JSON
JSON
JSONデータタイプはJSONデータをバイナリ形式で効率的に格納し、JSON関数を通じてその内部フィールドへのアクセスを可能にします。
デフォルトでは、最大1048576バイト（1MB）をサポートし、最大2147483643バイト（2GB）まで増加できます。これはstring_type_length_soft_limit_bytes設定で調整できます。
通常のSTRING型にJSON文字列を格納することと比較して、JSON型には2つの主な利点があります：
- データ挿入時のJSON形式検証。
- より効率的なバイナリストレージ形式により、
get_json_xx関数と比較してjson_extractなどの関数を使用したJSON内部フィールドへの高速アクセスが可能。
注意: バージョン1.2.xでは、JSON型はJSONBという名前でした。MySQLとの互換性を維持するため、バージョン2.0.0からJSONに名前が変更されました。古いTableでは以前の名前を引き続き使用できます。
構文
定義：
json_column_name JSON
挿入:
INSERT INTO VALUESを使用し、フォーマットをクォートで囲まれた文字列として指定します。例：
INSERT INTO table_name(id, json_column_name) VALUES (1, '{"k1": "100"}')
- STREAM LOADの場合、対応する列の形式は追加の引用符を含まない文字列です。例：
12 {"k1":"v31", "k2": 300}
13 []
14 [123, 456]
- JSON内で
'\'を含む特殊文字（'\r'、'\t'など）が出現する場合、replace関数を使用して"\"を"\\"に置換する必要があります。例えば、"\n"を"\\n"に置換する必要があります。
Query:
- JSON列全体を直接選択する：
SELECT json_column_name FROM table_name;
- JSON関数を使用してJSONから特定のフィールドやその他の情報を抽出します。例：
SELECT json_extract(json_column_name, '$.k1') FROM table_name;
JSON型は整数、文字列、BOOLEAN、ARRAY、MAPとの間でキャストできます。例えば：
SELECT CAST('{"k1": "100"}' AS JSON);
SELECT CAST(json_column_name AS STRING) FROM table_name;
SELECT CAST(json_extract(json_column_name, '$.k1') AS INT) FROM table_name;
ヒント
JSON型は現在、GROUP BY、ORDER BY、または比較演算には使用できません。
example
Table作成、データロード、クエリを含むJSONデータ型のチュートリアル。
データベースとTableの作成
CREATE DATABASE testdb;
USE testdb;
CREATE TABLE test_json (
id INT,
j JSON
)
DUPLICATE KEY(id)
DISTRIBUTED BY HASH(id) BUCKETS 10
PROPERTIES("replication_num" = "1");
データの読み込み
stream load test_json.csv テストデータ
- 2つの列があり、1列目はidで2列目はjson文字列です
- 25行あり、最初の18行は有効なjsonで最後の7行は無効です
1 \N
2 null
3 true
4 false
5 100
6 10000
7 1000000000
8 1152921504606846976
9 6.18
10 "abcd"
11 {}
12 {"k1":"v31", "k2": 300}
13 []
14 [123, 456]
15 ["abc", "def"]
16 [null, true, false, 100, 6.18, "abc"]
17 [{"k1":"v41", "k2": 400}, 1, "a", 3.14]
18 {"k1":"v31", "k2": 300, "a1": [{"k1":"v41", "k2": 400}, 1, "a", 3.14]}
19 ''
20 'abc'
21 abc
22 100x
23 6.a8
24 {x
25 [123, abc]
- 28%の行が無効なため、デフォルト設定でのstream loadは「too many filtered rows」というエラーメッセージで失敗します
curl --location-trusted -u root: -T test_json.csv http://127.0.0.1:8840/api/testdb/test_json/_stream_load
{
"TxnId": 12019,
"Label": "744d9821-9c9f-43dc-bf3b-7ab048f14e32",
"TwoPhaseCommit": "false",
"Status": "Fail",
"Message": "too many filtered rows",
"NumberTotalRows": 25,
"NumberLoadedRows": 18,
"NumberFilteredRows": 7,
"NumberUnselectedRows": 0,
"LoadBytes": 380,
"LoadTimeMs": 48,
"BeginTxnTimeMs": 0,
"StreamLoadPutTimeMs": 1,
"ReadDataTimeMs": 0,
"WriteDataTimeMs": 45,
"CommitAndPublishTimeMs": 0,
"ErrorURL": "http://172.21.0.5:8840/api/_load_error_log?file=__shard_2/error_log_insert_stmt_95435c4bf5f156df-426735082a9296af_95435c4bf5f156df_426735082a9296af"
}
- header設定'max_filter_ratio: 0.3'を設定した後、stream loadは成功します
curl --location-trusted -u root: -H 'max_filter_ratio: 0.3' -T test_json.csv http://127.0.0.1:8840/api/testdb/test_json/_stream_load
{
"TxnId": 12017,
"Label": "f37a50c1-43e9-4f4e-a159-a3db6abe2579",
"TwoPhaseCommit": "false",
"Status": "Success",
"Message": "OK",
"NumberTotalRows": 25,
"NumberLoadedRows": 18,
"NumberFilteredRows": 7,
"NumberUnselectedRows": 0,
"LoadBytes": 380,
"LoadTimeMs": 68,
"BeginTxnTimeMs": 0,
"StreamLoadPutTimeMs": 2,
"ReadDataTimeMs": 0,
"WriteDataTimeMs": 45,
"CommitAndPublishTimeMs": 19,
"ErrorURL": "http://172.21.0.5:8840/api/_load_error_log?file=__shard_0/error_log_insert_stmt_a1463f98a7b15caf-c79399b920f5bfa3_a1463f98a7b15caf_c79399b920f5bfa3"
}
- stream loadによって読み込まれたデータを表示するにはSELECTを使用します。JSON型の列はプレーンなJSON文字列として表示されます。
mysql> SELECT * FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+
| id | j |
+------+---------------------------------------------------------------+
| 1 | NULL |
| 2 | null |
| 3 | true |
| 4 | false |
| 5 | 100 |
| 6 | 10000 |
| 7 | 1000000000 |
| 8 | 1152921504606846976 |
| 9 | 6.18 |
| 10 | "abcd" |
| 11 | {} |
| 12 | {"k1":"v31","k2":300} |
| 13 | [] |
| 14 | [123,456] |
| 15 | ["abc","def"] |
