[python] support table scan with row range #6944
Merged
Conversation
This file contains hidden or 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
universe-hcy
force-pushed
the
paimon_ali
branch
2 times, most recently
from
9ab6fee to
12df60e
Compare
JingsongLi
reviewed
Jan 4, 2026
| (self.idx_of_this_subtask - remainder) * base_rows_per_shard) | ||
|
|
||
| end_row = start_row + num_row | ||
| def with_row_range(self, start_row, end_row) -> 'FullStartingScanner': |
Contributor
There was a problem hiding this comment.
Can we raise a exception for primary key table?
Contributor
There was a problem hiding this comment.
Can we use with_row_range and with_shard at the same time or not?
Contributor
Author
There was a problem hiding this comment.
not, It is difficult for all users to have a completely consistent behavior when using at the same time. Added an exception to avoid this situation.
| self.starting_scanner.with_shard(idx_of_this_subtask, number_of_para_subtasks) | ||
| return self | ||
|
|
||
| def with_row_range(self, start_row, end_row) -> 'TableScan': |
Contributor
There was a problem hiding this comment.
Are we returning this exact number of lines, or can it be approximate? This needs to be specified clearly in the comments.
|
|
||
| if self.idx_of_this_subtask is not None: | ||
| print("self.start_row_of_this_subtask:{}".format(self.start_row_of_this_subtask)) | ||
| if self.start_row_of_this_subtask is not None: |
Contributor
There was a problem hiding this comment.
logger may be better
Contributor
Author
There was a problem hiding this comment.
This is for debugging purposes, I will delete it soon.
universe-hcy
force-pushed
the
paimon_ali
branch
from
12df60e to
f1973d6
Compare
universe-hcy
force-pushed
the
paimon_ali
branch
from
f1973d6 to
59a42d9
Compare
discivigour
reviewed
Jan 4, 2026
| """ | ||
| Filter file entries by row range. The row_id corresponds to the row position of the | ||
| file in all file entries in table scan's partitioned_files. | ||
| """ |
Contributor
There was a problem hiding this comment.
It might be better to clearly describe the inclusion and exclusion of idx within the range in the comments.
Contributor
|
+1 |
jerry-024
added a commit
to jerry-024/paimon
that referenced
this pull request
Jan 6, 2026
* upstream/master: (35 commits) [spark] Spark support vector search (apache#6950) [doc] update Apache Doris document with DLF 3.0 (apache#6954) [variant] Fix reading empty shredded variant via variantAccess (apache#6953) [python] support alterTable (apache#6952) [python] support ray data sink to paimon (apache#6883) [python] Rename to TableScan.withSlice to specific start_pos and end_pos [python] sync to_ray method args with ray data api (apache#6948) [python] light refactor for stats collect (apache#6941) [doc] Update cdc ingestion related docs [rest] Add tagNamePrefix definition for listTagsPaged (apache#6947) [python] support table scan with row range (apache#6944) [spark] Fix EqualNullSafe is not correct when column has null value. (apache#6943) [python] fix value_stats containing system fields for primary key tables (apache#6945) [test][rest] add test case for two sessions with cache for rest commitTable (apache#6438) [python] do not retry for connect exception in rest (apache#6942) [spark] Fix read shredded and unshredded variant both (apache#6936) [python] Let Python write file without value stats by default (apache#6940) [python] ray version compatible (apache#6937) [core] Unify conflict detect in FileStoreCommitImpl (apache#6932) [test] Fix unstable case in CompactActionITCase ...
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
Purpose
In model training and inference, it is common to record unread row intervals to checkpoints for recovery after worker failure. Currently, the with_stard method cannot support this feature and needs to support row_range for table scanning.
Linked issue: close #xxx
Tests
test_data_blob_writer_with_row_range in blob_table_test.py
API and Format
Documentation