Improve header deduplication logic and add edge case tests

Co-authored-by: Ivan Kyrylov <ikyrylov@griddynamics.com>

Author: Kvintochka

scautable/src/CSVUtils.scala

@@ -0,0 +1,22 @@
+package io.github.quafadas.scautable
+import scala.collection.immutable.ListMap
+import scala.annotation.tailrec
+
+object CSVUtils:
+  def uniquifyHeaders(headers: List[String]): List[String] =
+    if headers.toSet.sizeCompare(headers) == 0 then headers
+    else
+      val (newHeaders, _) = headers.foldLeft((List.empty[String], Map.empty[String, Int])):
+        case ((acc, counter), elem) =>
+          val (newElem, newCounter) = generateUniqueName(acc, counter, elem)
+          (newElem :: acc) -> newCounter
+      newHeaders.reverse
+
+  @tailrec
+  private def generateUniqueName(acc: List[String], counter: Map[String, Int], elem: String): (String, Map[String, Int]) =
+    if acc.contains(elem) then
+      val newCounter = counter.updatedWith(elem)(_.map(_ + 1).orElse(Some(1)))
+      val newElem = s"${elem}_${newCounter(elem)}"
+      generateUniqueName(acc, newCounter, newElem)
+    else elem -> counter
+end CSVUtils

scautable/src/csv.scala

@@ -16,10 +16,9 @@ import scala.collection.View.FlatMap
 import io.github.quafadas.scautable.ConsoleFormat.*
 import ColumnTyped.*
 
-import scala.annotation.tailrec
 import scala.math.Fractional.Implicits.*
 import scala.collection.View.Single
-import scala.collection.immutable.ListMap
+import io.github.quafadas.scautable.CSVUtils.*
 
 object CSV:
 
@@ -167,16 +166,4 @@ object CSV:
     end match
   end deduplicateHeadersCode
 
-  val startHeaderIndex = 1
-  def uniquifyHeaders(headers: List[String]): List[String] =
-    if headers.toSet.sizeCompare(headers) == 0 then headers
-    else
-      val updatedHeaders = headers.foldLeft(ListMap.empty[String, Int]):
-        case (acc, elem) =>
-          acc
-            .get(elem)
-            .map(value => (acc + (s"${elem}_${value}" -> startHeaderIndex)).updatedWith(elem)(_.map(_ + 1)))
-            .getOrElse(acc + (elem -> startHeaderIndex))
-      updatedHeaders.keys.toList
-
 end CSV

scautable/test/src-jvm/testCsv.scala

@@ -391,21 +391,12 @@ import CsvSchema.*"""
 
   }
 
-  // test("csv has duplicate headers") {
-  //   def csv: CsvIterator[("colA", "colA", "colA", "colB", "colC", "colA")] = CSV.resource("dups.csv")
-
-  //   // If the next two lines compile, this is a pretty good indicator that we've deduplicated the headers
-  //   def dedupCsv: CsvIterator[("colA", "colA_1", "colA_2", "colB", "colC", "colA_5")] = CSV.deduplicateHeaders(csv)
-  //   val testVal = dedupCsv.drop(1).next().colA_5
-  //   assert(testVal == "5")
-  // }
-
-    test("csv has duplicate headers") {
+  test("csv has duplicate headers") {
     def csv: CsvIterator[("colA", "colA", "colA", "colB", "colC", "colA")] = CSV.resource("dups.csv")
 
     // If the next two lines compile, this is a pretty good indicator that we've deduplicated the headers
-    def dedupCsv: CsvIterator[("colA", "colA_2", "colA_3", "colB", "colC", "colA_4")] = CSV.deduplicateHeaders(csv)
-    val testVal = dedupCsv.drop(1).next().colA_4
+    def dedupCsv: CsvIterator[("colA", "colA_1", "colA_2", "colB", "colC", "colA_3")] = CSV.deduplicateHeaders(csv)
+    val testVal = dedupCsv.drop(1).next().colA_3
     assert(testVal == "5")
   }
 

scautable/test/src/CSVUtilsSuite.scala

@@ -0,0 +1,15 @@
+package io.github.quafadas.scautable
+
+class CSVUtilsSuite extends munit.FunSuite:
+  test("uniquifyHeaders should suffix duplicates with index") {
+    val headers = List("id", "name", "name", "email", "id", "name")
+    val result = CSVUtils.uniquifyHeaders(headers)
+    assertEquals(result, List("id", "name", "name_1", "email", "id_1", "name_2"))
+  }
+
+  test("uniquifyHeaders should correctly suffix duplicates even when original headers contain numbered suffixes") {
+    val headers = List("test", "test_1", "test", "test_1", "test_2")
+    val result = CSVUtils.uniquifyHeaders(headers)
+    assertEquals(result, List("test", "test_1", "test_1_1", "test_1_2", "test_2"))
+  }
+end CSVUtilsSuite