Add BytePackLib for efficient content storage

[RogerPodacter]

• Introduced BytePackLib to pack small byte arrays (0-31 bytes) into a bytes32 slot.
• Updated Ethscriptions contract to use packed content or SSTORE2 pointers based on content size.
• Modified content storage mapping to accommodate both packed data and SSTORE2 pointers.
• Enhanced getEthscriptionContent function to handle unpacking of packed data.

---

Note

Introduce BytePackLib and switch content storage to a single bytes32 that inlines ≤31B data or stores an SSTORE2 pointer for larger payloads.

• Core (Ethscriptions.sol):
  • Replace contentPointerBySha with unified contentStorageBySha: bytes32 storing either packed inline bytes (≤31) or an SSTORE2 pointer (≥32).
  • Add BytePackLib usage in getEthscriptionContent to detect/unpack inline data; otherwise read via SSTORE2Unlimited.
  • Update _storeContent to pack small calldata via BytePackLib.packCalldata or write to SSTORE2Unlimited and store pointer as bytes32.
• Library:
  • Add BytePackLib with packCalldata, unpack, isPacked, and packedLength for 0–31 byte payloads.
• Testing:
  • New BytePackLib.t.sol covering packing/unpacking, bounds, and format.
  • Add PackUnpackTest.t.sol examples for pack/unpack behavior.
  • Extend EthscriptionsWithTestFunctions to expose storage/read helpers compatible with packed-or-pointer format.
  • Update SSTORE2ContentSizes.t.sol to assert inline storage for <32B and pointers for ≥32B, and to validate retrieval/deduplication.

^{Written by Cursor Bugbot for commit 08d416c. This will update automatically on new commits. Configure here.}

[Copilot]

Pull Request Overview

This PR optimizes storage for small ethscription content by packing byte arrays up to 31 bytes directly into a single bytes32 storage slot, avoiding the gas cost of SSTORE2 deployments for small content. The implementation introduces a new BytePackLib library that uses a tag byte to distinguish packed data from SSTORE2 pointers.

• Adds BytePackLib library for packing/unpacking small byte arrays (0-31 bytes) into bytes32
• Refactors content storage to use packed format for content ≤31 bytes, SSTORE2 for ≥32 bytes
• Updates contentPointerBySha mapping from address to bytes32 to support dual-mode storage

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
contracts/src/libraries/BytePackLib.sol	New library implementing byte packing/unpacking with tag-based format
contracts/src/Ethscriptions.sol	Updated content storage to use packed format for small content or SSTORE2 pointers for larger content

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The isPacked function only checks if the first byte is non-zero, which could lead to false positives. While Ethereum addresses (160 bits = 20 bytes) cast to bytes32 have their leftmost 12 bytes as zero under normal circumstances, the tag byte range (1-32) could theoretically overlap with the upper bits if an address somehow had data in those positions. Consider adding an additional check to validate that the tag byte is within the valid range (1-32) to make the distinction more robust: uint8 tag = uint8(uint256(value >> 248)); return tag > 0 && tag <= 32;

Suggested change

	// Packed data has a non-zero tag byte (1-32) in the first byte
	return uint8(uint256(value >> 248)) != 0;
	// Packed data has a tag byte in the range 1-32 in the first byte
	uint8 tag = uint8(uint256(value >> 248));
	return tag > 0 && tag <= 32;

[Copilot]

Pull Request Overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 3 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The condition content.length <= 31 is inconsistent with the BytePackLib implementation which supports packing data up to 31 bytes (if (len >= 32) revert). However, the comment on line 79 states for <32 bytes which is correct. The condition should use < instead of <= to match the boundary of 32 bytes mentioned in the comment, or the comment should be updated to say for <=31 bytes. The current code is correct but the comment inconsistency could cause confusion.

Suggested change

	// Pack small content directly into bytes32 (0-31 bytes)
	// Pack small content directly into bytes32 (for <=31 bytes)

[Copilot]

The boundary condition uses >= 32 to determine SSTORE2 storage, but line 516 in Ethscriptions.sol uses <= 31 for inline storage. While these are mathematically equivalent, they check opposite conditions. For consistency and clarity, this should use > 31 to match the conceptual model that content > 31 bytes uses SSTORE2, or both places should use the same boundary expression (>= 32 vs <= 31).

Suggested change

	if (content.length >= 32) {
	if (content.length > 31) {

[Copilot]

[nitpick] The code check pointer.code.length > 0 may return false positives during contract construction. While unlikely in this test context since SSTORE2 contracts are immediately deployed, a more robust check would be pointer.code.length > 0 && pointer != address(0), though the prior address(0) return on line 37 makes this redundant. Consider simplifying by removing this check entirely since isPacked already filters out inline content.