Skip to content

Latest commit

 

History

History
62 lines (42 loc) · 1.73 KB

Development.md

File metadata and controls

62 lines (42 loc) · 1.73 KB

Development

Developing the library requires only a working rust environment. Integration testing and code coverage report generation require docker and docker-compose.

Running tests

In the main directory, you can run cargo test --all-features to run all the unit tests.

Running integration tests (requires docker-compose)

In the root directory, run docker-compose up -d to run a bitcoin node and electrs instance.

Then in the dlc or dlc-manager directory run

cargo test -- --ignored test_name

replacing test_name with the test you want to run. For example within the dlc-manager folder:

cargo test -- --ignored two_of_five_oracle_numerical_test

Running fuzz tests

Some fuzz testing are implemented, check the documentation for details.

Generating code coverage report

Code coverage for unit tests

From the main folder run:

./scripts/generate_test_coverage.sh

This will generate a coverage report in coverage/index.html.

Code coverage for dlc-manager integration tests

From the main folder run:

./scripts/generate_integration_test_coverage.sh

The coverage report will be available at integration_coverage/index.html.

Benchmarking and profiling

At the moment the bottlenecks are mainly the adaptor signature verification and signing for numerical outcome DLC with a relatively large number of CETs. However it can sometimes be useful to have a look at benchmarks and do some profiling.

Benchmarking (requires rust nightly)

Some benchmarks are available for the dlc manager. To run them:

cargo +nightly bench  --features=unstable

Profiling (requires docker-compose)

Profiling is currently not working.