Repository navigation
Conversation
The README still said "after registration" and described the CI matrix as nightly on every OS; CI runs 1.10 and stable on three OSes and 1.11 and the latest prerelease on Linux. The module docstring now qualifies the byte-for-byte Arrow claim to little-endian hosts, as the README does, and drops all-caps emphasis and "an `DataString`" left from the rename. The changelog gains the 1.1.0 entry and the unreleased encoding fix from #3. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Small accuracy and wording fixes; no code changes.
pre, not nightly) on Linux.DataString" left from the ArrowStrings rename (also in theStringVector/BytesVectordocstrings).convert(DataBytes, ...),materializeon plain vectors, Add convert to DataBytes and materialize for plain vectors of values #2), and an Unreleased entry for the encoding fix in Normalize string encodings before column edits #3, which is on main but not in a release.Verified the unchanged claims with probes:
hash,cmp,==,isless, iteration,codeunit,length) allocate zero bytes on 1.10.11 and 1.12.6.Stringon all 139,810 byte sequences of length 0-4 drawn from 16 UTF-8 class-boundary bytes, inline and as views, on 1.10, 1.12, and 1.13.DataBytesequality and hashing matchVector{UInt8}.Pkg.test()passes on Julia 1.10.11 and 1.12.6.🤖 Generated with Claude Code