[SPARK-26982][SQL] Enhance describe framework to describe the output of a query. #23883

dilipbiswal · 2019-02-25T00:51:13Z

What changes were proposed in this pull request?

Currently we can use df.printSchema to discover the schema information for a query. We should have a way to describe the output schema of a query using SQL interface.

Example:

DESCRIBE SELECT * FROM desc_table
DESCRIBE QUERY SELECT * FROM desc_table

spark-sql> create table desc_table (c1 int comment 'c1-comment', c2 decimal comment 'c2-comment', c3 string);

spark-sql> desc select * from desc_table;
c1	int	        c1-comment
c2	decimal(10,0)	c2-comment
c3	string	        NULL

How was this patch tested?

Added a new test under SQLQueryTestSuite and SparkSqlParserSuite

SparkQA · 2019-02-25T02:29:54Z

Test build #102728 has finished for PR 23883 at commit c7e57e2.

This patch fails Spark unit tests.
This patch merges cleanly.
This patch adds no public classes.

dilipbiswal · 2019-02-25T02:34:50Z

retest this please

SparkQA · 2019-02-25T06:46:24Z

Test build #102731 has finished for PR 23883 at commit c7e57e2.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

dilipbiswal · 2019-02-25T06:49:04Z

cc @cloud-fan @gatorsmile

viirya · 2019-02-25T07:40:00Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

 * {{{
 *   DESCRIBE [EXTENDED|FORMATTED] table_name partitionSpec?;
+ *   DESCRIBE QUERY <query>
+ *   DESCRIBE <query>


DESCRIBE QUERY <query> and DESCRIBE <query> equal to DESCRIBE [QUERY] <query>?

Yeah... would change.

viirya · 2019-02-25T07:41:08Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

 * {{{
 *   DESCRIBE [EXTENDED|FORMATTED] table_name partitionSpec?;
+ *   DESCRIBE QUERY <query>
+ *   DESCRIBE <query>
 * }}}
 */
 case class DescribeTableCommand(


As this is named as DescribeTableCommand, shall we have individual command for describing query?

Otherwise, shall we rename it to DescribeTableOrQueryCommand?

@viirya We could. Just that there are not that much changes. Lets see if others feel the same way. If so, i would change it. Renaming the command sounds good to me :-)

maropu · 2019-02-25T10:05:35Z

Thanks for the work! I have one question in adivance; any other database-like system supprting this syntax? Also, can you add a entry Query in https://github.com/apache/spark/blob/master/docs/sql-reserved-and-non-reserved-keywords.md and update entries in TableIdentifierParserSuite?

dilipbiswal · 2019-02-25T11:22:35Z

@maropu Thanks. I will update the entries. I got this idea from DB2 which supports this syntax. Here is the sample output for your perusal.

db2 => describe select * from desc_tab;

 Column Information

 Number of columns: 2

 SQL type              Type length  Column name                     Name length
 --------------------  -----------  ------------------------------  -----------
 497   INTEGER                   4  C1                                        2
 449   VARCHAR                  10  C2                                        2

gatorsmile · 2019-02-25T19:47:33Z

Thanks! This Describe command is pretty useful to SQL users!

Without this command, our Spark users need to create a temp/persistent view first and then describe it.

maropu · 2019-02-26T01:03:00Z

sql/core/src/main/scala/org/apache/spark/sql/execution/SparkSqlParser.scala

-      DescribeTableCommand(
-        visitTableIdentifier(ctx.tableIdentifier),
+      DescribeTableOrQueryCommand(
+        Option(visitTableIdentifier(ctx.tableIdentifier)),


visitTableIdentifier can return null?

@maropu with the change, DescribeTableOrQueryCommand takes a optional table identifier. Thats why i wrap it under an Option.

sql/core/src/main/scala/org/apache/spark/sql/execution/SparkSqlParser.scala

+    DescribeTableOrQueryCommand(None,
+      Option(visitQuery(ctx.query())),
+      Map.empty[String, String], false)
+  }


maropu · 2019-02-26T01:11:51Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

 * {{{
 *   DESCRIBE [EXTENDED|FORMATTED] table_name partitionSpec?;
+ *   DESCRIBE [QUERY] <query>


nit: DESCRIBE [QUERY] query? We need <>?

If its under the bracket , it means optional isn't it ? Like in the previous example extended and formatted are optional ?

Yea, [QUERY] means optional. You mean <query> means optional, too?

@maropu Sorry .. i misinterpreted. Here , i am just trying to say that they have to specify a query. Its not optional. How should i say it ? Does the following look better ?

