README.md

Introduction

This crate introduces a default command line tool with a pre-defined mock runtime (wasm runtime and Host APIs) of Smart Intermediate Representation.

How to Use

First, build the binary of ir_cli

cd ir_cli
make

Then, there is a binary called ir_cli in the directory target/debug

Second, build the wasm bytecode of the textual IR file. For example:

./target/debug/ir_cli build ../ir_example/hello_world.ir

Third, run the bytecode in the pre-defined mock runtime environment:

./target/debug/ir_cli run --func greeting a.out.wasm a.out.abi.json
./target/debug/ir_cli run --func greeting2 a.out.wasm a.out.abi.json [string]

Demo

Run the function greeting2 of example ir_example/hello_world.ir, you will get following result:

./target/debug/ir_cli run --func greeting2 a.out.wasm a.out.abi.json test    
CALL PRINTLN: Hello Smart Intermediate Representation
result: test

AI Intent

In order to use the capabilities of AI Intent, some preparations are required.

First, you need to install solcjs dependencies. We need to use solcjs to compile the solidity contract. Need to be executed in the root directory of this project

npm install

Second, need to build ir_cli according to the above steps and add the ir_cli to PATH.

Finally, execute the subcommand of ir_cli to generate tensor data of AI Intent.

ir_cli sol2tensor $input_dir_absolute_path

├- input
| ├- file1.json
| ├- file2.json
| └- ...
└- output
  ├- source_info -- information about contract
  ├- standard_input -- solc compiler standard input json
  ├- standard_output -- solc compiler standard output json
  ├- yul -- yul src code
  ├- sir -- Smart IR
  └- tensor -- tensor data