feat: support configure compression level #240
Conversation
Signed-off-by: tison <wander4096@gmail.com>
|
According to #239 (comment), pending to add a compression abstraction to hold these configs. |
Signed-off-by: tison <wander4096@gmail.com>
Signed-off-by: tison <wander4096@gmail.com>
This should be derived once 10XGenomics/lz4-rs#30 released.
Signed-off-by: tison <wander4096@gmail.com>
Signed-off-by: tison <wander4096@gmail.com>
|
Updated. cc @zamazan4ik @FlorentinDUBOIS @DonghunLouisLee Since I change the type of |
Signed-off-by: tison <wander4096@gmail.com>
| @@ -422,26 +423,26 @@ impl<Exe: Executor> TopicProducer<Exe> { | |||
|
|
|||
| let topic = topic.clone(); | |||
| let batch_size = options.batch_size; | |||
| let compression = options.compression; | |||
| let compression = options.compression.clone(); | |||
There was a problem hiding this comment.
This should be Copy. Blocked by upstream 10XGenomics/lz4-rs#30.
|
FMPOV looks fine. Thank you for the PR! |
tisonkun
changed the title
Signed-off-by: tison <wander4096@gmail.com>
|
@zamazan4ik do you have a suggestion on which version to release including this patch? 4.3.0 or 5.0.0. Semantically we should release 5.0.0 but actually this crate is not stable yet I would say (refer #224). |
|
@zamazan4ik No, I have no. For me personally, there is no considerable difference :) So I suggest just bumping the version to 4.3.0 . |
tisonkun
force-pushed
the
support-compression-level-zstd
branch
from
b5acebd to
34bf64a
Compare
Signed-off-by: tison <wander4096@gmail.com>
This reverts commit 362b9b3.
Signed-off-by: tison <wander4096@gmail.com>
Signed-off-by: tison <wander4096@gmail.com>
|
Updated. cc @fantapsody @FlorentinDUBOIS |
tisonkun
force-pushed
the
support-compression-level-zstd
branch
from
ab3de4b to
c889a08
Compare
| @@ -422,27 +423,7 @@ impl<Exe: Executor> TopicProducer<Exe> { | |||
|
|
|||
| let topic = topic.clone(); | |||
| let batch_size = options.batch_size; | |||
| let compression = options.compression; | |||
|
|
|||
| match compression { | |||
There was a problem hiding this comment.
Now the compression modes are structs under control. We exclude the corresponding enum members if features are unset. Thus, we're sure that there's no config-implementation mismatch.
Signed-off-by: tison <wander4096@gmail.com>
|
Pending to merge by the end of this week. I'm going to release a 5.0.0 for breaking changes. @fantapsody @FlorentinDUBOIS @DonghunLouisLee what do you think? For the increasing version number, the only risk I can see is that they're all beyond 1.0.0 but as stated in #224, they're actually untested for compatibility with a specific official Pulsar version. |
|
Pending to merge tomorrow and stage as 5.0.0. Perhaps I'll add rustfmt/clippy tools to reformat code stably before calling the release. |
This closes #239.