From 333c9aeeccbe10d0cdc437be15ab15820676bb47 Mon Sep 17 00:00:00 2001
From: chertus <chertus@gmail.com>
Date: Thu, 23 May 2019 19:01:17 +0300
Subject: [PATCH] support IN in CrossToInnerJoin push down

---
 .../Interpreters/CrossToInnerJoinVisitor.cpp  | 27 +++++++++++++++++++
 .../0_stateless/00863_comma_join_in.reference |  3 +++
 .../0_stateless/00863_comma_join_in.sql       | 21 +++++++++++++++
 3 files changed, 51 insertions(+)
 create mode 100644 dbms/tests/queries/0_stateless/00863_comma_join_in.reference
 create mode 100644 dbms/tests/queries/0_stateless/00863_comma_join_in.sql

diff --git a/dbms/src/Interpreters/CrossToInnerJoinVisitor.cpp b/dbms/src/Interpreters/CrossToInnerJoinVisitor.cpp
index ba97ca0186b..468f96c8352 100644
--- a/dbms/src/Interpreters/CrossToInnerJoinVisitor.cpp
+++ b/dbms/src/Interpreters/CrossToInnerJoinVisitor.cpp
@@ -4,6 +4,7 @@
 #include <Interpreters/CrossToInnerJoinVisitor.h>
 #include <Interpreters/DatabaseAndTableWithAlias.h>
 #include <Interpreters/IdentifierSemantic.h>
+#include <Interpreters/QueryNormalizer.h> // for functionIsInOperator
 #include <Parsers/ASTSelectQuery.h>
 #include <Parsers/ASTTablesInSelectQuery.h>
 #include <Parsers/ASTIdentifier.h>
@@ -120,6 +121,12 @@ public:
         {
             /// leave other comparisons as is
         }
+        else if (functionIsInOperator(node.name)) /// IN, NOT IN
+        {
+            if (auto ident = node.arguments->children.at(0)->as<ASTIdentifier>())
+                if (size_t min_table = checkIdentifier(*ident))
+                    asts_to_join_on[min_table].push_back(ast);
+        }
         else
         {
             ands_only = false;
@@ -202,6 +209,26 @@ private:
         }
         return 0;
     }
+
+    size_t checkIdentifier(const ASTIdentifier & identifier)
+    {
+        size_t best_match = 0;
+        size_t berst_table_pos = 0;
+
+        for (size_t i = 0; i < tables.size(); ++i)
+        {
+            size_t match = IdentifierSemantic::canReferColumnToTable(identifier, tables[i].table);
+            if (match > best_match)
+            {
+                best_match = match;
+                berst_table_pos = i;
+            }
+        }
+
+        if (best_match && tables[berst_table_pos].canAttachOnExpression())
+            return berst_table_pos;
+        return 0;
+    }
 };
 
 using CheckExpressionMatcher = OneTypeMatcher<CheckExpressionVisitorData, false>;
diff --git a/dbms/tests/queries/0_stateless/00863_comma_join_in.reference b/dbms/tests/queries/0_stateless/00863_comma_join_in.reference
new file mode 100644
index 00000000000..01e79c32a8c
--- /dev/null
+++ b/dbms/tests/queries/0_stateless/00863_comma_join_in.reference
@@ -0,0 +1,3 @@
+1
+2
+3
diff --git a/dbms/tests/queries/0_stateless/00863_comma_join_in.sql b/dbms/tests/queries/0_stateless/00863_comma_join_in.sql
new file mode 100644
index 00000000000..d67217ccc41
--- /dev/null
+++ b/dbms/tests/queries/0_stateless/00863_comma_join_in.sql
@@ -0,0 +1,21 @@
+drop table if exists test1;
+drop table if exists test2;
+drop table if exists test3;
+
+create table test1 (id UInt64, code String) engine = Memory;
+create table test3 (id UInt64, code String) engine = Memory;
+create table test2 (id UInt64, code String, test1_id UInt64, test3_id UInt64) engine = Memory;
+
+insert into test1 (id, code) select number, toString(number) FROM numbers(100000);
+insert into test3 (id, code) select number, toString(number) FROM numbers(100000);
+insert into test2 (id, code, test1_id, test3_id) select number, toString(number), number, number FROM numbers(100000);
+
+select test2.id
+from test1, test2, test3
+where test1.code in ('1', '2', '3')
+    and test2.test1_id = test1.id
+    and test2.test3_id = test3.id;
+
+drop table test1;
+drop table test2;
+drop table test3;