Refactored data source to cloud-init interface

This patch heavily refactors the data source so it follows the
cloud-init datasource interface standard. The hostname, network
configuration, and instance ID are all set via the "guestinfo.metadata"
property. Please see README.md for more information.
diff --git a/DataSourceVmxGuestinfo.py b/DataSourceVmxGuestinfo.py
index c5d7998..60b68dd 100644
--- a/DataSourceVmxGuestinfo.py
+++ b/DataSourceVmxGuestinfo.py
@@ -2,24 +2,54 @@
 #
 # Copyright (C) 2017 VMware Inc.
 #
-# Author: Anish Swaminathan <anishs@vmware.com>
+# Authors: Anish Swaminathan <anishs@vmware.com>
+#          Andrew Kutz <akutz@vmware.com>
 #
 import os
 import base64
+import zlib
+import json
 
 from cloudinit import log as logging
 from cloudinit import sources
 from cloudinit import util
+from cloudinit import safeyaml
 
 from distutils.spawn import find_executable
 
 LOG = logging.getLogger(__name__)
 
+# Used with CentOS 7 which installs cloud-init 0.7.9. The URL for
+# the DataSource class is http://bit.ly/cloudinit-datasource-0-7-9.
+#
+# Setting the hostname:
+#     The hostname is set by way of the metadata key "local-hostname".
+#
+# Setting the instance ID:
+#     The instance ID may be set by way of the metadata key "instance-id".
+#     However, if this value is absent then then the instance ID is
+#     read from the file /sys/class/dmi/id/product_uuid.
+#
+# Configuring the network:
+#     The network is configured by setting the metadata key "network"
+#     with a value consistent with Network Config Versions 1 or 2,
+#     depending on the Linux distro's version of cloud-init:
+#
+#         Network Config Version 1 - http://bit.ly/cloudinit-net-conf-v1
+#         Network Config Version 2 - http://bit.ly/cloudinit-net-conf-v2
+#
+#     For example, CentOS 7's official cloud-init package is version
+#     0.7.9 and does not support Network Config Version 2. However,
+#     this datasource still supports supplying Network Config Version 2
+#     data as long as the Linux distro's cloud-init package is new
+#     enough to parse the data.
+#
+#     The metadata key "network.encoding" may be used to indicate the
+#     format of the metadata key "network". Valid encodings are base64
+#     and gzip+base64.
 class DataSourceVmxGuestinfo(sources.DataSource):
     def __init__(self, sys_cfg, distro, paths, ud_proc=None):
         sources.DataSource.__init__(self, sys_cfg, distro, paths, ud_proc)
-        self.metadata = {}
-        self.userdata_raw = ''
         self.vmtoolsd = find_executable("vmtoolsd")
         if not self.vmtoolsd:
             LOG.error("Failed to find vmtoolsd")
@@ -28,118 +58,89 @@
         if not self.vmtoolsd:
             LOG.error("vmtoolsd is required to fetch guestinfo value")
             return False
-        hostname = self._get_guestinfo_value('hostname')
-        if hostname:
-            self.distro.set_hostname(hostname)
-        ud = self._get_guestinfo_value('userdata')
-        if ud:
-            LOG.debug("Decoding base64 format guestinfo.userdata")
-            self.userdata_raw = base64.b64decode(ud)
-        found = True
-        dev_index = 0
-        network_settings = ''
-        while found:
-            key_begin = 'interface.' + str(dev_index)
-            key_iname = key_begin + '.name'
-            interface_name = self._get_guestinfo_value(key_iname)
-            if interface_name:
-                network_settings += 'auto ' + interface_name + '\n'
-                network_settings += 'iface ' + interface_name
-                key_proto = key_begin + '.dhcp'
-                dhcp_enabled = self._get_guestinfo_value(key_proto)
-                key_address = key_begin + '.address'
-                address = self._get_guestinfo_value(key_address)
-                bootproto = 'dhcp'
-                if dhcp_enabled:
-                    if dhcp_enabled == 'yes':
-                        network_settings += ' dhcp\n'
-                    elif dhcp_enabled == 'no':
-                        network_settings += ' static\n'
-                        bootproto = 'static'
-                    else:
-                        LOG.warning("Invalid value for yes/no parameter for %s, setting to dhcp", key_proto)
-                elif address:
-                    bootproto = 'static'
-                    dhcp_enabled == 'no'
-                    network_settings += ' static\n'
-                else:
-                    dhcp_enabled == 'yes'
-                    network_settings += ' dhcp\n'
-                    LOG.debug("Setting network bootproto to dhcp by default")
-                key_mac = key_begin + '.mac'
-                mac = self._get_guestinfo_value(key_mac)
-                if address:
-                    network_settings += 'address ' + address + '\n'
-                if mac:
-                    network_settings += 'hwaddress ' + mac + '\n'
-                key_netmask = key_begin + '.netmask'
-                netmask = self._get_guestinfo_value(key_netmask)
-                if netmask:
-                    network_settings += 'netmask ' + netmask + '\n'
-                key_dnsserver = 'dns.servers'
-                dnsserver = self._get_guestinfo_value(key_dnsserver)
-                if dnsserver:
-                    network_settings += 'dns-nameservers '
-                    dnsserver = dnsserver.split(',')
-                    for d in dnsserver:
-                        network_settings += d + ' '
-                    network_settings += '\n'
-                key_dnsdomain = 'dns.domains'
-                dnsdomain = self._get_guestinfo_value(key_dnsdomain)
-                if dnsdomain:
-                    network_settings += 'dns-search '
-                    dnsdomain = dnsdomain.split(',')
-                    for d in dnsdomain:
-                        network_settings += d + ' '
-                    network_settings += '\n'
-                route_index = 0
-                default_destination_set = False
-                while True:
-                    key_route = key_begin + '.route.' + str(route_index)
-                    route = self._get_guestinfo_value(key_route)
-                    if route:
-                        network_settings += "routes.%s " % (route_index)
-                        route = route.split(',')
-                        if len(route) > 2:
-                            LOG.debug("Route information for %s route in %s device incorrect - ", 
-                                                "expected 2 values", route_index, dev_index)
-                            continue
-                        elif len(route) == 2:
-                            network_settings += route[0] + ' ' + route[1] + '\n'# Gateway Destination
-                        else: #length = 1
-                            if not default_destination_set:
-                                network_settings += route[0] + ' 0.0.0.0/0' + '\n'
-                                default_destination_set = True
-                            else:
-                                LOG.debug("Default destination set previously, not setting route %s", route_index) 
-                    else:
-                        break
-                    route_index += 1
-            else:
-                found = False
-            dev_index += 1
-        self.distro.apply_network(network_settings, False)
+
+        # Get the JSON metadata. Can be plain-text, base64, or gzip+base64.
+        metadata = self._get_encoded_guestinfo_data('metadata')
+        if metadata:
+            self.metadata = json.loads(metadata)
+
+        # Get the YAML userdata. Can be plain-text, base64, or gzip+base64.
+        self.userdata_raw = self._get_encoded_guestinfo_data('userdata')
+
+        # Get the YAML vendordata. Can be plain-text, base64, or gzip+base64.
+        self.vendordata_raw = self._get_encoded_guestinfo_data('vendordata')
+
         return True
 
