andygrove commented on code in PR #5037:
URL: https://github.com/apache/datafusion-comet/pull/5037#discussion_r3661103796


##########
docs/source/user-guide/latest/expressions.md:
##########
@@ -562,7 +562,7 @@ expression-level). The `outer` variants are wired but 
marked `Incompatible`; the
 | `contains` | ✅ | — |  |
 | `decode` | ✅ | — |  |
 | `elt` | ✅ | Codegen dispatch |  |
-| `encode` | 🔜 | — | Lowers to `StaticInvoke(encode)` (not allowlisted); falls 
back |
+| `encode` | ✅ | — |  |

Review Comment:
   Verified by regeneration rather than by hand. I ran `GenerateDocs` (the same 
invocation `dev/generate-release-docs.sh` uses) against a scratch copy of 
`docs/source/user-guide/latest` under three profiles:
   
   | profile | generated `encode` cell |
   | --- | --- |
   | spark-3.5 | `Codegen dispatch` |
   | spark-4.0 | `—` |
   | spark-4.1 | `—` |
   
   The committed cell matches the 4.x generation exactly, so the file is 
unchanged. The 3.5 run is also a useful independent check that `CometEncode` 
really is keyed on `classOf[Encode]` in that profile.
   
   One thing the run surfaced that is out of scope here: the tracked file has 
pre-existing drift on unrelated rows. Under 4.0 and 4.1 the generator emits 
`Native` for `<<`, `>>`, `>>>` (committed `—`) and `—` for `to_char`, 
`to_varchar` (committed `Codegen dispatch`); 4.1 additionally differs on 
`hour`, `minute`, `second`, `make_timestamp`. Since the published site 
regenerates into a temp tree at build time, this only affects the in-repo copy. 
I left it alone rather than adding unrelated churn to this PR, but it may be 
worth a separate cleanup, or a CI check that the committed column matches 
generation.



-- 
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]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to