revision):
     """Resets the git repository in a clean state.
 
     Checks it out if not present and deletes the working branch.
     """
+    assert self.git_url
     assert self.remote_branch
-    assert self.git_url
+
+    self._check_call_git(
+        ['cache', 'populate', self.git_url], timeout=FETCH_TIMEOUT)

[rmistry, 2014/02/25 13:23:14]

Disclaimer: I have never used git cache before.

Do you need to specify -c dir_name here?
When I tried
$ git cache populate https://skia.googlesource.com/buildbot
It gave me:

Usage: git cache populate [options] [url of repo to add to or update in cache]
git cache: error: No cache directory specified on command line or in
cache.cachepath.

Same for the cache exists call below.


Also, when I tried 'git cache' on my desktop (git version
1.9.0.rc1.175.g0b1dcb5), git cache was unrecognized.
But on another machine with git version 1.7.9.5 it worked (but needed -c
specified). If it only exists in particular versions then that might be a
problem.

+    cache_path = self._check_output_git(
+        ['cache', 'exists', self.git_url]).strip()
 
     if not os.path.isdir(self.project_path):
       # Clone the repo if the directory is not present.
-      logging.info(
-          'Checking out %s in %s', self.project_name, self.project_path)
       self._check_call_git(
-          ['clone', self.git_url, '-b', self.remote_branch, self.project_path],
+          ['clone', '--shared', cache_path, self.project_path],
           cwd=None, timeout=FETCH_TIMEOUT)
-    else:
-      # Throw away all uncommitted changes in the existing checkout.
-      self._check_call_git(['checkout', self.remote_branch])
-      self._check_call_git(
-          ['reset', '--hard', '--quiet',
-           '%s/%s' % (self.remote, self.remote_branch)])
 
-    if revision:
-      try:
-        # Look if the commit hash already exist. If so, we can skip a
-        # 'git fetch' call.
-        revision = self._check_output_git(['rev-parse', revision])
-      except subprocess.CalledProcessError:
-        self._check_call_git(
-            ['fetch', self.remote, self.remote_branch, '--quiet'])
-        revision = self._check_output_git(['rev-parse', revision])
-      self._check_call_git(['checkout', '--force', '--quiet', revision])
-    else:
-      branches, active = self._branches()
-      if active != self.master_branch:
-        self._check_call_git(
-            ['checkout', '--force', '--quiet', self.master_branch])
-      self._sync_remote_branch()
+    if not revision:
+      revision = self.remote_branch
 
-      if self.working_branch in branches:
-        self._call_git(['branch', '-D', self.working_branch])
+    self._check_call_git(['fetch', self.remote, revision])
+
+    self._check_call_git(['checkout', '--force', '--quiet', revision])
+    self._call_git(['clean', '-fdx'])
+
+    branches, _ = self._branches()
+    if self.working_branch in branches:
+      self._call_git(['branch', '-D', self.working_branch])
+
     return self._get_head_commit_hash()
 
-  def _sync_remote_branch(self):
-    """Syncs the remote branch."""
-    # We do a 'git pull origin master:refs/remotes/origin/master' instead of
-    # 'git pull origin master' because from the manpage for git-pull: 
-    #   A parameter <ref> without a colon is equivalent to <ref>: when
-    #   pulling/fetching, so it merges <ref> into the current branch without
-    #   storing the remote branch anywhere locally.
-    remote_tracked_path = 'refs/remotes/%s/%s' % (
-        self.remote, self.remote_branch)
-    self._check_call_git(
-        ['pull', self.remote,
-         '%s:%s' % (self.remote_branch, remote_tracked_path),
-         '--quiet'])
-
   def _get_head_commit_hash(self):
     """Gets the current revision (in unicode) from the local branch."""
     return unicode(self._check_output_git(['rev-parse', 'HEAD']).strip())
 
   def apply_patch(self, patches, post_processors=None, verbose=False):
     """Applies a patch on 'working_branch' and switches to it.
 
     Also commits the changes on the local branch.
 
     Ignores svn properties and raise an exception on unexpected ones.
@@ skipping 81 matching lines @@
     assert sorted(patches.filenames) == sorted(found_files), (
         sorted(patches.filenames), sorted(found_files))
 
   def commit(self, commit_message, user):
     """Commits, updates the commit message and pushes."""
     assert self.commit_user
     assert isinstance(commit_message, unicode)
     current_branch = self._check_output_git(
         ['rev-parse', '--abbrev-ref', 'HEAD']).strip()
     assert current_branch == self.working_branch
- 
+
     commit_cmd = ['commit', '--amend', '-m', commit_message]
     if user and user != self.commit_user:
       # We do not have the first or last name of the user, grab the username
       # from the email and call it the original author's name.
       # TODO(rmistry): Do not need the below if user is already in
       #                "Name <email>" format.
       name = user.split('@')[0]
       commit_cmd.extend(['--author', '%s <%s>' % (name, user)])
     self._check_call_git(commit_cmd)
 
     # Push to the remote repository.
     self._check_call_git(
         ['push', 'origin', '%s:%s' % (self.working_branch, self.remote_branch),
          '--force', '--quiet'])
     # Get the revision after the push.
     revision = self._get_head_commit_hash()
     # Switch back to the remote_branch and sync it.
     self._check_call_git(['checkout', self.remote_branch])
     self._sync_remote_branch()

[rmistry, 2014/02/24 23:17:04]

You removed this method but the call still exists, this tells me you have not
tested this end-to-end yet, I highly recommend testing it against a dummy
repository first, I can provide one to you if you do not have one to test
against.

[agable, 2014/02/24 23:24:20]

Correct, this patchset does not represent any end-to-end testing including
committing yet. It is so far just a rewrite of the prepare() method.

[rmistry, 2014/02/25 13:23:14]

The method call was here to make sure that the branch is synced appropriately
before we start again, should we try to maintain the same behavior?

     # Delete the working branch since we are done with it.
     self._check_call_git(['branch', '-D', self.working_branch])
 
     return revision
 
   def _check_call_git(self, args, **kwargs):
     kwargs.setdefault('cwd', self.project_path)
     kwargs.setdefault('stdout', self.VOID)
     kwargs.setdefault('timeout', GLOBAL_TIMEOUT)
     return subprocess2.check_call_out(['git'] + args, **kwargs)
@@ skipping 68 matching lines @@
   def revisions(self, rev1, rev2):
     return self.checkout.revisions(rev1, rev2)
 
   @property
   def project_name(self):
     return self.checkout.project_name
 
   @property
   def project_path(self):
     return self.checkout.project_path