DESCRIBE [QUERY] query_statement

maropu · 2019-02-26T01:37:37Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

+          throw new AnalysisException("Describe command is not supported for insert statements.")
+        case u @ Union(children) if children.forall(_.isInstanceOf[InsertIntoTable]) =>
+          throw new AnalysisException("Describe command is not supported for insert statements.")
+        case u: UnresolvedRelation =>


Can we put the two matches above together?

case p if p.collectFirst { case _: InsertIntoTable | _: InsertIntoDir => true }.isDefined =>

@maropu This is fantastic :-). Thanks.. Let me try it.

maropu · 2019-02-26T02:03:26Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

+        case u @ Union(children) if children.forall(_.isInstanceOf[InsertIntoTable]) =>
+          throw new AnalysisException("Describe command is not supported for insert statements.")
+        case u: UnresolvedRelation =>
+          describeTable(sparkSession, u.tableIdentifier, partitionSpec, isExtended, result)


Can you add a test for this match in describe-query.sql? Also, it might be better to add comments here about which query does this match cover.

@maropu this should have been already covered in describe-table tests. Basically,
a simpledescribe table mytable will go through it. Thats because its matched by queryPrimary in sqlBase.g4. I will add comments.

maropu

I left some minor comments. Looks good except the comments.

cloud-fan · 2019-02-26T03:19:17Z

sql/catalyst/src/main/antlr4/org/apache/spark/sql/catalyst/parser/SqlBase.g4

@@ -784,7 +785,7 @@ ansiNonReserved
    | RESET | RESTRICT | REVOKE | RLIKE | ROLE | ROLES | ROLLBACK | ROLLUP | ROW | ROWS | SCHEMA | SEPARATED | SERDE
    | SERDEPROPERTIES | SET | SETS | SHOW | SKEWED | SORT | SORTED | START | STATISTICS | STORED | STRATIFY | STRUCT
    | TABLES | TABLESAMPLE | TBLPROPERTIES | TEMPORARY | TERMINATED | TOUCH | TRANSACTION | TRANSACTIONS | TRANSFORM
-    | TRUE | TRUNCATE | UNARCHIVE | UNBOUNDED | UNCACHE | UNLOCK | UNSET | USE | VALUES | VIEW | WINDOW
+    | TRUE | TRUNCATE | UNARCHIVE | UNBOUNDED | UNCACHE | UNLOCK | UNSET | USE | VALUES | VIEW | WINDOW | QUERY


let's order it with alphabet

cloud-fan · 2019-02-26T03:20:57Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

 * }}}
 */
-case class DescribeTableCommand(
-    table: TableIdentifier,
+case class DescribeTableOrQueryCommand(


shall we create 2 commands?

SparkQA · 2019-02-26T03:46:33Z

Test build #102771 has finished for PR 23883 at commit db20bfb.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

SongYadong · 2019-02-26T10:46:21Z

...atalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/TableIdentifierParserSuite.scala

@@ -48,7 +48,7 @@ class TableIdentifierParserSuite extends SparkFunSuite {
    "false", "fetch", "float", "for", "grant", "group", "grouping", "import", "in",
    "insert", "int", "into", "is", "pivot", "lateral", "like", "local", "none", "null",
    "of", "order", "out", "outer", "partition", "percent", "procedure", "range", "reads", "revoke",
-    "rollup", "row", "rows", "set", "smallint", "table", "timestamp", "to", "trigger",
+    "rollup", "row", "rows", "set", "smallint", "table", "query", "timestamp", "to", "trigger",


super nit: if with alphabet order, put it between "procedure" and "range" may be better :)

… query

SparkQA · 2019-02-27T02:03:58Z

Test build #102804 has finished for PR 23883 at commit b6a13da.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

dilipbiswal · 2019-02-27T03:41:22Z

@cloud-fan I have addressed your comments in my last checkin - fyi

cloud-fan · 2019-02-27T04:54:18Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

+    val result = new ArrayBuffer[Row]
+    query match {
+      case p if p.collectFirst { case _: InsertIntoTable | _: InsertIntoDir => true }.isDefined =>
+        throw new AnalysisException("Describe command is not supported for insert statements.")


is this behavior from DB2? sql("INSERT ...").printSchema works and prints empty schema, shall we make the behavior consistent?

@cloud-fan In DB2, we get an error when we describe on "insert"

db2 => describe insert into t1 values (1,1) SQL0104N An unexpected token "insert" was found following "DESCRIBE".

Personally, i feel its better to let users know that this command is not meant for "insert" than returning an empty schema. But if you feel otherwise, i will change. Let me know please. You have a point about printSchema though :-)

I don't have a strong preference here, but the DB2 does this in the parser, I'm wondering if we should do this in the parser as well.

instead of (DESC | DESCRIBE) QUERY? query, how about (DESC | DESCRIBE) QUERY? querySpecification?

