SEZ9 commented on PR #12730:
URL: https://github.com/apache/seatunnel/pull/12730#issuecomment-6136890799
Thanks @zhangshenghang — the summary for `2433602e` matches what I raised,
and I'll verify each point against the diff before resolving:
- LocalFile `file_format_type` back to required (en/zh)
- `schema_save_mode` / `data_save_mode` rows removed from BosFile, CosFile,
ObsFile and OssJindoFile, plus the BosFile `schema_evolution_enabled` row
- `UInt8 / UInt16` and `Int8 / Int16` restored in the Clickhouse sink INT row
- `STAIN_TRACE` row added to the EventType table; no objection to the
`stain-trace-sample-rate` key change as part of that
A few open questions:
1. Clickhouse: did the BIGINT row (`UInt32 / UInt64 / Interval*`) change on
this head too, or is it intentionally kept as a write-compatibility list? If
the latter, a short note or a split between DDL-generation and
write-compatibility mappings would make the two semantics explicit. The source
table also still omits Bool/UUID/IPv4/IPv6.
2. Your comment appears cut off at "will take these in a f" — I assume
"follow-up". Deferring the ObsFile STS section rewrite is fine with me; please
confirm and drop a link once it exists.
3. Two earlier points I didn't see mentioned — handled or also deferred?
- OssFile source: `enable_file_split` / `file_split_size` documented but
not declared in `OssFileSourceFactory.optionRule()`.
- ObsFile/BosFile sink: newly documented rows (`tmp_path`, `encoding`,
`csv_string_quote_mode`, `enable_header_write`, excel/CDC options) not in the
respective `*FileSinkFactory.optionRule()`.
Happy to take another pass once those are clarified.
<!-- streview-comment:1705 -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
