feat: reuse buffer when serializing #160
Merged
Conversation
This file contains 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
This commit tries to minimize the number of allocations performed when serializing data. For this end, the connector and built-in Source now create internal `Vec<u8>` buffers that are reused whenever a message is serialized. This change should hopefully improve performance.
gabrik
approved these changes
Apr 28, 2023
J-Loudet
added a commit
that referenced
this pull request
May 2, 2023
* feat: support for custom (de)serializer
This commit introduces the following changes to Zenoh-Flow:
1. types used within a Zenoh-Flow application do not have to implement the
(cumbersome) `ZFData` trait; instead,
2. the types used must be `Send + Sync + 'static`,
3. an Output must know how to serialize the provided type `T` and, respectively,
an Input must know how to deserialize bytes into `T`.
With these changes, any SerDe compatible data serialization format is supported
as well as, for instance, ProtoBuf.
In terms of API, the major difference is how the Input and Output are obtained:
```rust
input: inputs
.take("in", |bytes| todo!("Provide your deserializer here"))
.expect("No input called 'in' found"),
output: outputs
.take("out", |data| todo!("Provide your serializer here"))
.expect("No output called 'out' found"),
```
* fix: move `DeserializerFn` in types/message.rs
* fix: remove explicit call to `Arc::clone`
* fix: add `as_mut_any` on `SendSyncAny` trait
* fix: implement/derive Debug for the different messages
* feat: InputBuilder and OutputBuilder
Instead of having the methods `take` and `take_raw` on the Inputs and Outputs,
that give, respectively, the Typed and Raw Input/Output, this commit removes
the `take_raw` variant and introduces builders.
The builders can be turned into the Typed or Raw variant using the corresponding
methods: `build_typed` and `build_raw`.
* doc: improve documentation of Input / Output
* feat: reuse buffer when serializing (#160)
* feat: reuse buffer when serializing
This commit tries to minimize the number of allocations performed when
serializing data. For this end, the connector and built-in Source now create
internal `Vec<u8>` buffers that are reused whenever a message is serialized.
This change should hopefully improve performance.
* doc: fix typos
* refacto: renamed `build_raw` and `build_typed` to `raw` and `typed`
---------
Signed-off-by: Julien Loudet <julien.loudet@zettascale.tech>
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.
These commits try to minimize the number of allocations performed when
serializing data. For this end, the connector and built-in Source now create
internal
Vec<u8>buffers that are reused whenever a message is serialized.This change should hopefully (slightly) improve performance.