Merge pull request #45 from todvora/master

Fixed maxDepth inconsistency, fixed total counts and percents

test/src/test/java/com/github/variety/test/LimitResultsAnalysisTest.java

@@ -32,7 +32,6 @@ public class LimitResultsAnalysisTest {
         analysis.verifyResult("_id", 5, 100, "ObjectId");
         analysis.verifyResult("name", 5, 100, "String");
 
-        // TODO: there is only one document with 'someBinData'. Why variety returns 5/100% instead of 1/20% ?
-        // FIXME: analysis.verifyResult("someBinData", 1, 20, "BinData-old");
+        analysis.verifyResult("someBinData", 1, 20, "BinData-old");
     }
 }

test/src/test/java/com/github/variety/test/MaxDepthAnalysisTest.java

@@ -11,7 +11,7 @@ import org.junit.Test;
 
 public class MaxDepthAnalysisTest {
 
-    private static final double EXPECTED_PERCENTS = 100; //TODO: why documentation mentions be 16.66666666666666, when there is only one document at all ?
+    private static final double EXPECTED_PERCENTS = 100;
     private Variety variety;
 
     @Before
@@ -47,8 +47,7 @@ public class MaxDepthAnalysisTest {
     public void testLimitedDepthAnalysis() throws Exception {
         final VarietyAnalysis analysis = variety.withMaxDepth(3).runAnalysis();
 
-        // TODO: depth 3 means 'someNestedObject.a.b' or 'someNestedObject.a.b.c'? Documentation describes the first variant, variety counts also second.
-        // FIXME: Assert.assertEquals("Variety results have not correct count of entries", 5, analysis.getResultsCollection().count()); // 5 results, including '_id' and 'name'
+        Assert.assertEquals("Variety results have not correct count of entries", 5, analysis.getResultsCollection().count()); // 5 results, including '_id' and 'name'
 
         analysis.verifyResult("_id", 1, EXPECTED_PERCENTS, "ObjectId");
         analysis.verifyResult("name", 1, EXPECTED_PERCENTS, "String");

test/src/test/java/com/github/variety/test/QueryLimitedAnalysisTest.java

@@ -27,9 +27,9 @@ public class QueryLimitedAnalysisTest {
         final VarietyAnalysis analysis = variety.withQuery("{someBinData:{$exists: true}}").runAnalysis();
         Assert.assertEquals(3, analysis.getResultsCollection().count());
 
-        // TODO: are those percentContaining numbers correct? Should percents be limited to all data or query data?
-        analysis.verifyResult("_id", 1, 20, "ObjectId");
-        analysis.verifyResult("name", 1, 20, "String");
+        // we analyzed only the keys of objects defined by query. But total counts and percents are computed from the complete collection
+        analysis.verifyResult("_id", 5, 100, "ObjectId");
+        analysis.verifyResult("name", 5, 100, "String");
         analysis.verifyResult("someBinData", 1, 20, "BinData-old");
 
     }

test/src/test/java/com/github/variety/test/SortedAnalysisTest.java

@@ -49,14 +49,11 @@ public class SortedAnalysisTest {
 
         Assert.assertEquals(5, analysis.getResultsCollection().count());
 
-        // TODO: are those percentContaining numbers correct? Should percents be limited to all data or query data?
-        // Why total counts are always 5, when 'someWeirdLegacyKey' has only one object?
-        // Keys and types are correct, total count and percents seems not right.
         analysis.verifyResult("_id", 5, 100, "ObjectId");
         analysis.verifyResult("name", 5, 100, "String");
-        analysis.verifyResult("bio", 5, 100, "String");
-        analysis.verifyResult("pets", 5, 100, "Array");
-        analysis.verifyResult("someWeirdLegacyKey", 5, 100, "String");
+        analysis.verifyResult("bio", 3, 60, "String");
+        analysis.verifyResult("pets", 2, 40, "Array");
+        analysis.verifyResult("someWeirdLegacyKey", 1, 20, "String");
 
     }
 }

test/src/test/java/com/github/variety/test/UnnamedObjectsAnalysisTest.java

@@ -5,6 +5,7 @@ import com.github.variety.VarietyAnalysis;
 import com.mongodb.DBObject;
 import com.mongodb.util.JSON;
 import org.junit.After;
+import org.junit.Assert;
 import org.junit.Before;
 import org.junit.Test;
 
@@ -33,20 +34,15 @@ public class UnnamedObjectsAnalysisTest {
     public void testUnnamedObjects() throws Exception {
         final VarietyAnalysis analysis = variety.runAnalysis();
 
+        Assert.assertEquals(6, analysis.getResultsCollection().count());
+
         analysis.verifyResult("_id", 2, 100, "ObjectId");
         analysis.verifyResult("title", 2, 100, "String");
         analysis.verifyResult("comments", 2, 100, "Array");
 
-
-        // TODO: current version of variety is not able to handle unnamed inside objects. Earlier they were marked with XX. key prefix.
-        // Now the unnamed object are skipped and not analysed at all. Example of earlier version results can be seen
-        // in issue https://github.com/variety/variety/issues/29
-
-        // There should be 6 different keys: _id, title, comments and three from anonymous objects: comments.XX.author, comments.XX.body, comments.XX.visible
-        // FIXME: Assert.assertEquals(6, analysis.getResultsCollection().count());
-
-        // FIXME: analysis.verifyResult("comments.XX.author", 2, 100, "String");
-        // FIXME: analysis.verifyResult("comments.XX.body", 2, 100, "String");
-        // FIXME: analysis.verifyResult("comments.XX.visible", 1, 50, "Boolean");
+        // unnamed objects are prefixed with .XX key
+        analysis.verifyResult("comments.XX.author", 2, 100, "String");
+        analysis.verifyResult("comments.XX.body", 2, 100, "String");
+        analysis.verifyResult("comments.XX.visible", 1, 50, "Boolean");
     }
 }

variety.js

@@ -128,7 +128,7 @@ function serializeDoc(doc, maxDepth){
 			//if(typeof value != 'object') 
 			result[parentKey+key] = value;
 			//it's an object, recurse...only if we haven't reached max depth
-			if(isHash(value) && (maxDepth > 0)) {
+			if(isHash(value) && (maxDepth > 1)) {
 				serialize(value, parentKey+key+'.',maxDepth-1);
 			}
 		}
@@ -209,7 +209,7 @@ resultsDB[resultsCollectionName].find({}).forEach(function(key) {
   if(limit < numDocuments) {
     var existsQuery = {};
     existsQuery[keyName] = {$exists: true};
-    key.totalOccurrences = db[collection].find(query).count(existsQuery);
+    key.totalOccurrences = db[collection].count(existsQuery);
   }  
   key.percentContaining = (key.totalOccurrences / numDocuments) * 100.0;
   resultsDB[resultsCollectionName].save(key);