Token dumper

blockzilla dump-tokens reads an epoch and writes every recorded token balance as one JSON line. Use it for a one-time export, or to load a database such as ClickHouse.

For a database that resumes after a stop, use the token index instead.

Export USDC balances

blockzilla dump-tokens --epoch 200 --max-blocks 20 \
  --mint EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v > usdc.jsonl

This writes 1,334 rows. Remove --mint to export all mints. Remove --max-blocks to export the whole epoch.

The command accepts the same options as stream.

Output fields

One line is one balance:

Field	Meaning
source_id	ID of the archive build that the row came from
epoch, block_ordinal, slot, tx_index	Where the transaction is
execution_status	succeeded or failed
side	pre (before the transaction) or post (after it)
balance_index	Position in Solana’s pre or post balance list
account_index	Position of the token account in the transaction’s account list
mint, owner, token_program	Base58 addresses, or null when not recorded
amount	Raw amount as a decimal string, so no precision is lost
decimals	Decimal places of the mint

The token account address is not in the output yet. account_index gives its position in the transaction.

What the rows mean

• Rows are balances recorded before and after each transaction. They are not transfers. To get a change, pair the pre and post rows of the same transaction and account_index.
• Failed transactions are included. Filter on execution_status = 'succeeded' for committed changes.
• If a range has no rows for your mint, no transaction in that range recorded a balance for it. It does not mean the mint has no history.
• With --mint, rows with a null mint stay in the output, because they can be a match.
• If execution status or token data is missing, the command stops. It does not skip rows.

Save a complete file

The command has no resume position. If it fails, the output is incomplete. Write to a temporary name and rename only after success:

blockzilla dump-tokens --epoch 200 > tokens.partial.jsonl &&
  mv tokens.partial.jsonl tokens-200.jsonl

After a failure, run the full export again. Do not append to the partial file.

Change the code

The command is dump_tokens.rs. To write your own version, see Read token balances.