| 16 | [null,true,false,100,6.18,"abc"] |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} |
+------+---------------------------------------------------------------+
18 rows in set (0.03 sec)
insert into を使用したデータの書き込み
- 1行を insert した後、総行数が18から19に増加
mysql> INSERT INTO test_json VALUES(26, '{"k1":"v1", "k2": 200}');
Query OK, 1 row affected (0.09 sec)
{'label':'insert_4ece6769d1b42fd_ac9f25b3b8f3dc02', 'status':'VISIBLE', 'txnId':'12016'}
mysql> SELECT * FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+
| id | j |
+------+---------------------------------------------------------------+
| 1 | NULL |
| 2 | null |
| 3 | true |
| 4 | false |
| 5 | 100 |
| 6 | 10000 |
| 7 | 1000000000 |
| 8 | 1152921504606846976 |
| 9 | 6.18 |
| 10 | "abcd" |
| 11 | {} |
| 12 | {"k1":"v31","k2":300} |
| 13 | [] |
| 14 | [123,456] |
| 15 | ["abc","def"] |
| 16 | [null,true,false,100,6.18,"abc"] |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} |
| 26 | {"k1":"v1","k2":200} |
+------+---------------------------------------------------------------+
19 rows in set (0.03 sec)
Query
json_extract関数を使用してjsonから一部のフィールドを抽出する
- json全体を抽出する場合、'$'はjsonパスのルートを表す
+------+---------------------------------------------------------------+---------------------------------------------------------------+
| id | j | json_extract(`j`, '$') |
+------+---------------------------------------------------------------+---------------------------------------------------------------+
| 1 | NULL | NULL |
| 2 | null | null |
| 3 | true | true |
| 4 | false | false |
| 5 | 100 | 100 |
| 6 | 10000 | 10000 |
| 7 | 1000000000 | 1000000000 |
| 8 | 1152921504606846976 | 1152921504606846976 |
| 9 | 6.18 | 6.18 |
| 10 | "abcd" | "abcd" |
| 11 | {} | {} |
| 12 | {"k1":"v31","k2":300} | {"k1":"v31","k2":300} |
| 13 | [] | [] |
| 14 | [123,456] | [123,456] |
| 15 | ["abc","def"] | ["abc","def"] |
| 16 | [null,true,false,100,6.18,"abc"] | [null,true,false,100,6.18,"abc"] |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | [{"k1":"v41","k2":400},1,"a",3.14] |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} |
| 26 | {"k1":"v1","k2":200} | {"k1":"v1","k2":200} |
+------+---------------------------------------------------------------+---------------------------------------------------------------+
19 rows in set (0.03 sec)
- k1フィールドを抽出し、存在しない場合はNULLを返す
mysql> SELECT id, j, json_extract(j, '$.k1') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+----------------------------+
| id | j | json_extract(`j`, '$.k1') |
+------+---------------------------------------------------------------+----------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | "v31" |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | "v31" |
| 26 | {"k1":"v1","k2":200} | "v1" |
+------+---------------------------------------------------------------+----------------------------+
19 rows in set (0.03 sec)
- トップレベル配列の要素0を抽出する
mysql> SELECT id, j, json_extract(j, '$[0]') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+----------------------------+
| id | j | json_extract(`j`, '$[0]') |
+------+---------------------------------------------------------------+----------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | NULL |
| 13 | [] | NULL |
| 14 | [123,456] | 123 |
| 15 | ["abc","def"] | "abc" |
| 16 | [null,true,false,100,6.18,"abc"] | null |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | {"k1":"v41","k2":400} |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | NULL |
| 26 | {"k1":"v1","k2":200} | NULL |
+------+---------------------------------------------------------------+----------------------------+
19 rows in set (0.03 sec)
- 名前a1のjson配列全体を抽出する
mysql> SELECT id, j, json_extract(j, '$.a1') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+------------------------------------+
| id | j | json_extract(`j`, '$.a1') |
+------+---------------------------------------------------------------+------------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | NULL |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | [{"k1":"v41","k2":400},1,"a",3.14] |
