Repository navigation
Mark empty char values with 0x7F on every platform - #14
Merged
Merged
Conversation
TILEDB_EMPTY_CHAR was CHAR_MAX, which is 127 where char is signed and 255 where it is unsigned, as on Linux aarch64, so arrays written there marked empty char cells with 0xFF and arrays written on x86-64 or macOS with 0x7F, and neither read the other's as empty. It is now 0x7F everywhere. Char data already written on Linux aarch64 keeps its 0xFF markers, which are no longer read as empty.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
TILEDB_EMPTY_CHARwasCHAR_MAX, which is 127 wherecharis signed (x86-64, macOS) and 255 where it's unsigned (Linux aarch64 and other non-x86 Linux targets). Arrays written on Linux aarch64 therefore marked empty char cells with 0xFF and arrays written elsewhere with 0x7F, and each platform read the other's markers as values; GenomicsDB also uses it as the missing value for char fields, so a workspace imported on one platform and queried on the other read missing string fields as one-character strings. This defines it as 0x7F everywhere, replacing #11. Char data already written on Linux aarch64 keeps its 0xFF markers, which will no longer read as empty.I also went through the rest of what TileDB writes for the same problem and found nothing else: the other empty values are fixed-width integers or IEEE floating point,
TILEDB_VAR_NUMisINT_MAX, which is 32 bits everywhere TileDB builds, and the book-keeping and array schema are written asint64_t,off_tandsize_t, all 64-bit on every 64-bit platform, or as non-negative bytes, apart from the compression levels in the array schema, which #13 fixes. The empty-value test now checks for 0x7F rather thanCHAR_MAX. I checked it on Linux aarch64 in an AlmaLinux 8 container with GCC 15, together with #12 and #13: every test and example passes, and the test fails without this change.