[incubator-gobblin] branch master updated: [GOBBLIN-1295] Provide escape for each character of '"' where hive doesn't do it properly

[su...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

suvasude pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-gobblin.git


The following commit(s) were added to refs/heads/master by this push:
     new 73a2a1c  [GOBBLIN-1295] Provide escape for each character of '"' where hive doesn't do it properly
73a2a1c is described below

commit 73a2a1c945918a6a606b1c467fcac099d40a0f7e
Author: Lei Sun <au...@gmail.com>
AuthorDate: Tue Oct 20 14:27:57 2020 -0700

    [GOBBLIN-1295] Provide escape for each character of '"' where hive doesn't do it properly
    
    Replacement of replaceAll() by replace() to avoid
    confusion in terms of the number of slashes
    
    Closes #3133 from autumnust/escapeToMakeUpForHive
---
 .../src/main/java/org/apache/gobblin/util/AvroUtils.java    |  3 ++-
 .../test/java/org/apache/gobblin/util/AvroUtilsTest.java    | 13 +++++++++++++
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/gobblin-utility/src/main/java/org/apache/gobblin/util/AvroUtils.java b/gobblin-utility/src/main/java/org/apache/gobblin/util/AvroUtils.java
index 65d3279..cd72883 100644
--- a/gobblin-utility/src/main/java/org/apache/gobblin/util/AvroUtils.java
+++ b/gobblin-utility/src/main/java/org/apache/gobblin/util/AvroUtils.java
@@ -891,7 +891,8 @@ public class AvroUtils {
    * Therefore the escaping behavior won't cause correctness issues.
    */
   public static String sanitizeSchemaString(String schemaString) {
-    return schemaString.replaceAll(";",  "\\\\;").replaceAll("'", "\\\\'");
+    return schemaString.replace("\\\"", "\\\\\\\"").replace(";",  "\\;")
+        .replace("'", "\\'");
   }
 
   /**
diff --git a/gobblin-utility/src/test/java/org/apache/gobblin/util/AvroUtilsTest.java b/gobblin-utility/src/test/java/org/apache/gobblin/util/AvroUtilsTest.java
index e7a8c35..991be36 100644
--- a/gobblin-utility/src/test/java/org/apache/gobblin/util/AvroUtilsTest.java
+++ b/gobblin-utility/src/test/java/org/apache/gobblin/util/AvroUtilsTest.java
@@ -272,6 +272,19 @@ public class AvroUtilsTest {
     Assert.assertEquals(actualString, expectedString);
     // Verify that there's only one slash being added.
     Assert.assertEquals(actualString.length(), invalidString.length() + 2);
+
+    // An instance of invalid string that contains a slash followed by a quote, both of which should be escaped.
+    String invalidStringWithSlash = "abc\\\"";
+    // Should have a slash before the actual slash, and a slash before the actual quote.
+    actualString = AvroUtils.sanitizeSchemaString(invalidStringWithSlash);
+    // Meaning for each slash:
+    // first two: escape in java and the actual escape the output string
+    // second pair: escape in java and the actual slash
+    // third pair: escape for the actual quote
+    // last pair: java escape slash with the actual quote.
+    expectedString = "abc\\\\\\\"";
+    Assert.assertEquals(actualString, expectedString);
+    Assert.assertEquals(actualString.length(), invalidStringWithSlash.length() + 2);
   }
 
   public static List<GenericRecord> getRecordFromFile(String path)