feat: add additional layers support

[Retrospection]

Summary

This PR adds support for additional layers to the Java bindings to align with Python and Node.js bindings.

Added Layers

• MimeGuessLayer: Automatically set Content-Type based on file extension
• TimeoutLayer: Add timeout for operations (both IO and non-IO)
• LoggingLayer: Add structured logging for all operations
• HotpathLayer: Add hotpath profiling for performance analysis

Changes

• Added 4 new Java layer classes in bindings/java/src/main/java/org/apache/opendal/layer/
• Added corresponding JNI implementations in bindings/java/src/layer.rs
• Updated bindings/java/Cargo.toml to enable required features
• Added tests in bindings/java/src/test/java/org/apache/opendal/test/LayerTest.java

Testing

All 7 layer tests pass:

• testOperatorWithRetryLayer
• testOperatorWithConcurrentLimitLayer
• testOperatorWithMimeGuessLayer
• testOperatorWithTimeoutLayer
• testOperatorWithLoggingLayer
• testOperatorWithThrottleLayer
• testOperatorWithHotpathLayer

Issue

Closes #6738

[PsiACE]

for toml format , u can use taplo to fix it

[Xuanwo]

Hi, please understand your changes first. LoggingLayer integrate rust's log crate and thus requires we to provide a set_logger call. Users can't configure that through log4j2 or java.util.logging.

Maybe we can start a discussion thread on how we do that.

[Retrospection]

Sorry for rough pr.

Honestly I'm not that familiar with this binding layer, and just think this can be implemented with ai just because it has other language implementation as reference.

I'll do some research to understand what the binding layer is for and then do discussion about implementation.

Thanks for kindly review.

[Retrospection]

Hi, please understand your changes first. LoggingLayer integrate rust's log crate and thus requires we to provide a set_logger call. Users can't configure that through log4j2 or java.util.logging.

Maybe we can start a discussion thread on how we do that.

I have been studying the abstraction role of the OpenDAL layer and now understand the issues with the Java LoggingLayer implementation in my previous PR.

I would like to ask:

1. Is it okay if I temporarily abandon the implementation of the LoggingLayer part for now?
2. Alternatively, should I pursue a deeper integration with SLF4J, Log4j, or java.util.logging instead?