map/reduce result objects now only have 'key', 'value', and 'object' properties; MapReduceDocument.key_object now returns proper Document subclass; added finalize with Reddit ranking simulation; MapReduceDocuments now yielded;

mongoengine/document.py

@@ -124,34 +124,31 @@ class MapReduceDocument(object):
     :param key: Document/result key, often an instance of 
                 :class:`~pymongo.objectid.ObjectId`. If supplied as 
                 an ``ObjectId`` found in the given ``collection``, 
-                the object can be accessed via the ``key_object`` property.
-    :param value: The result(s) for this key. If given as a dictionary,
-                  each key in the dictionary will be available as
-                  an instance attribute.
+                the object can be accessed via the ``object`` property.
+    :param value: The result(s) for this key.
     
     .. versionadded:: 0.2.2
  
     """
     
-    def __init__(self, collection, key, value):
+    def __init__(self, document, collection, key, value):
+        self._document = document
         self._collection = collection
         self.key = key
         self.value = value
-        
-        if isinstance(value, dict):
-            # create attributes for each named result
-            for k, v in value.iteritems():
-                setattr(self, k, v)
     
     @property
     def object(self):
         """Lazy-load the object referenced by ``self.key``. If ``self.key``
         is not an ``ObjectId``, simply return ``self.key``.
         """
-        if not isinstance(self.key, pymongo.objectid.ObjectId):
-            return self.key
+        if not isinstance(self.key, (pymongo.objectid.ObjectId)):
+            try:
+                self.key = pymongo.objectid.ObjectId(self.key)
+            except:
+                return self.key
         if not hasattr(self, "_key_object"):
-            self._key_object = self._collection.find_one(self.key)
+            self._key_object = self._document.objects.with_id(self.key)
             return self._key_object
         return self._key_object
 

mongoengine/queryset.py

@@ -125,6 +125,7 @@ class QuerySet(object):
         self._query = {}
         self._where_clause = None
         self._ordering = []
+        self._limit = None
 
         # If inheritance is allowed, only return instances and instances of
         # subclasses of the class being used
@@ -380,7 +381,8 @@ class QuerySet(object):
     def __len__(self):
         return self.count()
 
-    def map_reduce(self, map_f, reduce_f, scope=None, keep_temp=False):
+    def map_reduce(self, map_f, reduce_f, finalize_f=None, limit=None,
+                   scope=None, keep_temp=False):
         """Perform a map/reduce query using the current query spec 
         and ordering. While ``map_reduce`` respects ``QuerySet`` chaining, 
         it must be the last call made, as it does not return a maleable 
@@ -402,6 +404,8 @@ class QuerySet(object):
         :param reduce_f: reduce function, as 
                          :class:`~pymongo.code.Code` or string
         :param scope: values to insert into map/reduce global scope. Optional.
+        :param limit: number of objects from current query to provide
+                      to map/reduce method
         :param keep_temp: keep temporary table (boolean, default ``True``)
         
         Returns a list of :class:`~mongoengine.document.MapReduceDocument`.
@@ -427,13 +431,16 @@ class QuerySet(object):
         
         mr_args = {'query': self._query, 'keeptemp': keep_temp}
 
+        if finalize_f:
+            if not isinstance(finalize_f, pymongo.code.Code):
+                finalize_f = pymongo.code.Code(finalize_f)
+            mr_args['finalize'] = finalize_f
+            
         if scope:
             mr_args['scope'] = scope
         if limit:
             mr_args['limit'] = limit
 
-        docs = []
-
         results = self._collection.map_reduce(map_f, reduce_f, **mr_args)
         results = results.find()
 
@@ -441,10 +448,8 @@ class QuerySet(object):
             results = results.sort(self._ordering)
         
         for doc in results:
-            mrd = MapReduceDocument(self._collection, doc['_id'], doc['value'])
-            docs.append(mrd)
-        
-        return docs
+            yield MapReduceDocument(self._document, self._collection, 
+                                    doc['_id'], doc['value'])
 
     def limit(self, n):
         """Limit the number of returned documents to `n`. This may also be
@@ -452,6 +457,7 @@ class QuerySet(object):
 
