ArkLib

This is the home of core ARK library.

Being implemented in Rust, it provides us capability to port all our apps to all common platforms. Right now, Android is supported by using the arklib-android project. And for Linux/macOS/Windows, the library can be used as-is and easily embedded into an app, e.g. built with Tauri. Development docs will come sometime.

The Concept of the library

The purpose of the library is to manage resource index of folders with various user data, as well as to manage user-defined metadata: tags, scores, arbitrary properties like movie title or description. Such a metadata is persisted to filesystem for easier sync and backup. The resource index provides us with content addressing and allows easier storage and versions tracking. We also believe it'll allow easier cross-device sync implementation.

Prerequisites

• PDFium prebuilt (bblanchon)

Build

Like most of Rust projects:

cargo build --release

Run unit tests:

cargo test

Development

For easier testing and debugging, we have the ARK-CLI tool working with ARK-enabled folders.

Benchmarks

arklib relies on the criterion crate for benchmarking to ensure optimal performance. Benchmarks are crucial for evaluating the efficiency of various functionalities within the library.

Running Benchmarks

To execute the benchmarks, run this command:

cargo bench

This command runs all benchmarks and generates a report in HTML format located at target/criterion/report. If you wish to run a specific benchmark, you can specify its name as an argument as in:

cargo bench index_build

Benchmarking Local Files

Our benchmark suite includes tests on local files and directories. These benchmarks are located in the benches/ directory. Each benchmark sets a time limit using group.measurement_time(), which you can adjust manually based on your requirements.

You have the flexibility to benchmark specific files or folders by modifying the variables within the benchmark files. By default, the benchmarks operate on the tests/ directory and its contents. You can change the directory/files by setting the DIR_PATH and FILE_PATHS variables to the desired values.

For pre-benchmark assessment of required time to index a huge local folder, you can modify test_build_resource_index test case in src/index.rs.