feat: implement batching for multiple extrinsic calls to reduce trans…

[mimalsm]

…action fees

This commit implements comprehensive batching support for CLI commands that can submit multiple extrinsic calls, significantly reducing transaction fees for users. This addresses GitHub issue #567.

Core Changes:

1. New Batching Interface (subtensor_interface.py)

   • Added sign_and_send_batch_extrinsic() method to support batching multiple GenericCall objects into a single transaction
   • Supports three batch types: 'batch', 'batch_all', and 'force_batch'
   • Defaults to 'batch_all' to ensure atomicity (all succeed or all fail)
   • Fully integrates with existing features: MEV protection, proxy transactions, era management, and nonce handling
   • Returns detailed results including success status, extrinsic receipt, and individual call results

2. Stake Add Command Refactoring (stake/add.py)

   • Refactored to collect multiple stake operations (across different netuids and/or hotkeys) into a single batch transaction
   • Automatically uses batching when multiple --netuid or --hotkey arguments are provided
   • Optimized balance fetching using asyncio.gather for parallel queries
   • Enhanced display to show:
     • Individual stake changes per netuid/hotkey combination
     • Final coldkey balance change (previous_balance -> current_balance) * Clear visual feedback for each operation in the batch

3. Stake Remove Command Refactoring (stake/remove.py)

   • Refactored unstake() to batch multiple remove_stake operations
   • Refactored unstake_all() to batch multiple unstake_all operations
   • Supports batching when:
     • Multiple --netuid arguments are provided
     • Multiple --hotkey arguments are provided * --include-hotkeys with multiple hotkeys is used
   • Optimized stake balance fetching using asyncio.gather
   • Enhanced display to show:
     • Individual unstake changes per netuid/hotkey
     • Remaining stake balances after operations * Final coldkey balance change (previous_balance -> current_balance)
   • Captures initial balance at function start for accurate display

4. CLI Validation Updates (cli.py)

   • Removed restriction preventing unstake_all with multiple --include-hotkeys
   • Updated logic to properly initialize wallet when include_hotkeys is provided
   • Now supports batching unstake_all operations across multiple hotkeys

Technical Details:

• Batch Type: Uses Utility.batch_all by default, ensuring atomic execution
• Fee Savings: Multiple operations combined into single transaction fee
• Backward Compatibility: Single operations still work as before (no batching)
• Error Handling: batch_all ensures all operations succeed or fail together
• Display Consistency: Balance changes shown consistently across all stake operations

Benefits:

• Significant fee reduction when performing multiple stake/unstake operations
• Improved user experience with clear feedback on all operations
• Atomic operations prevent partial failures
• Maintains all existing functionality (MEV protection, proxy, etc.)

Closes #567

[thewhaleking]

Looks nice, needs tests.

[thewhaleking]

you'll want to get the block_hash = chain head at the beginning of the function, and then, in these compose_call calls, pass the block_hash=block_hash, which will significantly speed things up (otherwise it needs to make a network call for each compose_call call.

[mimalsm]

I just noticed that many of my prs are all closed. giving up gittensor