HIVE-4577: hive CLI can't handle hadoop dfs command with space and qu…

…otes (Bing Li reviewed by Vaibhav Gumashta)
akitanaka · Jul 14, 2017 · adca35a · adca35a
1 parent 4af4624
commit adca35a
Show file tree

Hide file tree

Showing 4 changed files with 68 additions and 2 deletions.
diff --git a/ql/src/java/org/apache/hadoop/hive/ql/processors/DfsProcessor.java b/ql/src/java/org/apache/hadoop/hive/ql/processors/DfsProcessor.java
@@ -20,6 +20,7 @@
 
 import java.io.PrintStream;
 import java.util.Arrays;
+import java.util.ArrayList;
 import java.util.Map;
 
 import org.slf4j.Logger;
@@ -30,6 +31,7 @@
 import org.apache.hadoop.hive.conf.VariableSubstitution;
 import org.apache.hadoop.hive.metastore.api.FieldSchema;
 import org.apache.hadoop.hive.metastore.api.Schema;
+import org.apache.hadoop.hive.ql.CommandNeedRetryException;
 import org.apache.hadoop.hive.ql.security.authorization.plugin.HiveOperationType;
 import org.apache.hadoop.hive.ql.session.SessionState;
 import org.apache.hadoop.hive.ql.session.SessionState.LogHelper;
@@ -74,7 +76,7 @@ public Map<String, String> getHiveVariable() {
         }
       }).substitute(ss.getConf(), command);
 
-      String[] tokens = command.split("\\s+");
+      String[] tokens = splitCmd(command);
       CommandProcessorResponse authErrResp =
           CommandUtil.authorizeCommand(ss, HiveOperationType.DFS, Arrays.asList(tokens));
       if(authErrResp != null){
@@ -104,4 +106,60 @@ public Map<String, String> getHiveVariable() {
     }
   }
 
+  private String[] splitCmd(String command) throws CommandNeedRetryException {
+
+    ArrayList<String> paras = new ArrayList<String>();
+    int cmdLng = command.length();
+    char y = 0;
+    int start = 0;
+
+    for (int i = 0; i < cmdLng; i++) {
+      char x = command.charAt(i);
+
+      switch(x) {
+        case ' ':
+          if ((int) y == 0) {
+            String str = command.substring(start, i).trim();
+            if (!str.equals("")) {
+              paras.add(str);
+              start = i + 1;
+            }
+          }
+          break;
+        case '"':
+          if ((int) y == 0) {
+            y = x;
+            start = i + 1;
+          } else if ('"' == y) {
+            paras.add(command.substring(start, i).trim());
+            y = 0;
+            start = i + 1;
+          }
+          break;
+        case '\'':
+          if ((int) y == 0) {
+            y = x;
+            start = i + 1;
+          } else if ('\'' == y) {
+            paras.add(command.substring(start, i).trim());
+            y = 0;
+            start = i + 1;
+          }
+          break;
+        default:
+          if (i == cmdLng-1 && start < cmdLng) {
+            paras.add(command.substring(start, cmdLng).trim());
+          }
+          break;
+      }
+    }
+
+    if ((int) y != 0) {
+      console.printError("Syntax error on hadoop options: dfs " + command);
+      throw new CommandNeedRetryException();
+    }
+
+    return paras.toArray(new String[paras.size()]);
+  }
+
 }
diff --git a/ql/src/test/queries/clientpositive/dfscmd.q b/ql/src/test/queries/clientpositive/dfscmd.q
@@ -0,0 +1,7 @@
+dfs -mkdir "hello";
+dfs -mkdir 'world';
+dfs -mkdir "bei jing";
+dfs -rmr 'hello';
+dfs -rmr "world";
+dfs -rmr 'bei jing';
+
diff --git a/ql/src/test/results/clientpositive/dfscmd.q.out b/ql/src/test/results/clientpositive/dfscmd.q.out
@@ -0,0 +1 @@
+#### A masked pattern was here ####
diff --git a/ql/src/test/results/clientpositive/perf/query14.q.out b/ql/src/test/results/clientpositive/perf/query14.q.out
@@ -1,6 +1,6 @@
-Warning: Shuffle Join MERGEJOIN[890][tables = [$hdt$_1, $hdt$_2, $hdt$_0]] in Stage 'Reducer 5' is a cross product
 Warning: Shuffle Join MERGEJOIN[891][tables = [$hdt$_1, $hdt$_2, $hdt$_0]] in Stage 'Reducer 12' is a cross product
 Warning: Shuffle Join MERGEJOIN[892][tables = [$hdt$_2, $hdt$_3, $hdt$_1]] in Stage 'Reducer 16' is a cross product
+Warning: Shuffle Join MERGEJOIN[890][tables = [$hdt$_1, $hdt$_2, $hdt$_0]] in Stage 'Reducer 5' is a cross product
 PREHOOK: query: explain
 with  cross_items as
  (select i_item_sk ss_item_sk