Home

Awesome

fast-dat-parser

Parses the blockchain about as fast as your IO can pipe it out. For a typical SSD, this can be around ~450 MiB/s.

All memory is allocated up front.

Output goes to stdout, stderr is used for logging.

WARNING: Not actively maintained, use with caution.

Usage

A fast blk*.dat parser for bitcoin blockchain analysis.

Important to note is that the implementation skips bitcoind allocated zero-byte gaps, and includes orphan blocks unless -w omits them.

Transforms (-t)

Each of these pre-included functions write their output as raw data (binary, not hex). You can easily write your own though!

Use a whitelist (see -w) to stop orphan blocks from being parsed. (see below for filtering by best chain)

Examples

Output all scripts for the local-best blockchain

# parse the local-best blockchain
cat ~/.bitcoin/blocks/blk*.dat | ./parser -t0 | ./bestchain > chain.dat

# output every script found in the local-best blockchain
cat ~/.bitcoin/blocks/blk*.dat | ./parser -j4 -t1 -wchain.dat > ~/.bitcoin/scripts.dat

Useful tools

These tools are for the CLI, but will aid in preparing/using data produced by the above.

bestchain

A best-chain filter for block headers.

Accepts 80-byte block headers until EOF, then finds the best-chain in the set, and outputs the best-chain in the form of a sorted hash map (see HMap<K, V>).

LICENSE MIT

The constants and getOpString function in include/bitcoin-ops.hpp is copied from https://github.com/bitcoin/bitcoin/.