Add DataFusion 55.0.0 release blog post - #203
Conversation
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
| partitions based on ordered split points ([#22395], design discussion | ||
| [#21992]). Range partitioning preserves ordering across partitions, which is a | ||
| natural fit for pre-sorted data and for distributed engines that shuffle by | ||
| range. |
There was a problem hiding this comment.
maybe a quick example or link to documented example to show how split points are used 👍
There was a problem hiding this comment.
I can of course make this change, just throwing idea out there first
There was a problem hiding this comment.
@gene-bordegaray -- I am very behind on this post (I have been working on the ALP blog post with @sdf-jkl apache/parquet-site#195)
I now have some time and will likely focus on this over the next day or two
I also added you as a collaborator on my fork, so you should now be able to push directly to this branch -- please feel free to do so
There was a problem hiding this comment.
sweet I will take time this weekend to help with edits here and in general 🙇
There was a problem hiding this comment.
Thank you @gene-bordegaray -- the content about range partitioning was great, but I think a little too detailed for this release announcement. I filed a ticket about writing a self contained blog post
There was a problem hiding this comment.
(I will trim the section down a bit but wanted to capture the content first)
9b4a74b to
0b1d694
Compare
e8e763b to
54d4c5f
Compare
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
This is great, @alamb! Nice work. And congrats to everyone on another awesome release! |
|
I think you should also mention that you can now opt in for better ordering equivalence (for example custom cast) |
|
stuhood
left a comment
There was a problem hiding this comment.
Thanks for doing this! Looks great.
kumarUjjawal
left a comment
There was a problem hiding this comment.
Thank you @alamb
This looks great. And thank you everyone involved in the release.
Co-authored-by: Stu Hood <stuhood@gmail.com>
| the design discussion ([#21992]). Thanks to [@gene-bordegaray], [@saadtajwar], [@peterxcli], [@stuhood], | ||
| [@gmhelmold], [@mattp5657], [@mithuncy], [@JSOD11], [@EdsonPetry], | ||
| [@Rich-T-kid], and [@blinding-pixels] for driving this substantial community | ||
| effort. |
…ex()` (apache#24634) ## Which issue does this PR close? - N/A -- documentation only. ## Rationale for this change While writing a blog post - apache/datafusion-site#203 Which mentiones `file_row_index()` and `input_file_name()` -- I added a simple example for the post and I think it would help to have docs in the main site too. ## What changes are included in this PR? Replaces the `sql_example` with ```sql > COPY (SELECT * from values (100), (200), (300)) to '/tmp/foo.parquet'; > select *, input_file_name(), file_row_index() from '/tmp/foo.parquet'; +---------+-------------------+------------------+ | column1 | input_file_name() | file_row_index() | +---------+-------------------+------------------+ | 100 | tmp/foo.parquet | 0 | | 200 | tmp/foo.parquet | 1 | | 300 | tmp/foo.parquet | 2 | +---------+-------------------+------------------+ ``` The same example is used for both functions, since showing them together is likely common and I think makes it clear what they do ## Are these changes tested? By CI ## Are there any user-facing changes? Yes, but documentation only -- the rendered examples for `input_file_name()` and `file_row_index()` in the SQL function reference. No API or behavior changes.
Co-authored-by: Qi Zhu <821684824@qq.com>
Co-authored-by: Qi Zhu <821684824@qq.com>
|
I'll plan to merge this on Friday unless anyone else woul like more time to review |
|
|
||
| Built-in table providers do not yet implement the hook, but custom | ||
| [`TableProvider`][TableProvider] implementations (such as lakehouse table formats) can now plug | ||
| in their own merge execution. |
There was a problem hiding this comment.
@alamb Thanks for highlighting MERGE INTO. One small note: 55.0 still rejects two target-alias edge cases—target-correlated subqueries and a source qualifier matching the target’s real table name. apache/datafusion#24429 fixes both in the planned 56.0 release with a breaking for the MergeIntoOp construction. Would it be useful to mention this briefly as a known limitation?
There was a problem hiding this comment.
Thanks @wirybeaver that is a good point -- I added a note in 00cc646 that we plan more MERGE INTO work in the following releases
… into datafusion-55-blog
|
ok, here we go. |
| --- | ||
| layout: post | ||
| title: Apache DataFusion 55.0.0 Released | ||
| date: 2026-08-25 |
There was a problem hiding this comment.
🤔 the date will be a few days early, but I guess that is ok
|
Publishing workflow seems to be waiting on a builder: https://github.com/apache/datafusion-site/actions/runs/33159387574 |
Which issue does this PR close?
55.0.0release datafusion#2421655.0.0(Jul / Aug 2026) datafusion#22393Rationale for this change
Post announcing the DataFusion 55.0.0 release, following the model of the 54.0.0 post.
Changes
Rendered Preview: https://alamb.github.io/datafusion-site/blog/2026/08/25/datafusion-55.0.0/