@cloud-fan the only thing is we may leave out a few statements if we restrict it to querySpecification like VALUES (...), WITH.. etc .. Is it okay if we create a parserRule like following ?

queryDescribe
: stmtType1
| stmtType2
;

@cloud-fan I tried to add a rule, but it actually gets complicated as the rules are intertwined. In order to do this, i have to create multiple parallel rules for describequery and then implement the visit method. So lets throw analysis exception for now ?

then how about go with querySpecification, to support only the SELECT ... query at the beginning? we can support more in the future.

@cloud-fan OK.. let me give that a try.

SparkQA · 2019-02-27T12:11:38Z

Test build #102815 has finished for PR 23883 at commit e742ae9.

This patch fails Spark unit tests.
This patch merges cleanly.
This patch adds the following public classes (experimental):
case class DescribeTableCommand(

SparkQA · 2019-02-28T08:05:02Z

Test build #102853 has finished for PR 23883 at commit cfbe8db.

This patch fails due to an unknown error code, -9.
This patch merges cleanly.
This patch adds the following public classes (experimental):
abstract class DescribeCommandBase extends RunnableCommand

cloud-fan · 2019-02-28T11:35:21Z

retest this please

SparkQA · 2019-02-28T16:18:45Z

Test build #102858 has finished for PR 23883 at commit cfbe8db.

This patch passes all tests.
This patch merges cleanly.
This patch adds the following public classes (experimental):
abstract class DescribeCommandBase extends RunnableCommand

SparkQA · 2019-03-01T00:52:12Z

Test build #102873 has finished for PR 23883 at commit c519fc7.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

maropu · 2019-03-01T01:23:06Z

sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala

@@ -117,6 +117,11 @@ class AstBuilder(conf: SQLConf) extends SqlBaseBaseVisitor[AnyRef] with Logging
    }
  }

+  override def visitDescQuery(ctx: DescQueryContext): LogicalPlan = withOrigin(ctx) {
+    plan(ctx.queryTerm).
+      optionalMap(ctx.queryOrganization)(withQueryResultClauses)


nit:

override def visitDescQuery(ctx: DescQueryContext): LogicalPlan = withOrigin(ctx) { plan(ctx.queryTerm).optionalMap(ctx.queryOrganization)(withQueryResultClauses) }

maropu · 2019-03-01T01:24:59Z

sql/core/src/main/scala/org/apache/spark/sql/execution/HiveResult.scala

@@ -22,7 +22,7 @@ import java.sql.{Date, Timestamp}

 import org.apache.spark.sql.Row
 import org.apache.spark.sql.catalyst.util.{DateFormatter, DateTimeUtils, TimestampFormatter}
-import org.apache.spark.sql.execution.command.{DescribeTableCommand, ExecutedCommandExec, ShowTablesCommand}
+import org.apache.spark.sql.execution.command._


import org.apache.spark.sql.execution.command.{DescribeCommandBase, ExecutedCommandExec, ShowTablesCommand}?

maropu · 2019-03-01T01:38:32Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

@@ -516,6 +505,35 @@ case class DescribeTableCommand(
      new MetadataBuilder().putString("comment", "comment of the column").build())()
  )

+  def describeSchema(


How about this?

protected def describeSchema(schema: StructType, header: Boolean): Seq[Row] = { val buffer = new ArrayBuffer[Row] if (header) { buffer += Row(s"# ${output.head.name}", output(1).name, output(2).name) } schema.foreach { column => buffer += Row(column.name, column.dataType.simpleString, column.getComment().orNull) } buffer }

@maropu describeSchema has two callers. One from the describeTable code path and one from describeQuery. The other caller passes the row buffer.

Ah, I see. Can you add protected, anyway?

cloud-fan · 2019-03-01T05:10:27Z

sql/catalyst/src/main/antlr4/org/apache/spark/sql/catalyst/parser/SqlBase.g4

@@ -255,6 +256,10 @@ query
    : ctes? queryNoWith
    ;

+descQuery


nit: queryToDesc

cloud-fan · 2019-03-01T05:12:49Z

LGTM

dongjoon-hyun · 2019-03-01T05:46:54Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

+  }
+}
+/**
+ * Command can of the following form.


can of the -> can be the?