| 26 | {"k1":"v1","k2":200} | NULL |
+------+---------------------------------------------------------------+------------------------------------+
19 rows in set (0.02 sec)
- 配列内のオブジェクトからネストされたフィールドを抽出する
mysql> SELECT id, j, json_extract(j, '$.a1[0]'), json_extract(j, '$.a1[0].k1') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+-------------------------------+----------------------------------+
| id | j | json_extract(`j`, '$.a1[0]') | json_extract(`j`, '$.a1[0].k1') |
+------+---------------------------------------------------------------+-------------------------------+----------------------------------+
| 1 | NULL | NULL | NULL |
| 2 | null | NULL | NULL |
| 3 | true | NULL | NULL |
| 4 | false | NULL | NULL |
| 5 | 100 | NULL | NULL |
| 6 | 10000 | NULL | NULL |
| 7 | 1000000000 | NULL | NULL |
| 8 | 1152921504606846976 | NULL | NULL |
| 9 | 6.18 | NULL | NULL |
| 10 | "abcd" | NULL | NULL |
| 11 | {} | NULL | NULL |
| 12 | {"k1":"v31","k2":300} | NULL | NULL |
| 13 | [] | NULL | NULL |
| 14 | [123,456] | NULL | NULL |
| 15 | ["abc","def"] | NULL | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | {"k1":"v41","k2":400} | "v41" |
| 26 | {"k1":"v1","k2":200} | NULL | NULL |
+------+---------------------------------------------------------------+-------------------------------+----------------------------------+
19 rows in set (0.02 sec)
- 特定のデータ型を持つフィールドを抽出
- json_extract_string は文字列型のフィールドを抽出し、フィールドが文字列でない場合は文字列に変換します
mysql> SELECT id, j, json_extract_string(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+---------------------------------------------------------------+
| id | j | json_extract_string(`j`, '$') |
+------+---------------------------------------------------------------+---------------------------------------------------------------+
| 1 | NULL | NULL |
| 2 | null | null |
| 3 | true | true |
| 4 | false | false |
| 5 | 100 | 100 |
| 6 | 10000 | 10000 |
| 7 | 1000000000 | 1000000000 |
| 8 | 1152921504606846976 | 1152921504606846976 |
| 9 | 6.18 | 6.18 |
| 10 | "abcd" | abcd |
| 11 | {} | {} |
| 12 | {"k1":"v31","k2":300} | {"k1":"v31","k2":300} |
| 13 | [] | [] |
| 14 | [123,456] | [123,456] |
| 15 | ["abc","def"] | ["abc","def"] |
| 16 | [null,true,false,100,6.18,"abc"] | [null,true,false,100,6.18,"abc"] |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | [{"k1":"v41","k2":400},1,"a",3.14] |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} |
| 26 | {"k1":"v1","k2":200} | {"k1":"v1","k2":200} |
+------+---------------------------------------------------------------+---------------------------------------------------------------+
19 rows in set (0.02 sec)
mysql> SELECT id, j, json_extract_string(j, '$.k1') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+-----------------------------------+
| id | j | json_extract_string(`j`, '$.k1') |
+------+---------------------------------------------------------------+-----------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | v31 |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | v31 |
| 26 | {"k1":"v1","k2":200} | v1 |
+------+---------------------------------------------------------------+-----------------------------------+
19 rows in set (0.03 sec)
- json_extract_intはint型のフィールドを抽出し、フィールドがintでない場合はNULLを返します
mysql> SELECT id, j, json_extract_int(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+-----------------------------+
| id | j | json_extract_int(`j`, '$') |
+------+---------------------------------------------------------------+-----------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | 100 |
| 6 | 10000 | 10000 |
| 7 | 1000000000 | 1000000000 |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | NULL |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | NULL |
| 26 | {"k1":"v1","k2":200} | NULL |
+------+---------------------------------------------------------------+-----------------------------+
19 rows in set (0.02 sec)
