Schema & File Structure

R2 File Structure

We're hosting indexed.xyz on Cloudflare's R2. R2 has an S3 API, so you can grab these files either with the Goldsky CLI (the quickest way to get started), the AWS CLI tool, or rclone.

The prefix structure for data in R2 is:

s3://indexed-xyz-wnam/<chain>/(decoded|raw)/logs/v2.0.0/dt=<yyyy-MM-dd>

Right now these chains are supported, but if you'd like to see other chains here, shoot us an email and we'll consider adding it.

You'll probably want the decoded files, as that's what this document describes.

The data is partitioned by day. In most tools you can leave that part of the prefix off and download all data recursively, but to limit downloads and local storage, you may want to pull a smaller subset of the data to get started.

Decoded Logs

The objects in R2 are Parquet files.

The Parquet file scheme we're using is:

column_name	column_type
block_timestam	BIGINT
address	VARCHAR
event_signature	VARCHAR
event_params	VARCHAR[]
block_number	BIGINT
block_hash	VARCHAR
log_index	BIGINT
transaction_hash	VARCHAR
transaction_index	BIGINT
data	VARCHAR
topics	VARCHAR
id	VARCHAR
dt	VARCHAR

Here’s an example from one of the files, queried using DuckDB:

SELECT * FROM 'ethereum_decoded_logs_v1.2.0_partition_key=00_dt=2021_part-3acd096e-279f-42aa-9c5b-b03397cf6f7f-24-134.parquet' LIMIT 1;

block_time	address	event_signature	event_params	block_number	block_hash	log_index	transaction_hash	transaction_index	data	topics	id
1609514571	0x056fd409e1d7a124bd7017459dfea2f387b6d5cd	Transfer(address,address,uint256)	[0xd24400ae8BfEBb18cA49Be86258a3C749cf46853, 0xe2983860a018FB295ec08C3B63A487874E4d1469, 25000]	11569225	0x8911437fe4d3dad8e887ab58dc4107595cb2541bc3d1fbb27ba6da9dfb12bd50	4	0xc9e3cb4357516616abddc10f5e63795694e5dbd7bcfa88b757a9aab076263066	11	0x00000000000000000000000000000000000000000000000000000000000061a8	0xddf252ad1be2c89b69c2b068fc378daa952ba7f163c4a11628f55a4df523b3ef,0x000000000000000000000000d24400ae8bfebb18ca49be86258a3c749cf46853,0x000000000000000000000000e2983860a018fb295ec08c3b63a487874e4d1469	decoded_log_0x8911437fe4d3dad8e887ab58dc4107595cb2541bc3d1fbb27ba6da9dfb12bd50_4

Some caveats to keep in mind:

• We're still validating this dataset.
• Though the prefix key we use is based on an all-lowercased address, and md5 digest, the addresses within the parquet files do not have a consistent case in all the fields.

Raw Blocks

column_name	column_type
id	VARCHAR
number	BIGINT
hash	VARCHAR
parent_hash	VARCHAR
nonce	VARCHAR
sha3_uncles	VARCHAR
logs_bloom	VARCHAR
transactions_root	VARCHAR
state_root	VARCHAR
receipts_root	VARCHAR
miner	VARCHAR
difficulty	DOUBLE
total_difficulty	DOUBLE
size	BIGINT
extra_data	VARCHAR
gas_limit	BIGINT
gas_used	BIGINT
timestamp	BIGINT
transaction_count	BIGINT
base_fee_per_gas	BIGINT
dt	VARCHAR

Raw Transactions

column_name	column_type
id	VARCHAR
hash	VARCHAR
nonce	BIGINT
block_hash	VARCHAR
block_number	BIGINT
transaction_index	BIGINT
from_address	VARCHAR
to_address	VARCHAR
value	VARCHAR
gas	VARCHAR
gas_price	VARCHAR
input	VARCHAR
max_fee_per_gas	VARCHAR
max_priority_fee_per_gas	VARCHAR
transaction_type	BIGINT
block_timestamp	BIGINT
dt	VARCHAR

Raw Logs

column_name	column_type
block_timestamp	BIGINT
block_number	BIGINT
block_hash	VARCHAR
transaction_hash	VARCHAR
transaction_index	BIGINT
log_index	BIGINT
address	VARCHAR
data	VARCHAR
topics	VARCHAR
id	VARCHAR
dt	VARCHAR

Arweave Raw Blocks

column_name	column_type
id	VARCHAR
height	INTEGER
mined_at	TIMESTAMP
created_at	TIMESTAMP
previous_block	VARCHAR
mined_at_utc	BIGINT

Arweave Raw Transactions

column_name	column_type
id	VARCHAR
owner	VARCHAR
tags	VARCHAR
target	VARCHAR
quantity	VARCHAR
reward	VARCHAR
signature	VARCHAR
last_tx	VARCHAR
data_size	BIGINT
content_type	VARCHAR
format	INTEGER
created_at	TIMESTAMP
deleted_at	TIMESTAMP
height	INTEGER
owner_address	VARCHAR
data_root	VARCHAR
parent	VARCHAR