From 485a181d0f651ac50983ab8539817ed25e786fd7 Mon Sep 17 00:00:00 2001
From: Oleg Ilyenko <oleg.ilyenko@gmail.com>
Date: Sat, 30 Apr 2016 16:20:34 +0200
Subject: [PATCH] `suggestionList` implementation (#126)

---
 src/main/scala/sangria/util/StringUtil.scala  | 48 +++++++++++++++++++
 .../scala/sangria/util/StringUtilSpec.scala   | 14 ++++++
 2 files changed, 62 insertions(+)

diff --git a/src/main/scala/sangria/util/StringUtil.scala b/src/main/scala/sangria/util/StringUtil.scala
index d697dc29..80c46330 100644
--- a/src/main/scala/sangria/util/StringUtil.scala
+++ b/src/main/scala/sangria/util/StringUtil.scala
@@ -1,5 +1,7 @@
 package sangria.util
 
+import scala.collection.mutable.ListBuffer
+
 object StringUtil {
   private val camelToUpper = "_*([A-Z][a-z\\d]+)".r
 
@@ -20,4 +22,50 @@ object StringUtil {
       if (start.nonEmpty) s"${start mkString ", "} or $last"
       else last
     }
+
+  /**
+    * Given an invalid input string and a list of valid options, returns a filtered
+    * list of valid options sorted based on their similarity with the input.
+    */
+  def suggestionList(input: String, options: Seq[String]): Seq[String] = {
+    val inputThreshold = input.length / 2
+
+    options
+      .map (opt ⇒ opt → lexicalDistance(input, opt))
+      .filter (opt ⇒ opt._2 <= math.max(math.max(inputThreshold, opt._1.length / 2), 1))
+      .sortBy (_._2)
+      .map (_._1)
+  }
+
+  /**
+    * Computes the lexical distance between strings A and B.
+    *
+    * The "distance" between two strings is given by counting the minimum number
+    * of edits needed to transform string A into string B. An edit can be an
+    * insertion, deletion, or substitution of a single character, or a swap of two
+    * adjacent characters.
+    *
+    * This distance can be useful for detecting typos in input or sorting
+    *
+    * @return distance in number of edits
+    */
+  def lexicalDistance(a: String, b: String): Int = {
+    val d = for (i ← 0 to a.length) yield ListBuffer.fill(b.length + 1)(i)
+
+    for (j ← 1 to b.length) {
+      d(0)(j) = j
+    }
+
+    for (i ← 1 to a.length; j ← 1 to b.length) {
+      val cost = if (a(i - 1) == b(j - 1)) 0 else 1
+
+      d(i)(j) = math.min(math.min(d(i - 1)(j) + 1, d(i)(j - 1) + 1), d(i - 1)(j - 1) + cost)
+
+      if (i > 1 && j > 1 && a(i - 1) == b(j - 2) && a(i - 2) == b(j - 1)) {
+        d(i)(j) = math.min(d(i)(j), d(i - 2)(j - 2) + cost)
+      }
+    }
+
+    d(a.length)(b.length)
+  }
 }
diff --git a/src/test/scala/sangria/util/StringUtilSpec.scala b/src/test/scala/sangria/util/StringUtilSpec.scala
index 40909a23..0975a51c 100644
--- a/src/test/scala/sangria/util/StringUtilSpec.scala
+++ b/src/test/scala/sangria/util/StringUtilSpec.scala
@@ -34,4 +34,18 @@ class StringUtilSpec extends WordSpec with Matchers {
       quotedOrList(Seq("A", "B", "C", "D", "E", "F")) should be ("\"A\", \"B\", \"C\", \"D\" or \"E\"")
     }
   }
+
+  "suggestionList" should {
+    "Returns results when input is empty" in {
+      suggestionList("", Seq("a")) should be (Seq("a"))
+    }
+
+    "Returns empty array when there are no options" in {
+      suggestionList("input", Seq.empty) should be (Seq.empty)
+    }
+    
+    "Returns options sorted based on similarity" in {
+      suggestionList("abc", Seq("a", "ab", "abc")) should be (Seq("abc", "ab"))
+    }
+  }
 }