SEZ9 commented on PR #12355:
URL: https://github.com/apache/seatunnel/pull/12355#issuecomment-5806330422

   Thanks @wgzhao - and apologies for the delay, this fell off my radar. Happy 
to take another look at `4f6663a9f` now.
   
   Before I do, a few quick things so I can close each of the five points 
cleanly:
   
   1. **Length leaking into `columnLength` (the first point)** - you mention 
the `SET`/`ENUM` length is now derived from the option list in 
`MySqlTypeUtils`. Can you confirm the rule you landed on (e.g. `ENUM` = length 
of the longest option, `SET` = sum of option lengths plus `n-1` separators), 
and whether it works off the raw quoted option text or the unquoted values? 
Escaped quotes / multi-byte options would give different answers depending on 
which one is used, so I want to check that against the test expectations.
   2. **Per-item breakdown** - you reference a reply "just above" with the 
item-by-item breakdown, but I don't see it in the thread on my side. Could you 
re-post or link it? It would save me guessing which hunk maps to which point.
   3. **Test coverage (points 3 and 4)** - the new shapes you list 
(multi-character, embedded comma, escaped quote, `CHARACTER SET`) cover the 
verbatim-join cases. Do the tests also assert `columnLength` (not only 
`getSourceType()`) for those shapes, and is the single-option `SET` case 
included? Those were the two remaining gaps I had flagged.
   4. **Javadocs (points 2 and 5)** - just confirm the test Javadoc now 
describes the pre-fix value as `SET(5)` and refers to `Column#getSourceType()` 
on the SeaTunnel `Column`, and that the override's Javadoc uses a qualified 
reference so it doesn't resolve to the Debezium `Column`.
   
   If those all check out I don't expect anything further on the code side, and 
I'll leave a review on the head accordingly. No need to touch the branch in the 
meantime.
   
   <!-- streview-comment:1268 -->


-- 
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]

Reply via email to