event_reader.multicall_timestamp

Documentation for eth_defi.event_reader.multicall_timestamp Python module.

Read timestamps of blocks using multiprocess.

Module Attributes

HYPERSYNC_SPARSE_TIMESTAMP_MIN_STEP

Use exact sampled Hypersync reads when less than one per 100 blocks is needed.

Functions

fetch_block_timestamps_multiprocess(...[, ...])	Extract timestamps using fast multiprocessing.
fetch_block_timestamps_multiprocess_auto_backend(...)	Fetch block timestamps, choose backend.

HYPERSYNC_SPARSE_TIMESTAMP_MIN_STEP = 100

Use exact sampled Hypersync reads when less than one per 100 blocks is needed.

fetch_block_timestamps_multiprocess(chain_id, web3factory, start_block, end_block, step, display_progress=True, max_workers=8, timeout=120, cache_path=PosixPath('/home/runner/.tradingstrategy/block-timestamp'), checkpoint_freq=20000, rpc_request_stats=None)

Extract timestamps using fast multiprocessing.

• Subprocess entrypoint

• This is called by a joblib.Parallel

• The subprocess is recycled between different batch jobs

• We cache reader Web3 connections between batch jobs

• joblib never shuts down this process

Note

Because this method aggressively uses step to skip blocks, it results to non-reuseable timestamp cache (only valid for one scan and subsequent scans of the same task).

:param cache_path

Cache timestamps across runs and commands.

Set to None to disable, or remove the file. .

Parameters

• checkpoint_freq (int) – Block number frequency how often to save.

• rpc_request_stats (Optional[eth_defi.provider.rpcdb.RPCRequestStats]) – Optional parent accumulator receiving successful subprocess task calls.

• chain_id (int) –

• web3factory (eth_defi.event_reader.web3factory.Web3Factory) –

• start_block (int) –

• end_block (int) –

• step (int) –

• cache_path (Optional[pathlib.Path]) –

Return type

eth_defi.event_reader.timestamp_cache.BlockTimestampSlicer

fetch_block_timestamps_multiprocess_auto_backend(chain_id, web3factory, start_block, end_block, step, display_progress=True, max_workers=8, timeout=120, cache_path=PosixPath('/home/runner/.tradingstrategy/block-timestamp'), checkpoint_freq=20000, hypersync_client=None, rpc_request_stats=None)

Fetch block timestamps, choose backend.

• If Hypersync is available, use the optimised code path

For arguments see fetch_block_timestamps_multiprocess().

Parameters

• step (int) – Sampling interval. Wide intervals use a sparse, cache-aware Hypersync path instead of downloading every intervening block header.

• chain_id (int) –

• web3factory (eth_defi.event_reader.web3factory.Web3Factory) –

• start_block (int) –

• end_block (int) –

• cache_path (Optional[pathlib.Path]) –

• checkpoint_freq (int) –

• hypersync_client (hypersync.HypersyncClient | None) –

• rpc_request_stats (Optional[eth_defi.provider.rpcdb.RPCRequestStats]) –

Returns

Pandas series block number (int) -> block timestamp (datetime)

The mapping is deliberately keyed by block number. Modern chains such as Monad can produce multiple blocks in one second, while block-header timestamps retain one-second precision. Do not reverse this mapping or treat timestamp values as unique; retaining the same timestamp for multiple blocks is correct for the scanner’s intended precision.

Return type

eth_defi.event_reader.timestamp_cache.BlockTimestampSlicer