refactor: streamline column-aware row encoding impl #20583
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
BugenZhao
commented
Feb 25, 2025
Comment on lines
-35
to
-37
| const OFFSET8 = 0b01; | ||
| const OFFSET16 = 0b10; | ||
| const OFFSET32 = 0b11; |
There was a problem hiding this comment.
The OFFSET32 here overlaps with OFFSET8 and OFFSET16, for which the bitflag doesn't seem to provide clear semantics. That's why I decide to re-implement with bitfield.
| offsets: Vec<u8>, | ||
| buf: Vec<u8>, | ||
| } | ||
|
|
||
| /// A trait unifying [`ToDatumRef`] and already encoded bytes. |
There was a problem hiding this comment.
To unify encode and encode_slice.
Comment on lines
+138
to
+139
| // Use 0 if there's no data. | ||
| let max_offset = usize_offsets.last().copied().unwrap_or(0); |
There was a problem hiding this comment.
So that we support ser/de for empty rows.
| .reserve_exact(usize_offsets.len() * offset_width.width()); | ||
| for &offset in usize_offsets { | ||
| self.offsets | ||
| .put_uint_le(offset as u64, offset_width.width()); |
There was a problem hiding this comment.
Unifies put_u8, put_u16_le, and put_u32_le.
Comment on lines
+223
to
+227
| /// A view of the encoded bytes, which can be iterated over to get the column id and data. | ||
| /// Used for deserialization. | ||
| // TODO: can we unify this with `RowEncoding`, which is for serialization? | ||
| #[derive(Clone)] | ||
| struct EncodedBytes<'a> { |
There was a problem hiding this comment.
To unify Deserializer and try_drop_invalid_columns.
BugenZhao
requested review from
fuyufjh and
zwang28
and removed request for
a team
8 tasks
BugenZhao
requested review from
stdrc and
hzxa21
and removed request for
stdrc
BugenZhao
force-pushed
the
bz/col-aware-encoding-refactor
branch
from
04f2c7b to
8acd50f
Compare
This was referenced Feb 26, 2025
There was a problem hiding this comment.
PR Overview
This PR refactors the column-aware row encoding implementation to improve clarity, conciseness, and reduce potential errors.
- Replaces the bitflags-based header with a bitfield struct and an enum representing offset width.
- Consolidates encoding logic by introducing a generic Encode trait and refines the serializer/deserializer implementations.
- Updates tests and configuration to match the new encoding scheme.
Reviewed Changes
| File | Description |
|---|---|
| src/common/src/util/value_encoding/column_aware_row_encoding.rs | Refactored row encoding/decoding using a bitfield-based header and generic encoding trait. |
| src/storage/src/row_serde/value_serde.rs | Adjusted tests to validate new behavior for handling dropped columns. |
| src/common/Cargo.toml | Added the bitfield-struct dependency to support new header implementation. |
| src/common/src/util/value_encoding/error.rs | Changed the error message format for invalid flags to use binary representation. |
Copilot reviewed 5 out of 5 changed files in this pull request and generated no comments.
Comments suppressed due to low confidence (1)
src/common/src/util/value_encoding/error.rs:45
- [nitpick] Using binary formatting for the invalid flag may obscure its numeric value. Consider providing additional context or including a decimal representation for improved clarity.
#[error("Invalid flag: {0:b}")]
5bc558d to
55c1865
Compare
BugenZhao
force-pushed
the
bz/col-aware-encoding-refactor
branch
from
55c1865 to
4d6f827
Compare
8 tasks
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
BugenZhao
force-pushed
the
bz/col-aware-encoding-refactor
branch
from
4d6f827 to
fe5cdd0
Compare
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I hereby agree to the terms of the RisingWave Labs, Inc. Contributor License Agreement.
What's changed and what's your intention?
Reimplement column-aware row encoding to make the code more clear, concise and less error-prone. See review comments for explanation of changes.
Checklist
Documentation
Release note