Repository navigation
Mark Vortex ID metadata as base64 - #11300
Merged
Merged
Conversation
connortsui20
force-pushed
the
ct/arrow-ext-format-and-guards
branch
from
November 6, 2026 15:43
7d3d64c to
5aca0f4
Compare
connortsui20
force-pushed
the
ct/arrow-ext-format-and-guards
branch
from
November 6, 2026 15:53
5aca0f4 to
b202f2c
Compare
connortsui20
changed the title
connortsui20
changed the base branch from
develop
to
ct/arrow-ext-export-guard
November 6, 2026 15:54
connortsui20
force-pushed
the
ct/arrow-ext-export-guard
branch
from
November 6, 2026 15:58
21038a8 to
4820228
Compare
connortsui20
force-pushed
the
ct/arrow-ext-format-and-guards
branch
2 times, most recently
from
November 6, 2026 16:00
d433b00 to
32206cb
Compare
connortsui20
force-pushed
the
ct/arrow-ext-export-guard
branch
from
November 6, 2026 16:10
4820228 to
0679fa4
Compare
connortsui20
force-pushed
the
ct/arrow-ext-format-and-guards
branch
from
November 6, 2026 16:12
32206cb to
6154db5
Compare
connortsui20
changed the base branch from
ct/arrow-ext-export-guard
to
develop
November 6, 2026 16:12
connortsui20
marked this pull request as ready for review
November 6, 2026 16:16
Follow-up to #11224, from its review. The metadata of an extension type without an Arrow plugin was unmarked base64, but other Arrow plugins write the same keys with plain text metadata, so one string had two meanings. The metadata now carries `vortex:extension:metadata:encoding = base64`, as #10651 wrote it. The import decodes a marked value as base64 and reads an unmarked value as its UTF-8 bytes. The new public `ext_metadata` module owns the format, and `ArrowSession::to_arrow_storage_field` builds the field. Signed-off-by: Connor Tsui <[email protected]>
connortsui20
force-pushed
the
ct/arrow-ext-format-and-guards
branch
from
November 6, 2026 16:25
6154db5 to
9127e3b
Compare
connortsui20
enabled auto-merge (squash)
November 6, 2026 16:29
robert3005
approved these changes
Nov 6, 2026
12 of 49 tasks
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in 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.
Summary
Since #11224, the metadata of an extension type without an Arrow plugin is unmarked base64, but other Arrow plugins write the same keys with plain text metadata. So
"/w=="means the byte0xffto Vortex and the four bytes/w==to them. 0.92.0 and 0.93.0 already write this format, and every release that writes it into Parquet or IPC files makes it harder to change.Changes
The metadata now carries
vortex:extension:metadata:encoding = base64, as #10651 did, and unmarked metadata imports as its text. Writing text whenever the bytes are valid UTF-8 was rejected, because binary metadata such as02 00 00would then be written as text. Note that fields written by 0.92.0 and 0.93.0 no longer import.API Changes
vortex_arrow::ext_metadatamodule andArrowSession::to_arrow_storage_field.