From ca79a488929f7777b5c2262c12f85bfa0272ca5d Mon Sep 17 00:00:00 2001
From: yangjie01 <yangjie01@baidu.com>
Date: Fri, 27 Nov 2020 11:16:25 +0800
Subject: [PATCH] refactor test

---
 .../unescaped-quotes-unescaped-delimiter.csv  |  3 --
 .../execution/datasources/csv/CSVSuite.scala  | 33 ++++++++++++-------
 2 files changed, 21 insertions(+), 15 deletions(-)
 delete mode 100644 sql/core/src/test/resources/test-data/unescaped-quotes-unescaped-delimiter.csv

diff --git a/sql/core/src/test/resources/test-data/unescaped-quotes-unescaped-delimiter.csv b/sql/core/src/test/resources/test-data/unescaped-quotes-unescaped-delimiter.csv
deleted file mode 100644
index a1d91b6d27a79..0000000000000
--- a/sql/core/src/test/resources/test-data/unescaped-quotes-unescaped-delimiter.csv
+++ /dev/null
@@ -1,3 +0,0 @@
-c1,c2
-"a,""b,c","xyz"
-"a,b,c","x""yz"
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVSuite.scala
index 97c0fe11c17ad..30f0e45d04eab 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVSuite.scala
@@ -75,8 +75,6 @@ abstract class CSVSuite
   private val valueMalformedFile = "test-data/value-malformed.csv"
   private val badAfterGoodFile = "test-data/bad_after_good.csv"
   private val malformedRowFile = "test-data/malformedRow.csv"
-  private val unescapedQuotesAndUnescapedDelimiterFile =
-    "test-data/unescaped-quotes-unescaped-delimiter.csv"
 
   /** Verifies data and schema. */
   private def verifyCars(
@@ -2432,16 +2430,27 @@ abstract class CSVSuite
   }
 
   test("SPARK-33566: configure UnescapedQuoteHandling to parse " +
-    "unescapedQuotesAndUnescapedDelimiterFile correctly") {
-    // Without configure UnescapedQuoteHandling to STOP_AT_CLOSING_QUOTE,
-    // the result will be Row(""""a,""b""", """c""""), Row("""a,b,c""", """"x""yz"""")
-    val result = spark.read
-      .option("inferSchema", "true")
-      .option("header", "true")
-      .option("unescapedQuoteHandling", "STOP_AT_CLOSING_QUOTE")
-      .csv(testFile(unescapedQuotesAndUnescapedDelimiterFile)).collect()
-    val exceptResults = Array(Row("""a,""b,c""", "xyz"), Row("""a,b,c""", """x""yz"""))
-    assert(result.sameElements(exceptResults))
+    "unescaped quotes and unescaped delimiter data correctly") {
+    withTempPath { path =>
+      val dataPath = path.getCanonicalPath
+      val row1 = Row("""a,""b,c""", "xyz")
+      val row2 = Row("""a,b,c""", """x""yz""")
+      // Generate the test data, use `,` as delimiter and `"` as quotes, but they didn't escape.
+      Seq(
+        """c1,c2""",
+        s""""${row1.getString(0)}","${row1.getString(1)}"""",
+        s""""${row2.getString(0)}","${row2.getString(1)}"""")
+        .toDF().repartition(1).write.text(dataPath)
+      // Without configure UnescapedQuoteHandling to STOP_AT_CLOSING_QUOTE,
+      // the result will be Row(""""a,""b""", """c""""), Row("""a,b,c""", """"x""yz"""")
+      val result = spark.read
+        .option("inferSchema", "true")
+        .option("header", "true")
+        .option("unescapedQuoteHandling", "STOP_AT_CLOSING_QUOTE")
+        .csv(dataPath).collect()
+      val exceptResults = Array(row1, row2)
+      assert(result.sameElements(exceptResults))
+    }
   }
 }