Scale complexity based on overage

This mimics the algorithm from Classic analysis to determine the
remediation points for a complex function. The formula is

    baseCost + (costPer * (score - threshold))

With this, a complexity score at the threshold will be 1,000,000 points,
with 100,000 points for every additional branch within the function.

Author: gdiggs

bin/eslint.js

@@ -49,7 +49,7 @@ function buildIssueJson(message, path) {
         }
       }
     },
-    remediation_points: 50000
+    remediation_points: checks.remediationPoints(checkName, message)
   };
   return JSON.stringify(issue);
 }

lib/checks.js

@@ -27,6 +27,29 @@ var categories = function(checkName) {
   return [checkCategoryMapping[checkName] || "Style"]
 };
 
+var remediationPoints = function(checkName, message) {
+  if (checkName === "complexity") {
+    // (@base_cost + (overage * @cost_per))*1_000_000
+    // cost_per: 0.1,  base: 1
+    var costPer = 100000,
+        points = 1000000,
+        threshold = 10, // TODO: Get this from the eslintrc
+        overage, complexity;
+
+    complexity = message.message.match(/complexity of (\d+)/)[1];
+    overage = complexity - threshold;
+
+    if (overage > 0) {
+      points += (costPer * overage);
+    }
+
+    return points;
+  } else {
+    return 50000;
+  }
+};
+
 module.exports = {
-  categories: categories
+  categories: categories,
+  remediationPoints: remediationPoints
 };