Otherwise, shall we reuse the previous description, Command that looks like?

dongjoon-hyun · 2019-03-01T05:47:09Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

-    }
-  }
+/**
+ * Command can of the following form.


dongjoon-hyun · 2019-03-01T05:53:17Z

sql/core/src/test/resources/sql-tests/inputs/describe-query.sql

+
+
+-- Error cases.
+DESC WITH s AS (SELECT 'hello' as col1) SELECT * FROM s;


Hi, @dilipbiswal .
Is this the only exception case in SELECT syntax? Do we have a plan to support this SELECT query with WITH?

@dongjoon-hyun Yeah.. Wenchen suggested that we start with simple selects and then improve on it. I am planning to look into CTEs next.

dongjoon-hyun · 2019-03-01T06:15:37Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

+/**
+ * Command can of the following form.
+ * {{{
+ *   DESCRIBE [QUERY] query_statement


query_statement -> select_statement?

Also, if possible, let's add some comment below line 627 that we don't support CTE syntax yet.

@dongjoon-hyun Now we support statements of the form :

SELECT

SELECTS wrapped under a set op (union/intersect etc)

VALUES(....)

TABLE
Thats why i didn't use select_statement.. Any other name you can think of ?

I got it. BTW, TABLE and VALUES doesn't look like query statement. What about enumerating them all clearly?

dongjoon-hyun · 2019-03-01T06:27:57Z

sql/core/src/test/resources/sql-tests/inputs/describe-query.sql

+-- Test tables
+CREATE table  desc_temp1 (key int COMMENT 'column_comment', val string) USING PARQUET;
+CREATE table  desc_temp2 (key int, val string) USING PARQUET;
+CREATE table  desc_temp3 (key int, val string) USING PARQUET;


Shall we remove this 3rd CREATE table desc_temp3? That is only used at multi-insert test case, FROM ... INSERT ... INSERT. And, we can use like the following.

FROM desc_temp1 a INSERT INTO desc_temp1 SELECT * INSERT INTO desc_temp2 SELECT *;

dongjoon-hyun · 2019-03-01T06:29:46Z

sql/core/src/test/resources/sql-tests/inputs/describe-query.sql

+DESC SELECT key, COUNT(*) as count FROM desc_temp1 group by key;
+DESC SELECT 10.00D as col1;
+DESC QUERY SELECT key FROM desc_temp1 UNION ALL select CAST(1 AS DOUBLE);
+DESC QUERY VALUES(1.00D, 'hello') as tab1(col1, col2);


Shall we add DESC QUERY FROM desc_temp1 a SELECT * at line 13?

SparkQA · 2019-03-01T08:01:47Z

Test build #102898 has finished for PR 23883 at commit ee858a2.

This patch fails Spark unit tests.
This patch merges cleanly.
This patch adds no public classes.

dongjoon-hyun · 2019-03-01T21:01:34Z

sql/core/src/test/resources/sql-tests/inputs/describe-query.sql

+-- cleanup
+DROP TABLE desc_temp1;
+DROP TABLE desc_temp2;
+DROP TABLE desc_temp3;


We need to remove this line, too. And, could you add a new line at the end? GitHub shows a warning here.

dongjoon-hyun · 2019-03-01T21:02:14Z

sql/core/src/test/resources/sql-tests/results/describe-query.sql.out

+struct<>
+-- !query 16 output
+org.apache.spark.sql.AnalysisException
+Table or view not found: desc_temp3;


To remove this Table or view not found exception, please don't forget to regenerate this~

SparkQA · 2019-03-02T01:14:00Z

Test build #102927 has finished for PR 23883 at commit 7556506.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

SparkQA · 2019-03-02T02:10:34Z

Test build #102930 has finished for PR 23883 at commit 7d7d7d2.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

cloud-fan · 2019-03-02T02:38:34Z

sql/core/src/main/scala/org/apache/spark/sql/execution/command/tables.scala

+ * 1. SELECT statements
+ * 2. SELECT statements inside set operators (UNION, INTERSECT etc)
+ * 3. VALUES statement.
+ * 4. TABLE statement. Example : TABLE table_name


do you mean we support DESCRIBE QUERY TABLE t?

nvm, just checked and we do support it...

cloud-fan · 2019-03-02T03:19:50Z

thanks, merging to master!