         :param n: the maximum number of objects to return
         """
+        self._limit = n
         self._cursor.limit(n)
         # Return self to allow chaining
         return self

tests/queryset.py

@@ -1,6 +1,9 @@
+# -*- coding: utf-8 -*-
+
+
 import unittest
 import pymongo
-from datetime import datetime
+from datetime import datetime, timedelta
 
 from mongoengine.queryset import (QuerySet, MultipleObjectsReturned, 
                                   DoesNotExist)
@@ -489,16 +492,121 @@ class QuerySetTest(unittest.TestCase):
         
         # ensure both artists are found
         results = Song.objects.map_reduce(map_f, reduce_f)
+        results = list(results)
         self.assertEqual(len(results), 2)
         
         # query for a count of Songs per artist, ordered by -count.
         # Patti Smith has 3 song credits, and should therefore be first.
         results = Song.objects.order_by("-value").map_reduce(map_f, reduce_f)
+        results = list(results)
         self.assertEqual(results[0].key, "Patti Smith")
         self.assertEqual(results[0].value, 3.0)
         
         Song.drop_collection()
         
+    def test_map_reduce_finalize(self):
+        """Ensure scope and finalize are working correctly by simulating
+        "hotness" ranking with Reddit algorithm.
+        """
+        from time import mktime
+        
+        class Link(Document):
+            title = StringField()
+            up_votes = IntField()
+            down_votes = IntField()
+            submitted = DateTimeField()
+
+        Link.drop_collection()
+
+        now = datetime.utcnow()
+
+        # Note: Test data taken from a custom Reddit homepage on
+        #       Fri, 12 Feb 2010 14:36:00 -0600.
+        
+        Link(title = "Google Buzz auto-followed a woman's abusive ex ...", 
+             up_votes = 1079,
+             down_votes = 553,
+             submitted = now-timedelta(hours=4)).save()
+        Link(title = "We did it! Barbie is a computer engineer.",
+             up_votes = 481,
+             down_votes = 124,
+             submitted = now-timedelta(hours=2)).save()
+        Link(title = "This Is A Mosquito Getting Killed By A Laser", 
+             up_votes = 1446,
+             down_votes = 530,
+             submitted=now-timedelta(hours=13)).save()
+        Link(title = "Arabic flashcards land physics student in jail.",
+             up_votes = 215, 
+             down_votes = 105,
+             submitted = now-timedelta(hours=6)).save()
+        Link(title = "The Burger Lab: Presenting, the Flood Burger",
+             up_votes = 48,
+             down_votes = 17,
+             submitted = now-timedelta(hours=5)).save()
+        Link(title="How to see polarization with the naked eye",
+             up_votes = 74,
+             down_votes = 13,
+             submitted = now-timedelta(hours=10)).save()
+
+        map_f = """
+            function() {
+                emit(this._id, {up_delta: this.up_votes - this.down_votes,
+                                reddit_epoch: new Date(2005, 12, 8, 7, 46, 43, 0).getTime(),
+                                sub_date: this.submitted.getTime()})
+            }
+        """
+
+        reduce_f = """
+            function(key, values) {
+                data = values[0];
+                                
+                x = data.up_delta;
+                
+                // calculate time diff between reddit epoch and submission
+                sec_since_epoch = data.sub_date - data.reddit_epoch;
+                sec_since_epoch /= 1000;
+                
+                // calculate 'Y'
+                if(x > 0) {
+                    y = 1;
+                } else if (x = 0) {
+                    y = 0;
+                } else {
+                    y = -1;
+                }
+                
+                // calculate 'Z', the maximal value
+                if(Math.abs(x) >= 1) {
+                    z = Math.abs(x);
+                } else {
+                    z = 1;
+                }
+                
+                return {x: x, y: y, z: z, t_s: sec_since_epoch};
+            }
+        """
+        
+        finalize_f = """
+            function(key, value) {
+            
+                // f(sec_since_epoch,y,z) = log10(z) + ((y*sec_since_epoch) / 45000)
+                z_10 = Math.log(value.z) / Math.log(10);
+                weight = z_10 + ((value.y * value.t_s) / 45000);
+                return weight;
+
+            }
+        """
+
+        # ensure both artists are found
+        results = Link.objects.order_by("-value")
+        results = results.map_reduce(map_f, reduce_f, finalize_f=finalize_f)
+        results = list(results)
+        
+        self.assertTrue(results[0].object.title.startswith("Google Buzz"))
+        self.assertTrue(results[-1].object.title.startswith("How to see"))
+
+        Link.drop_collection()
+        
     def test_item_frequencies(self):
         """Ensure that item frequencies are properly generated from lists.
         """