Die build.py, Die: Part 1 (BranchUtility and SubversionFetcher)

chrome/common/extensions/docs/server2/branch_utility.py

+# 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.
+
+import logging
+import re
+import json
+
+from google.appengine.api import memcache
+from google.appengine.api import urlfetch
+
+DEFAULT_CACHE_TIME = 300
+
+class BranchUtility(object):
+  """Utility class for dealing with different doc branches.
+  """
+
+  def GetChannelNameFromURL(self, url):

[cduvall, 2012/05/22 00:07:40]

Can probably make this shorter, but works for now.

[Aaron Boodman, 2012/05/22 00:20:34]

It should only count if it's the first url component.

[cduvall, 2012/05/22 01:23:55]

Done.

+    if '/dev/' in url:
+      return 'dev'
+    if '/beta/' in url:
+      return 'beta'
+    if '/trunk/' in url:
+      return 'trunk'
+    return 'stable'
+
+  def GetBranchNumberForChannelName(self, name):
+    """Returns an empty string if the branch number cannot be found.
+    """
+    if name == 'trunk':

[cduvall, 2012/05/22 00:07:40]

omahaproxy.appspot.com has no branch number for trunk?

[Aaron Boodman, 2012/05/22 00:20:34]

There's no branch number for trunk because ... it's trunk :). Instead of
http://src.chromium.org/viewvc/chrome/branches/<branchnum>, you look in
http://src.chromium.org/viewvc/chrome/trunk/

[cduvall, 2012/05/22 01:23:55]

Done.

+      return 'trunk'

[Aaron Boodman, 2012/05/22 00:20:34]

In this case I think you should throw an exception. Caller should not call this
function with 'trunk'.

[cduvall, 2012/05/22 01:23:55]

I think its cleaner to handle trunk in here, since the branch number from omaha
can be trunk too. I'm just going to return trunk and have the caller handle it.

+    result = memcache.get(name)

[Aaron Boodman, 2012/05/22 00:20:34]

We are going to be using memcache for all kinds of things, so you're going to
want to use the namespace feature.

In fact, maybe we should just cache the url response from omahaproxy. So your
memcache namespace would be 'urls', and we will put other URLs that we fetch in
there in the future.

[cduvall, 2012/05/22 01:23:55]

Done.

+    if result is not None:
+      return result
+    logging.info('Branch number cache miss: ' + name)
+
+    result = urlfetch.fetch(
+        url="http://omahaproxy.appspot.com/json",
+        method=urlfetch.GET)
+
+    # Throw exception or something instead?
+    if result.status_code != 200:

[Aaron Boodman, 2012/05/22 00:20:34]

Yeah, exception makes sense.

[cduvall, 2012/05/22 01:23:55]

Done.

+      return ''
+
+    version_json = json.loads(result.content)
+    branch_number = None
+    for entry in version_json:
+      if entry['os'] != 'cf':

[cduvall, 2012/05/22 00:07:40]

Not sure what os: cf was, but assumed it was the default or something.

[Aaron Boodman, 2012/05/22 00:20:34]

It stands for 'chromeframe' - you can search what it is. It's a good idea to
skip this one for our purposes; it frequently has a different version.

[cduvall, 2012/05/22 01:23:55]

Done.

+        continue
+      for version in entry['versions']:

[Aaron Boodman, 2012/05/22 00:20:34]

Can you look at what version each platform has and use the most popular one? :).
I'm not sure what else to do in the case they disagree, that should be pretty
rare though.

[cduvall, 2012/05/22 01:23:55]

Done.

+        if version['channel'] == name:
+          branch_number = version['true_branch']
+
+    if branch_number == None:
+      return ''
+
+    if not memcache.add(name, branch_number, DEFAULT_CACHE_TIME):
+      logging.error('Memcache set failed.')
+    return branch_number