disallow whitespace characters except space in data source names (#8465)

* disallow whitespace characters in data source names

* wrapped preconditions in a function, and simplify unit tests code

* Fixed regex to allow space, simplified repeat logic

* Fixed import style against mvn checkstyle

* Add msg in case test fails, use emptyMap(), improved naming

* Changes on assertion functions

* change wording of "whitespace" to "whitespace except space" to avoid misleading

server/src/main/java/org/apache/druid/segment/indexing/DataSchema.java

@@ -41,13 +41,17 @@ import java.util.Arrays;
 import java.util.HashSet;
 import java.util.Map;
 import java.util.Set;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
 
 /**
+ *
  */
 public class DataSchema
 {
   private static final Logger log = new Logger(DataSchema.class);
-
+  private static final Pattern INVALIDCHARS = Pattern.compile("(?s).*[^\\S ].*");
   private final String dataSource;
   private final Map<String, Object> parser;
   private final AggregatorFactory[] aggregators;
@@ -72,8 +76,7 @@ public class DataSchema
     this.parser = parser;
     this.transformSpec = transformSpec == null ? TransformSpec.NONE : transformSpec;
 
-    Preconditions.checkArgument(!Strings.isNullOrEmpty(dataSource), "dataSource cannot be null or empty. Please provide a dataSource.");
-    Preconditions.checkArgument(!dataSource.contains("/"), "dataSource cannot contain the '/' character.");
+    validateDatasourceName(dataSource);
     this.dataSource = dataSource;
 
     if (granularitySpec == null) {
@@ -98,6 +101,20 @@ public class DataSchema
     this.aggregators = aggregators == null ? new AggregatorFactory[]{} : aggregators;
   }
 
+  static void validateDatasourceName(String dataSource)
+  {
+    Preconditions.checkArgument(
+        !Strings.isNullOrEmpty(dataSource),
+        "dataSource cannot be null or empty. Please provide a dataSource."
+    );
+    Matcher m = INVALIDCHARS.matcher(dataSource);
+    Preconditions.checkArgument(
+        !m.matches(),
+        "dataSource cannot contain whitespace character except space."
+    );
+    Preconditions.checkArgument(!dataSource.contains("/"), "dataSource cannot contain the '/' character.");
+  }
+
   @JsonProperty
   public String getDataSource()
   {

server/src/test/java/org/apache/druid/segment/indexing/DataSchemaTest.java

@@ -22,6 +22,7 @@ package org.apache.druid.segment.indexing;
 import com.fasterxml.jackson.databind.JsonMappingException;
 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.google.common.collect.ImmutableList;
+import com.google.common.collect.ImmutableMap;
 import com.google.common.collect.ImmutableSet;
 import org.apache.druid.data.input.InputRow;
 import org.apache.druid.data.input.impl.DimensionsSpec;
@@ -52,6 +53,7 @@ import java.io.IOException;
 import java.nio.ByteBuffer;
 import java.nio.charset.StandardCharsets;
 import java.util.Arrays;
+import java.util.Collections;
 import java.util.Map;
 
 public class DataSchemaTest
@@ -286,7 +288,7 @@ public class DataSchemaTest
     // Jackson creates a default type parser (StringInputRowParser) for an invalid type.
     schema.getParser();
   }
-  
+
   @Test
   public void testEmptyDatasource()
   {
@@ -324,6 +326,42 @@ public class DataSchemaTest
     );
   }
 
+
+  @Test
+  public void testInvalidWhitespaceDatasource()
+  {
+    Map<String, String> invalidCharToDataSourceName = ImmutableMap.of(
+        "\\t", "\tab\t",
+        "\\r", "\rcarriage\return\r",
+        "\\n", "\nnew\nline\n"
+    );
+
+    for (Map.Entry<String, String> entry : invalidCharToDataSourceName.entrySet()) {
+      testInvalidWhitespaceDatasourceHelper(entry.getValue(), entry.getKey());
+    }
+  }
+
+  private void testInvalidWhitespaceDatasourceHelper(String dataSource, String invalidChar)
+  {
+    String testFailMsg = "dataSource contain invalid whitespace character: " + invalidChar;
+    try {
+      DataSchema schema = new DataSchema(
+          dataSource,
+          Collections.emptyMap(),
+          null,
+          null,
+          null,
+          jsonMapper
+      );
+      Assert.fail(testFailMsg);
+    }
+    catch (IllegalArgumentException errorMsg) {
+      String expectedMsg = "dataSource cannot contain whitespace character except space.";
+      Assert.assertEquals(testFailMsg, expectedMsg, errorMsg.getMessage());
+    }
+  }
+
+
   @Test
   public void testSerde() throws Exception
   {