mysql> SELECT id, j, json_extract_int(j, '$.k2') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+--------------------------------+
| id | j | json_extract_int(`j`, '$.k2') |
+------+---------------------------------------------------------------+--------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | 300 |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | 300 |
| 26 | {"k1":"v1","k2":200} | 200 |
+------+---------------------------------------------------------------+--------------------------------+
19 rows in set (0.03 sec)
- json_extract_bigint はbigint型のフィールドを抽出し、フィールドがbigintでない場合はNULLを返します
mysql> SELECT id, j, json_extract_bigint(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+--------------------------------+
| id | j | json_extract_bigint(`j`, '$') |
+------+---------------------------------------------------------------+--------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | 100 |
| 6 | 10000 | 10000 |
| 7 | 1000000000 | 1000000000 |
| 8 | 1152921504606846976 | 1152921504606846976 |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | NULL |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | NULL |
| 26 | {"k1":"v1","k2":200} | NULL |
+------+---------------------------------------------------------------+--------------------------------+
19 rows in set (0.03 sec)
mysql> SELECT id, j, json_extract_bigint(j, '$.k2') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+-----------------------------------+
| id | j | json_extract_bigint(`j`, '$.k2') |
+------+---------------------------------------------------------------+-----------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | 300 |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | 300 |
| 26 | {"k1":"v1","k2":200} | 200 |
+------+---------------------------------------------------------------+-----------------------------------+
19 rows in set (0.02 sec)
- json_extract_double は double 型のフィールドを抽出し、フィールドが double でない場合は NULL を返します
mysql> SELECT id, j, json_extract_double(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+--------------------------------+
| id | j | json_extract_double(`j`, '$') |
+------+---------------------------------------------------------------+--------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | 100 |
| 6 | 10000 | 10000 |
| 7 | 1000000000 | 1000000000 |
| 8 | 1152921504606846976 | 1.152921504606847e+18 |
| 9 | 6.18 | 6.18 |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | NULL |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | NULL |
| 26 | {"k1":"v1","k2":200} | NULL |
+------+---------------------------------------------------------------+--------------------------------+
19 rows in set (0.02 sec)
mysql> SELECT id, j, json_extract_double(j, '$.k2') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+-----------------------------------+
| id | j | json_extract_double(`j`, '$.k2') |
+------+---------------------------------------------------------------+-----------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | 300 |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | 300 |
| 26 | {"k1":"v1","k2":200} | 200 |
+------+---------------------------------------------------------------+-----------------------------------+
19 rows in set (0.03 sec)
- json_extract_boolはboolean型のフィールドを抽出し、そのフィールドがbooleanでない場合はNULLを返します
mysql> SELECT id, j, json_extract_bool(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+------------------------------+
| id | j | json_extract_bool(`j`, '$') |
+------+---------------------------------------------------------------+------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | 1 |
| 4 | false | 0 |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | NULL |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | NULL |
| 26 | {"k1":"v1","k2":200} | NULL |
+------+---------------------------------------------------------------+------------------------------+
19 rows in set (0.01 sec)
