Avoid recursively checking JSON inventory for Unicode

by moving to en-bloc unicode conversion to act on scripts stdout

Both python-json and simplejson always return unicode strings when using
their loads() method on unicode strings. This is true at least since
2009. This makes checking each substring unnecessary, because we do not
need to recursively check the strings contained in the inventory dict
later one-by-one

This commit makes parsing of large dynamic inventory at least 2 seconds
faster.

cf: https://github.com/towolf/ansible-large-inventory-testcase

lib/ansible/inventory/script.py

@@ -31,7 +31,7 @@ from ansible.errors import AnsibleError
 from ansible.inventory.host import Host
 from ansible.inventory.group import Group
 from ansible.module_utils.basic import json_dict_bytes_to_unicode
-from ansible.utils.unicode import to_str
+from ansible.utils.unicode import to_str, to_unicode
 
 
 class InventoryScript:
@@ -58,7 +58,13 @@ class InventoryScript:
         if sp.returncode != 0:
             raise AnsibleError("Inventory script (%s) had an execution error: %s " % (filename,stderr))
 
-        self.data = stdout
+        # make sure script output is unicode so that json loader will output
+        # unicode strings itself
+        try:
+            self.data = to_unicode(stdout, errors="strict")
+        except Exception as e:
+            raise AnsibleError("inventory data from {0} contained characters that cannot be interpreted as UTF-8: {1}".format(to_str(self.filename), to_str(e)))
+
         # see comment about _meta below
         self.host_vars_from_top = None
         self._parse(stderr)
@@ -78,8 +84,6 @@ class InventoryScript:
             sys.stderr.write(err + "\n")
             raise AnsibleError("failed to parse executable inventory script results from {0}: data needs to be formatted as a json dict".format(to_str(self.filename)))
 
-        self.raw  = json_dict_bytes_to_unicode(self.raw)
-
         group = None
         for (group_name, data) in self.raw.items():