EXPLAIN FORMAT=JSON

Add support for semi-join strategies: FirstMatch, DuplicateWeedout,
LooseScan.

mysql-test/r/explain_json.result

@@ -592,5 +592,77 @@ EXPLAIN
     }
   }
 }
+# 
+# First-Match
+# 
+explain
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	10	
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	10	Using where; FirstMatch(t2); Using join buffer (flat, BNL join)
+explain format=json
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+EXPLAIN
+{
+  "query_block": {
+    "select_id": 1,
+    "table": {
+      "table_name": "t2",
+      "access_type": "ALL",
+      "rows": 10,
+      "filtered": 100
+    },
+    "block-nl-join": {
+      "table": {
+        "table_name": "t1",
+        "access_type": "ALL",
+        "rows": 10,
+        "filtered": 100,
+        "first_match": "t2"
+      },
+      "buffer_type": "flat",
+      "join_type": "BNL",
+      "attached_condition": "((t1.b = t2.b) and (t1.a = t2.a))"
+    }
+  }
+}
+# 
+# Duplicate Weedout
+# 
+set @tmp= @@optimizer_switch;
+set optimizer_switch='firstmatch=off';
+explain
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	10	
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	10	Using where; Start temporary; End temporary; Using join buffer (flat, BNL join)
+explain format=json
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+EXPLAIN
+{
+  "query_block": {
+    "select_id": 1,
+    "table": {
+      "table_name": "t2",
+      "access_type": "ALL",
+      "rows": 10,
+      "filtered": 100
+    },
+    "duplicates_removal": {
+      "block-nl-join": {
+        "table": {
+          "table_name": "t1",
+          "access_type": "ALL",
+          "rows": 10,
+          "filtered": 100
+        },
+        "buffer_type": "flat",
+        "join_type": "BNL",
+        "attached_condition": "((t1.b = t2.b) and (t1.a = t2.a))"
+      }
+    }
+  }
+}
+set optimizer_switch=@tmp;
 drop table t1,t2;
 drop table t0;

mysql-test/t/explain_json.test

@@ -114,6 +114,25 @@ insert into t2 select * from t1;
 explain format=json
 select * from t1,t2 where t1.a in ( select a from t0);
 
+--echo # 
+--echo # First-Match
+--echo # 
+explain
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+explain format=json
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+
+--echo # 
+--echo # Duplicate Weedout
+--echo # 
+set @tmp= @@optimizer_switch;
+set optimizer_switch='firstmatch=off';
+explain
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+explain format=json
+select * from t2 where t2.a in ( select a from t1 where t1.b=t2.b);
+set optimizer_switch=@tmp;
+
 drop table t1,t2;
 drop table t0;
 

sql/sql_explain.cc

@@ -732,7 +732,13 @@ void Explain_basic_join::print_explain_json(Explain_query *query,
   writer->add_member("select_id").add_ll(select_id);
   for (uint i=0; i< n_join_tabs; i++)
   {
+    if (join_tabs[i]->start_dups_weedout)
+      writer->add_member("duplicates_removal").start_object();
+    
     join_tabs[i]->print_explain_json(query, writer, is_analyze);
+    
+    if (join_tabs[i]->end_dups_weedout)
+      writer->end_object();
   }
   print_explain_json_for_children(query, writer, is_analyze);
   writer->end_object();
@@ -1089,9 +1095,7 @@ void Explain_table_access::tag_to_json(Json_writer *writer, enum explain_extra_t
       write_item(writer, pushed_index_cond);
       break;
     case ET_USING_WHERE:
-      if (where_cond)
       {
-        writer->add_member("attached_condition");
         /*
           We are printing the condition that is checked when scanning this
           table.
@@ -1099,7 +1103,11 @@ void Explain_table_access::tag_to_json(Json_writer *writer, enum explain_extra_t
           - in other cases, it is where_cond.
         */
         Item *item= bka_type.is_using_jbuf()? cache_cond: where_cond;
-        write_item(writer, item);
+        if (item)
+        {
+          writer->add_member("attached_condition");
+          write_item(writer, item);
+        }
       }
       break;
     case ET_USING_INDEX:
@@ -1110,6 +1118,15 @@ void Explain_table_access::tag_to_json(Json_writer *writer, enum explain_extra_t
       break;
     case ET_USING_JOIN_BUFFER:
       /* Do nothing. Join buffer is handled differently */
+    case ET_START_TEMPORARY:
+    case ET_END_TEMPORARY:
+      /* Handled as "duplicates_removal: { ... } */
+      break;
+    case ET_FIRST_MATCH:
+      writer->add_member("first_match").add_str(firstmatch_table_name.c_ptr());
+      break;
+    case ET_LOOSESCAN:
+      writer->add_member("loose_scan").add_bool(true);
       break;
     default:
       DBUG_ASSERT(0);

sql/sql_explain.h

@@ -539,6 +539,8 @@ class Explain_table_access : public Sql_alloc
   Explain_table_access() :
     derived_select_number(0),
     non_merged_sjm_number(0),
+    start_dups_weedout(false),
+    end_dups_weedout(false),
     where_cond(NULL),
     cache_cond(NULL),
     pushed_index_cond(NULL),
@@ -615,6 +617,9 @@ class Explain_table_access : public Sql_alloc
   EXPLAIN_BKA_TYPE bka_type;
   
   StringBuffer<32> firstmatch_table_name;
+
+  bool start_dups_weedout;
+  bool end_dups_weedout;
   
   /*
     Note: lifespan of WHERE condition is less than lifespan of this object.

sql/sql_select.cc

@@ -23641,9 +23641,16 @@ void JOIN_TAB::save_explain_data(Explain_table_access *eta, table_map prefix_tab
     }
 
     if (tab->first_weedout_table)
+    {
+      eta->start_dups_weedout= true;
       eta->push_extra(ET_START_TEMPORARY);
+    }
     if (tab->check_weed_out_table)
+    {
       eta->push_extra(ET_END_TEMPORARY);
+      eta->end_dups_weedout= true;
+    }
+
     else if (tab->do_firstmatch)
     {
       if (tab->do_firstmatch == /*join->join_tab*/ first_top_tab - 1)