mysql> SELECT id, j, json_extract_bool(j, '$[1]') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+---------------------------------+
| id | j | json_extract_bool(`j`, '$[1]') |
+------+---------------------------------------------------------------+---------------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | NULL |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | 1 |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | NULL |
| 26 | {"k1":"v1","k2":200} | NULL |
+------+---------------------------------------------------------------+---------------------------------+
19 rows in set (0.01 sec)
- json_extract_isnullはjson null型のフィールドを抽出し、フィールドがjson nullの場合は1を返し、それ以外の場合は0を返します
- json nullはSQL NULLとは異なります。SQL NULLはフィールドに値が存在しないことを表しますが、json nullは特別な値nullを持つフィールドを表します。
mysql> SELECT id, j, json_extract_isnull(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+--------------------------------+
| id | j | json_extract_isnull(`j`, '$') |
+------+---------------------------------------------------------------+--------------------------------+
| 1 | NULL | NULL |
| 2 | null | 1 |
| 3 | true | 0 |
| 4 | false | 0 |
| 5 | 100 | 0 |
| 6 | 10000 | 0 |
| 7 | 1000000000 | 0 |
| 8 | 1152921504606846976 | 0 |
| 9 | 6.18 | 0 |
| 10 | "abcd" | 0 |
| 11 | {} | 0 |
| 12 | {"k1":"v31","k2":300} | 0 |
| 13 | [] | 0 |
| 14 | [123,456] | 0 |
| 15 | ["abc","def"] | 0 |
| 16 | [null,true,false,100,6.18,"abc"] | 0 |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | 0 |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | 0 |
| 26 | {"k1":"v1","k2":200} | 0 |
+------+---------------------------------------------------------------+--------------------------------+
19 rows in set (0.03 sec)
json_exists_pathでJSONにフィールドが存在するかチェックする
mysql> SELECT id, j, json_exists_path(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+-----------------------------+
| id | j | json_exists_path(`j`, '$') |
+------+---------------------------------------------------------------+-----------------------------+
| 1 | NULL | NULL |
| 2 | null | 1 |
| 3 | true | 1 |
| 4 | false | 1 |
| 5 | 100 | 1 |
| 6 | 10000 | 1 |
| 7 | 1000000000 | 1 |
| 8 | 1152921504606846976 | 1 |
| 9 | 6.18 | 1 |
| 10 | "abcd" | 1 |
| 11 | {} | 1 |
| 12 | {"k1":"v31","k2":300} | 1 |
| 13 | [] | 1 |
| 14 | [123,456] | 1 |
| 15 | ["abc","def"] | 1 |
| 16 | [null,true,false,100,6.18,"abc"] | 1 |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | 1 |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | 1 |
| 26 | {"k1":"v1","k2":200} | 1 |
+------+---------------------------------------------------------------+-----------------------------+
19 rows in set (0.02 sec)
mysql> SELECT id, j, json_exists_path(j, '$.k1') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+--------------------------------+
| id | j | json_exists_path(`j`, '$.k1') |
+------+---------------------------------------------------------------+--------------------------------+
| 1 | NULL | NULL |
| 2 | null | 0 |
| 3 | true | 0 |
| 4 | false | 0 |
| 5 | 100 | 0 |
| 6 | 10000 | 0 |
| 7 | 1000000000 | 0 |
| 8 | 1152921504606846976 | 0 |
| 9 | 6.18 | 0 |
| 10 | "abcd" | 0 |
| 11 | {} | 0 |
| 12 | {"k1":"v31","k2":300} | 1 |
| 13 | [] | 0 |
| 14 | [123,456] | 0 |
| 15 | ["abc","def"] | 0 |
| 16 | [null,true,false,100,6.18,"abc"] | 0 |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | 0 |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | 1 |
| 26 | {"k1":"v1","k2":200} | 1 |
+------+---------------------------------------------------------------+--------------------------------+
19 rows in set (0.03 sec)
mysql> SELECT id, j, json_exists_path(j, '$[2]') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+--------------------------------+
| id | j | json_exists_path(`j`, '$[2]') |
+------+---------------------------------------------------------------+--------------------------------+
| 1 | NULL | NULL |
| 2 | null | 0 |
| 3 | true | 0 |
| 4 | false | 0 |
| 5 | 100 | 0 |
| 6 | 10000 | 0 |
| 7 | 1000000000 | 0 |
