Add longest common prefix implementation, documentation and tests

Author: Leziak

presto-docs/src/main/sphinx/functions/string.rst

@@ -66,6 +66,10 @@ String Functions
     i.e. the minimum number of single-character edits (insertions,
     deletions or substitutions) needed to change ``string1`` into ``string2``.
 
+.. function:: longest_common_prefix(string1, string2) -> varchar
+
+    Returns the longest common prefix between ``string1`` and ``string2``
+
 .. function:: lower(string) -> varchar
 
     Converts ``string`` to lowercase.

presto-main-base/src/main/java/com/facebook/presto/operator/scalar/StringFunctions.java

@@ -756,6 +756,27 @@ public static Slice rightPad(@SqlType("varchar(x)") Slice text, @SqlType(Standar
         return pad(text, targetLength, padString, text.length());
     }
 
+    @Description("returns the longest common prefix shared by two strings")
+    @ScalarFunction("longest_common_prefix")
+    @LiteralParameters({"x", "y"})
+    @SqlType(StandardTypes.VARCHAR)
+    public static Slice longestCommonPrefix(@SqlType("varchar(x)") Slice left, @SqlType("varchar(y)") Slice right)
+    {
+        int i = 0;
+        int byteIndex = 0;
+        int[] leftCodePoints = castToCodePoints(left);
+        int[] rightCodePoints = castToCodePoints(right);
+        int leftLength = leftCodePoints.length;
+        int rightLength = rightCodePoints.length;
+
+        while (i < leftLength && i < rightLength && leftCodePoints[i] == rightCodePoints[i]) {
+            i++;
+            byteIndex += SliceUtf8.lengthOfCodePointSafe(left, byteIndex);
+        }
+
+        return Slices.wrappedBuffer(left.getBytes(0, byteIndex));
+    }
+
     @Description("computes Levenshtein distance between two strings")
     @ScalarFunction
     @LiteralParameters({"x", "y"})

presto-main-base/src/test/java/com/facebook/presto/operator/scalar/TestStringFunctions.java

@@ -164,6 +164,30 @@ public void testCharLength()
         assertFunction("LENGTH(CAST('\u4FE1\u5FF5,\u7231,\u5E0C\u671B' AS CHAR(20)))", BIGINT, 20L);
     }
 
+    @Test
+    public void testLongestCommonPrefix()
+    {
+        assertFunction("LONGEST_COMMON_PREFIX('', '')", VARCHAR, "");
+        assertFunction("LONGEST_COMMON_PREFIX('', 'hello')", VARCHAR, "");
+        assertFunction("LONGEST_COMMON_PREFIX('hello', '')", VARCHAR, "");
+        assertFunction("LONGEST_COMMON_PREFIX('hello', 'hello')", VARCHAR, "hello");
+        assertFunction("LONGEST_COMMON_PREFIX('hello world', 'hello')", VARCHAR, "hello");
+        assertFunction("LONGEST_COMMON_PREFIX('hello', 'hello world')", VARCHAR, "hello");
+        assertFunction("LONGEST_COMMON_PREFIX('hello world', 'hel wold')", VARCHAR, "hel");
+        assertFunction("LONGEST_COMMON_PREFIX('hel wold', 'hello world')", VARCHAR, "hel");
+
+        // Test for non-ASCII
+        assertFunction("LONGEST_COMMON_PREFIX('\u4FE1\u5FF5,\u7231,\u5E0C\u671B', '')", VARCHAR, "");
+        assertFunction("LONGEST_COMMON_PREFIX('', '\u4FE1\u5FF5,\u7231,\u5E0C\u671B')", VARCHAR, "");
+        assertFunction("LONGEST_COMMON_PREFIX('\u4FE1\u5FF5,\u7231,\u5E0C\u671B', '\u4FE1\u5FF5,\u7231,\u5E0C\u671B')", VARCHAR, "\u4FE1\u5FF5,\u7231,\u5E0C\u671B");
+        assertFunction("LONGEST_COMMON_PREFIX('\u4FE1\u5FF5,\u7221,\u5E0C\u671B', '\u4FE1\u5FF5,\u7231,\u5E0C\u671B')", VARCHAR, "\u4FE1\u5FF5,");
+        assertFunction("LONGEST_COMMON_PREFIX('hello na\u00EFve world', 'hello na\u00EFve')", VARCHAR, "hello na\u00EFve");
+
+        // Test for invalid-utf8 characters
+        assertInvalidFunction("LONGEST_COMMON_PREFIX('hello world', utf8(from_hex('81')))", "Invalid UTF-8 encoding in characters: �");
+        assertInvalidFunction("LONGEST_COMMON_PREFIX('hello world', utf8(from_hex('3281')))", "Invalid UTF-8 encoding in characters: 2�");
+    }
+
     @Test
     public void testLevenshteinDistance()
     {