Witness Items and Count

The Witness section of a transaction is a list of lists. Every input has its own "Witness Field," and every witness field can contain multiple "Witness Items."

1. Defining a Witness Item

A witness item is a piece of data required for validation. In a standard P2WPKH transaction, there are exactly two items:

1. The ECDSA Signature.

2. The Public Key.

In more complex scripts, like a 2-of-3 Multi-sig, there might be four items (OP_0, Sig1, Sig2, and the Redeem Script).

2. Encoding the Count

Before the items for a specific input are listed, there is a Count byte.

• If the input is Legacy, the count is 00.

• If the input is P2WPKH, the count is 02.

• If the input is Taproot, the count is usually 01 (since the pubkey is in the output).

3. The "Itemized" Format

Each item in the list is prefixed by its length as a VarInt. This makes it incredibly efficient for nodes to "pop" items off the witness and push them onto the execution stack.

[Item Count]
  [Item 1 Length][Item 1 Data]
  [Item 2 Length][Item 2 Data]

4. The wTXID (Witness Transaction ID)

Because the witness data is required for full validation but is not part of the TXID, Bitcoin uses a second hash called the wTXID.

• TXID: Hash of the transaction without Marker, Flag, or Witness.

• wTXID: Hash of the entire transaction including Marker, Flag, and Witness.

Miners use the wTXID to build the Witness Merkle Tree, which is committed to the coinbase transaction of every block.

In the final section, we will build a Python Witness Parser.