| 8 | 1152921504606846976 | 0 |
| 9 | 6.18 | 0 |
| 10 | "abcd" | 0 |
| 11 | {} | 0 |
| 12 | {"k1":"v31","k2":300} | 0 |
| 13 | [] | 0 |
| 14 | [123,456] | 0 |
| 15 | ["abc","def"] | 0 |
| 16 | [null,true,false,100,6.18,"abc"] | 1 |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | 1 |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | 0 |
| 26 | {"k1":"v1","k2":200} | 0 |
+------+---------------------------------------------------------------+--------------------------------+
19 rows in set (0.02 sec)
json_type による JSON フィールドのデータ型取得
- json path で指定されたフィールドのデータ型を返します。存在しない場合は NULL を返します。
mysql> SELECT id, j, json_type(j, '$') FROM test_json ORDER BY id;
+------+---------------------------------------------------------------+----------------------+
| id | j | json_type(`j`, '$') |
+------+---------------------------------------------------------------+----------------------+
| 1 | NULL | NULL |
| 2 | null | null |
| 3 | true | bool |
| 4 | false | bool |
| 5 | 100 | int |
| 6 | 10000 | int |
| 7 | 1000000000 | int |
| 8 | 1152921504606846976 | bigint |
| 9 | 6.18 | double |
| 10 | "abcd" | string |
| 11 | {} | object |
| 12 | {"k1":"v31","k2":300} | object |
| 13 | [] | array |
| 14 | [123,456] | array |
| 15 | ["abc","def"] | array |
| 16 | [null,true,false,100,6.18,"abc"] | array |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | array |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | object |
| 26 | {"k1":"v1","k2":200} | object |
+------+---------------------------------------------------------------+----------------------+
19 rows in set (0.02 sec)
mysql> select id, j, json_type(j, '$.k1') from test_json order by id;
+------+---------------------------------------------------------------+-------------------------+
| id | j | json_type(`j`, '$.k1') |
+------+---------------------------------------------------------------+-------------------------+
| 1 | NULL | NULL |
| 2 | null | NULL |
| 3 | true | NULL |
| 4 | false | NULL |
| 5 | 100 | NULL |
| 6 | 10000 | NULL |
| 7 | 1000000000 | NULL |
| 8 | 1152921504606846976 | NULL |
| 9 | 6.18 | NULL |
| 10 | "abcd" | NULL |
| 11 | {} | NULL |
| 12 | {"k1":"v31","k2":300} | string |
| 13 | [] | NULL |
| 14 | [123,456] | NULL |
| 15 | ["abc","def"] | NULL |
| 16 | [null,true,false,100,6.18,"abc"] | NULL |
| 17 | [{"k1":"v41","k2":400},1,"a",3.14] | NULL |
| 18 | {"k1":"v31","k2":300,"a1":[{"k1":"v41","k2":400},1,"a",3.14]} | string |
| 26 | {"k1":"v1","k2":200} | string |
+------+---------------------------------------------------------------+-------------------------+
19 rows in set (0.03 sec)
FAQ
- JSONのnullとSQLのNULL（つまり、IS NULL）に違いはありますか？
はい、違いがあります。JSONでは、null（例：{"key1": null}）は、キーkey1が存在し、その値が明示的にnullであることを意味します。これはJSONバイナリにエンコードされる特別な型です。
一方、SQL NULL（IS NULLを使用する場合）は、そのキーがJSONオブジェクト内に全く存在しないことを示す場合があります。
例：
mysql> SELECT JSON_EXTRACT_STRING('{"key1" : null}', "$.key1") IS NULL;
+----------------------------------------------------------+
| JSON_EXTRACT_STRING('{"key1" : null}', "$.key1") IS NULL |
+----------------------------------------------------------+
| 0 |
+----------------------------------------------------------+
1 row in set (0.00 sec)
mysql> SELECT JSON_EXTRACT_STRING('{"key1" : null}', "$.key_not_exist") IS NULL;
+-------------------------------------------------------------------+
| JSON_EXTRACT_STRING('{"key1" : null}', "$.key_not_exist") IS NULL |
+-------------------------------------------------------------------+
| 1 |
+-------------------------------------------------------------------+
1 row in set (0.01 sec)
GET_JSON_XXXとJSON_EXTRACT_XXX関数の違いは何ですか。どちらを選択すべきでしょうか。
GET_JSON_XXX関数は文字列型での使用を想定して設計されており、生のJSON文字列から直接値を抽出します。一方、JSON_EXTRACT_XXX関数はJSONデータ型専用に実装されており、そのために最適化されています。
keywords
JSONB, JSON, json_parse, json_parse_error_to_null, json_parse_error_to_value, json_extract, json_extract_isnull, json_extract_bool, json_extract_int, json_extract_bigint, json_extract_double, json_extract_string, json_exists_path, json_type