Zellic/smart-contract-fiesta

Dataset Overview

Dataset Name

• Name: Zellic 2023 Smart Contract Source Index
• Alias: Zellic Smart Contract Source Index

Dataset Description

• Purpose: Provide a publicly downloadable Ethereum mainnet smart contract source code dataset to advance smart contract security research.
• Applications: Includes static analysis, machine learning, etc.

Dataset Content

• Methodology:
  • Collect all contract addresses deployed on the Ethereum mainnet and their EVM bytecode Keccak256 hashes.
  • Build the index by fully syncing from the genesis block using a modified Geth instance.
  • De‑duplicate source code based on bytecode hash.
• Statistics:
  • Unique Source Codes: 149,386
  • Contracts with Code: 3,897,319
  • Total Smart Contracts in Global Index: 30,586,657
  • Character Count: 6,473,548,073
  • Word Count: 712,444,206
  • Lines of Code: 90,562,628
  • Comment Lines: 62,503,873
  • Blank Lines: 24,485,549
  • Total Lines: 177,552,050
  • Unique Words: 939,288

Dataset Structure

• Index:
  • Filename: address_bytecodehash_index
  • Content: Mapping of all deployed contract addresses to the Keccak256 hash of their EVM bytecode.
• Contract Source Code:
  • Storage Location: Under the organized_contracts directory, organized by bytecode hash.
  • Contents: Source files and metadata.json (includes compiler version, optimization settings, etc.).
  • Source Formats: Single‑file, multi‑file, Solidity compiler JSON input.

Additional Information

• Contract Languages: Not limited to Solidity; includes Vyper and other languages.
• Source Extraction: A Bash script is provided to extract all source code.