Skip to content

Latest commit

 

History

History
137 lines (99 loc) · 8.19 KB

README.md

File metadata and controls

137 lines (99 loc) · 8.19 KB
Raw

Calldata Compressor and Decompressor

Build Status Coverage Status NPM Package

This is a project with script for compress calldata and Solidity contract for decompress it.

Decompress algo

It's algorithm for decompress compressed data, which implements in _decompressed method of DecompressorExtension contract. The algorithm is to consider each of the compressed calldata's bytes in a row from 0 to the last and perform certain actions depending on the conditions. Depending on the first bits of a byte, it determines how the next bits or the next few bytes are handled. The following bitmasks are used to explain how to interpret a byte, or multiple bytes, to get a decompressed calldata:

  • 00XXXXXX - N zero bytes, where N is equal to (number in XXXXXX) + 1.
  • 01PXXXXX - copy next N bytes, where N is equal to (number in XXXXX) + 1, and if P=1 then pad this copied part with zeros to 32 bytes.
  • 10BBXXXX XXXXXXXX - use 12 bits (XX...X) as key for N bytes from storage _dict, where N is equal value in the array [32,20,4,31] with index K, where K equals to number in BB.
  • 11BBXXXX XXXXXXXX XXXXXXXX - use 20 bits (XX...X) as key for N bytes from storage _dict, where N is equal value in the array [32,20,4,31] with index K, where K equals to number in BB.

Compress algo

It's algorithm for compress data. It's checks all cases with bitmasks from Decompress algo section and uses dynamic algorithm to choose the best case with most zip functionallity. It calculates all the possible applications of different masks for each byte of the code, and then iterates over all the options using dynamic programming and chooses the best option.

Contract Overview

This contract provides a DecompressorExtension abstract contract that decompresses compressed calldata. It contains the following functions:

  • getData: retrieves data from the contract's dictionary and returns it in an array.
  • decompress: decompresses data stored in the contract using a compression algorithm.
  • decompressed: returns the decompressed data as bytes.
  • _setData: sets a single value in the contract's dictionary.
  • _setDataArray: sets multiple values in the contract's dictionary.

Usage

To use this contract, you must create a new contract that inherits from DecompressorExtension.

getData

To use the getData function, call it with a beginning and ending index, which will return an array of data values from the dictionary. First 2 positions are reserved with msg.senderand address(this), so you can't get it.

decompress

To use the decompress function, send a transaction to the contract with compressed data as input. This function will delegate call the decompressed function, which returns the decompressed data as bytes.

decompressed

To use the decompressed function, call it directly. This function returns the decompressed data as bytes.

_setData

The _setData is an internal function used within the contract or its inheritors. It sets a value in the dictionary at a specified offset, excluding the first two reserved slots for msg.sender and address(this).

_setDataArray

The _setDataArray is an internal function used within the contract or its inheritors. It's used to set an array of values in the dictionary starting from a specified offset. The first two slots, reserved for msg.sender and address(this), are excluded from this operation.

JS Compress script

To use compress script use compress method

const { compress } = require('@1inch/calldata-compressor/js/compressor.js');

Description of compress function

You have to fill dictionary, which you want to use in compress/decompress processes, using some functions with _setDataArray or _setData methods before use compress method. If you do not do this, then the compression will be without taking into account the dictionary. This is an asynchronous function that takes four parameters:

  1. calldata - This parameter is expected to contain some data that needs to be compressed.
  2. decompressorExt - This parameter is expected to be a Contract object of DecompressorExtension.
  3. wallet - This parameter is expected to be a msg.sender address.
  4. initDictAmount - This parameter is optional and has a default value of 1000. It represents the initial dictionary size to be used during compression.

This function compresses some data using dynamic programming and chooses the best compressed data.

CompressDataDescription Class

Represents a description of how to compress a specific portion of data. It has three properties:

  • startByte: Represents the starting byte index of the data portion to compress.
  • amountBytes: Represents the number of bytes to compress starting from the startByte index.
  • method: Represents the compression method (decompress mask) to use.

CompressDataPower Class

Represents the power of the compressed data. It has two properties:

  • decompressedSize: Represents the size of the original (decompressed) data in bytes.
  • compressedSize: Represents the size of the compressed data in bytes.

It also has two methods:

  • range(): Returns the difference between decompressedSize and compressedSize.
  • add(c): Adds the decompressedSize and compressedSize of another CompressDataPower instance c to the current instance.

CompressData Class

Represents the compressed data itself, along with its description and power. It has two properties:

  • power: An instance of CompressDataPower representing the power of the compressed data.
  • description: An instance of CompressDataDescription representing the description of how the data was compressed.

Calldata Class

This class provides a tool to compress a hexadecimal string representing a smart contract call. The compression is done by replacing repeating patterns with more concise descriptions, resulting in a smaller encoded string.

analyse()

  • Analyzes the hexadecimal data and computes compression information for each byte.
  • Returns:
    • Array - An array of objects containing information about compression for each byte. Each object has the following properties:
      • index: The index of the byte.
      • zeroCompress: An object with information about the number of contiguous zeros that can be compressed.
      • copyCompress: An object with information about the number of bytes that can be compressed by copying a next block of data.
      • storageCompress: An array with objects containing information about blocks that can be compressed using data from the dictionary.

#compressPart(fromByte, toByte)

  • Compresses a part of the hexadecimal data between the specified byte indexes.
  • Arguments:
    • fromByte - The index of the first byte to be compressed.
    • toByte - The index of the last byte to be compressed.
  • Returns:
    • CompressData - The compressed data information, including the compressed power and description.

compress()

  • Compresses all hex data by splitting it into parts, compressing each part and choosing the best combination to compress.
  • Returns:
    • Array - An array of compressed data parts, each containing the compressed power and description.

Tests environment

You can find examples of dict and calldatas in IPFS:

File IPFS CID
dict.json Qmb4fEEdQe9QB4ChBFK8bJVkefCjG4nB91hgQ2kcKsKGWK
tx-calldata.json QmQnjk5MKMiC6jZNjWgXKkbCMFFznNYaHRsS281jV94wpe

Just add this files to ./test dir

License

This contract is licensed under the MIT license, as indicated by the SPDX-License-Identifier.