…of a query. Currently we can use `df.printSchema` to discover the schema information for a query. We should have a way to describe the output schema of a query using SQL interface. Example: DESCRIBE SELECT * FROM desc_table DESCRIBE QUERY SELECT * FROM desc_table ```SQL spark-sql> create table desc_table (c1 int comment 'c1-comment', c2 decimal comment 'c2-comment', c3 string); spark-sql> desc select * from desc_table; c1 int c1-comment c2 decimal(10,0) c2-comment c3 string NULL ``` Added a new test under SQLQueryTestSuite and SparkSqlParserSuite Closes apache#23883 from dilipbiswal/dkb_describe_query. Authored-by: Dilip Biswal <dbiswal@us.ibm.com> Signed-off-by: Wenchen Fan <wenchen@databricks.com>

dilipbiswal changed the title ~~[SPARK-26982][SQL] Enhance describe frame work to describe the output of a query.~~ [SPARK-26982][SQL] Enhance describe framework to describe the output of a query. Feb 25, 2019

viirya reviewed Feb 25, 2019

View reviewed changes

maropu reviewed Feb 26, 2019

View reviewed changes

sql/core/src/main/scala/org/apache/spark/sql/execution/SparkSqlParser.scala

DescribeTableOrQueryCommand(None,

Option(visitQuery(ctx.query())),

Map.empty[String, String], false)

}

This comment was marked as outdated.

Sign in to view

maropu reviewed Feb 26, 2019

View reviewed changes

cloud-fan reviewed Feb 26, 2019

View reviewed changes

SongYadong reviewed Feb 26, 2019

View reviewed changes

dilipbiswal added 6 commits February 26, 2019 11:31

[SPARK-26982] Enhance describe frame work to describe the output of a…

dfc3e13

… query

Code review

9f6b879

code review2

03bf58f

review

c59c1f6

code review

0853284

Code review

b6a13da

dilipbiswal force-pushed the dkb_describe_query branch from db20bfb to b6a13da Compare February 26, 2019 21:28

cloud-fan reviewed Feb 27, 2019

View reviewed changes

Code review

cfbe8db

more review

c519fc7

maropu reviewed Mar 1, 2019

View reviewed changes

comments

ee858a2

cloud-fan reviewed Mar 1, 2019

View reviewed changes

dongjoon-hyun reviewed Mar 1, 2019

View reviewed changes

more ..

7556506

dongjoon-hyun reviewed Mar 1, 2019

View reviewed changes

remove extra drop table

7d7d7d2

cloud-fan reviewed Mar 2, 2019

View reviewed changes

cloud-fan closed this in 4a486d6 Mar 2, 2019



		-- Error cases.
		DESC WITH s AS (SELECT 'hello' as col1) SELECT * FROM s;

[SPARK-26982][SQL] Enhance describe framework to describe the output of a query. #23883

[SPARK-26982][SQL] Enhance describe framework to describe the output of a query. #23883

Conversation

dilipbiswal commented Feb 25, 2019

What changes were proposed in this pull request?

How was this patch tested?

SparkQA commented Feb 25, 2019

dilipbiswal commented Feb 25, 2019

SparkQA commented Feb 25, 2019

dilipbiswal commented Feb 25, 2019

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dilipbiswal Feb 25, 2019 • edited Loading

Choose a reason for hiding this comment

maropu commented Feb 25, 2019 • edited Loading

dilipbiswal commented Feb 25, 2019 • edited Loading

gatorsmile commented Feb 25, 2019 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment was marked as outdated.

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dilipbiswal Feb 26, 2019 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dilipbiswal Feb 26, 2019 • edited Loading

Choose a reason for hiding this comment

maropu Feb 26, 2019 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

maropu left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

SparkQA commented Feb 26, 2019

Choose a reason for hiding this comment

SparkQA commented Feb 27, 2019

dilipbiswal commented Feb 27, 2019

Choose a reason for hiding this comment

dilipbiswal Feb 27, 2019 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

SparkQA commented Feb 27, 2019

SparkQA commented Feb 28, 2019

cloud-fan commented Feb 28, 2019

SparkQA commented Feb 28, 2019

SparkQA commented Mar 1, 2019

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

maropu Mar 1, 2019 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

cloud-fan commented Mar 1, 2019

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dongjoon-hyun Mar 1, 2019 • edited Loading

Choose a reason for hiding this comment

SparkQA commented Mar 1, 2019

Choose a reason for hiding this comment

Choose a reason for hiding this comment

SparkQA commented Mar 2, 2019

dilipbiswal Feb 25, 2019 •

edited

Loading

maropu commented Feb 25, 2019 •

edited

Loading

dilipbiswal commented Feb 25, 2019 •

edited

Loading

gatorsmile commented Feb 25, 2019 •

edited

Loading

dilipbiswal Feb 26, 2019 •

edited

Loading

dilipbiswal Feb 26, 2019 •

edited

Loading

maropu Feb 26, 2019 •

edited

Loading

dilipbiswal Feb 27, 2019 •

edited

Loading

maropu Mar 1, 2019 •

edited

Loading

dongjoon-hyun Mar 1, 2019 •

edited

Loading