Closed nicklan closed 3 days ago
Attention: Patch coverage is 98.61111%
with 1 line
in your changes missing coverage. Please review.
Project coverage is 79.64%. Comparing base (
67aa7d5
) to head (b20642a
). Report is 1 commits behind head on main.
Files with missing lines | Patch % | Lines |
---|---|---|
test-utils/src/lib.rs | 98.21% | 0 Missing and 1 partial :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
weird semver fails due to
failed to parse /home/runner/work/delta-kernel-rs/delta-kernel-rs/target/semver-checks/git-9861282e77ba1e3ef3410780d490feab52e90b18/83da59f23e419fe81cc19b13b6750f4c634b28bb/Cargo.toml: no `package` table,
weird semver fails due to
failed to parse /home/runner/work/delta-kernel-rs/delta-kernel-rs/target/semver-checks/git-9861282e77ba1e3ef3410780d490feab52e90b18/83da59f23e419fe81cc19b13b6750f4c634b28bb/Cargo.toml: no `package` table,
Ahh, it's because it's trying to compare against main
which doesn't have test_utils
. Will only be an issue for this PR
nit: is
test-utils
ortest_utils
more idiomatic I don't remember? Also does naming not matter since this won't ever make it to crates.io?
Yeah, doesn't really matter, but we've gone camel_case
for all the other crates we have, so I've updated it to make it consistent.
What changes are proposed in this pull request?
Move some useful common utils for building an in-memory delta table into their own crate.
I want to start adding tests for actually reading data to the
ffi
crate. To do so we need to use anInMemory
object store, because all i/o in tokio fails in miri due to it using epoll, which miri doesn't support. So we'll need to build in-memory tables and use thememory:///
url to get a memory backedObjectStore
.We already do this in
read.rs
in kernel, but there's no way to get at that code inside the ffi crate. So this moves a bunch of utilities for this sort of thing into its own "test-utils" crate and uses it in read.rs.I've also cleaned up and renamed things a bit. Most of the code remains unchanged.
Follow-up PR(s) will use this in the ffi crate to test more.
How was this change tested?
Existing tests pass