-    def _get_guestinfo_value(self, key):
-        LOG.debug("Getting guestinfo value for key %s", key)
-        value = ''
-        try:
-            (value, _err) = util.subp([self.vmtoolsd, "--cmd", "info-get guestinfo." + key])
-            if _err:
-                LOG.error("Failed to get guestinfo value for key %s", key)
-        except util.ProcessExecutionError as error:
-            util.logexc(LOG,"Failed to get guestinfo value for key %s: %s", key, error)
-        except Exception:
-            util.logexc(LOG,"Unexpected error while trying to get guestinfo value for key %s", key)
-        return value.rstrip()
+    @property
+    def network_config(self):
+        # Pull the network configuration out of the metadata.
+        if self.metadata and 'network' in self.metadata:
+            data = self._get_encoded_metadata('network')
+            if data:
+                # Load the YAML-formatted network data into an object
+                # and return it.
+                net_config = safeyaml.load(data)
+                LOG.debug("Loaded network config: %s", net_config)
+                return net_config
+        return None
 
     def get_instance_id(self):
+        # Pull the instance ID out of the metadata if present. Otherwise
+        # read the file /sys/class/dmi/id/product_uuid for the instance ID.
+        if self.metadata and 'instance-id' in self.metadata:
+            return self.metadata['instance-id']
         with open('/sys/class/dmi/id/product_uuid', 'r') as id_file:
             return str(id_file.read()).rstrip()
 
+    def _get_encoded_guestinfo_data(self, key):
+        data = self._get_guestinfo_value(key)
+        if not data:
+            return None
+        enc_type = self._get_guestinfo_value(key + '.encoding')
+        return self._get_encoded_data('guestinfo.' + key, enc_type, data)
+
+    def _get_encoded_metadata(self, key):
+        if not self.metadata or not key in self.metadata:
+            return None
+        data = self.metadata[key]
+        enc_type = self.metadata.get(key + '.encoding')
+        return self._get_encoded_data('metadata.' + key, enc_type, data)
+
+    def _get_encoded_data(self, key, enc_type, data):
+        LOG.debug("Getting encoded data for key=%s, enc=%s", key, enc_type)
+        if enc_type == "gzip+base64" or enc_type == "gz+b64":
+            LOG.debug("Decoding %s format %s", enc_type, key)
+            return zlib.decompress(base64.b64decode(data), zlib.MAX_WBITS | 16)
+        elif enc_type == "base64" or enc_type == "b64":
+            LOG.debug("Decoding %s format %s", enc_type, key)
+            return base64.b64decode(data)
+        else:
+            LOG.debug("Plain-text data %s", key)
+            return data
+
+    def _get_guestinfo_value(self, key):
+        NOVAL = "No value found"
+        LOG.debug("Getting guestinfo value for key %s", key)
+        try:
+            (stdout, stderr) = util.subp([self.vmtoolsd, "--cmd", "info-get guestinfo." + key])
+            if stderr == NOVAL:
+                LOG.debug("No value found for key %s", key)
+            elif not stdout:
+                LOG.error("Failed to get guestinfo value for key %s", key)
+            else:
+                return stdout.rstrip()
+        except util.ProcessExecutionError as error:
+            if error.stderr == NOVAL:
+                LOG.debug("No value found for key %s", key)
+            else:
+                util.logexc(LOG,"Failed to get guestinfo value for key %s: %s", key, error)
+        except Exception:
+            util.logexc(LOG,"Unexpected error while trying to get guestinfo value for key %s", key)
+        return None
+
 def get_datasource_list(depends):
     """
     Return a list of data sources that match this set of dependencies
     """
-    return [DataSourceVmxGuestinfo]
\ No newline at end of file
+    return [DataSourceVmxGuestinfo]