[hotfix-#1853] [connector jdbc] jdbc toInternal methods should process data by field name rather than index #1853

[david-gao1]

Purpose of this pull request

Which issue you fix

Fixes # (issue).

Checklist:

• I have executed the 'mvn spotless:apply' command to format my code.
• I have a meaningful commit message (including the issue id, the template of commit message is '[label-type-#issue-id][fixed-module] a meaningful commit message.')
• I have performed a self-review of my own code.
• I have commented my code, particularly in hard-to-understand areas.
• I have made corresponding changes to the documentation.
• I have added tests that prove my fix is effective or that my feature works.
• New and existing unit tests pass locally with my changes.
• I have checked my code and corrected any misspellings.
• My commit is only one. (If there are multiple commits, you can use 'git squash' to compress multiple commits into one.)

[zoudaokoulife]

代码上看是合理的，不过涉及到核心功能的改动。我们本地先验证下数据同步 相关逻辑是否有影响之后再合并

[david-gao1]

好的

…

---Original--- From: ***@***.***> Date: Thu, Dec 14, 2023 19:54 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [DTStack/chunjun] [hotfix-#1853] [connector jdbc] jdbc toInternalmethods should process data by field name rather than index #1853 (PR #1854) 代码上看是合理的，不过涉及到核心功能的改动。我们本地先验证下数据同步 相关逻辑是否有影响之后再合并 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[zoudaokoulife]

@david-gao1
从上下文结构上看，问题的根本原因是在 构建SupportsProjectionPushDown 的情况下，使用了剪枝导致的。
上面提供的方式是能解决问题。但是从结构上看通过在查询的时候 使用剪枝后的column 会更合适，减少了查询的column 。
建议在com.dtstack.chunjun.connector.jdbc.source.JdbcDynamicTableSource#getScanRuntimeProvider 调整 columns 的构造逻辑。
final RowType rowType = (RowType) physicalRowDataType.getLogicalType(); TypeInformation<RowData> typeInformation = InternalTypeInfo.of(rowType); ResolvedSchema projectionSchema = ResolvedSchema.physical(rowType.getFieldNames(), physicalRowDataType.getChildren()) List<Column> columns = projectionSchema.getColumns()
你这边是否能更新一个版本呢？

[david-gao1]

好的 感谢提醒 我研究下

…

---Original--- From: ***@***.***&gt; Date: Mon, Dec 18, 2023 10:49 AM To: ***@***.***&gt;; Cc: ***@***.******@***.***&gt;; Subject: Re: [DTStack/chunjun] [hotfix-#1853] [connector jdbc] jdbc toInternalmethods should process data by field name rather than index #1853 (PR #1854) @david-gao1 从上下文结构上看，问题的根本原因是在 构建SupportsProjectionPushDown 的情况下，使用了剪枝导致的。 上面提供的方式是能解决问题。但是从结构上看通过在查询的时候 使用剪枝后的column 会更合适，减少了查询的column 。 建议在com.dtstack.chunjun.connector.jdbc.source.JdbcDynamicTableSource#getScanRuntimeProvider 调整 columns 的构造逻辑。 final RowType rowType = (RowType) physicalRowDataType.getLogicalType(); TypeInformation<RowData&gt; typeInformation = InternalTypeInfo.of(rowType); ResolvedSchema projectionSchema = ResolvedSchema.physical(rowType.getFieldNames(), physicalRowDataType.getChildren()) List<Column&gt; columns = projectionSchema.getColumns() 你这边是否能更新一个版本呢？ — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: ***@***.***&gt;