support PARSE_JSON / TRY_PARSE_JSON function

[b41sh]

I hereby agree to the terms of the CLA available at: https://databend.rs/dev/policies/cla/

Summary

Support PARSE_JSON function
Support TRY_PARSE_JSON function

Changelog

• New Feature
• Build/Testing/CI
• Need Documentation (Need to add documentation to https://databend.rs)

Related Issues

Fixes #4476

Test Plan

Unit Tests

Stateless Tests

[vercel]

This pull request is being automatically deployed with Vercel (learn more).
To see the status of your deployment, click below or on the icon next to each commit.

🔍 Inspect: https://vercel.com/databend/databend/q6ch5RiXBiEcBb1CusQaDyJMq7Hs
✅ Preview: https://databend-git-fork-b41sh-func-parse-json-databend.vercel.app

[mergify]

Thanks for the contribution!
I have applied any labels matching special text in your PR Changelog.

Please review the labels and make any necessary changes.

[sundy-li]

If SUPPRESS_PARSE_ERROR is false, the return type should not be nullable.

[b41sh]

parse_json may also yield null, for example:

mysql> select parse_json('');
+----------------+
| parse_json('') |
+----------------+
| NULL           |
+----------------+
1 row in set (0.01 sec)

[sundy-li]

So what's the difference between JsonValue::Null and DataValue::Null ?

[b41sh]

DataValue::Null is a SQL null, and JsonValue::Null is a JSON null. Detailed description can refer to this document:
https://docs.snowflake.com/en/sql-reference/functions/parse_json.html#usage-notes

mysql> select parse_json(null), parse_json(''), parse_json('null');
+------------------+----------------+--------------------+
| parse_json(NULL) | parse_json('') | parse_json('null') |
+------------------+----------------+--------------------+
|             NULL | NULL           | null               |
+------------------+----------------+--------------------+
1 row in set (0.02 sec)

[sundy-li]

So If SUPPRESS_PARSE_ERROR is false, it's JsonColumn not Nullable(JsonColumn).

[b41sh]

if SUPPRESS_PARSE_ERROR is false, it should also return Nullable(JsonColumn). Because if the input value is an empty string, it should return a SQL null.

[sundy-li]

So the right way is:

1. parse_json -> SUPPRESS_PARSE_ERROR true -> JsonColumn
2. try_parse_json -> SUPPRESS_PARSE_ERROR false -> Nullable(JsonColumn)