All bitcoin transactions have scripts embedded into its inputs and outputs. The scripts use a very simple programming language, which is evaluated from left to right using a stack. The language is designed such that it guarantees all scripts will execute in a limited amount of time (it is not Turing-Complete).
When a transaction is validated, the input scripts are concatenated with the output scripts and evaluated. To be valid, all transaction scripts must evaluate to true. A good analogy for how this works is that the output scripts are puzzles that specify in which conditions can those bitcoins be spent. The input scripts provide the correct data to make those output scripts evaluate to true.
The `Script` object provides an interface to construct, parse, and identify bitcoin scripts. It also gives simple interfaces to create most common script types. This class is useful if you want to create custom input or output scripts. In other case, you should probably use `Transaction`.
This is the most commonly used transaction output script. It's used to pay to a bitcoin address (a bitcoin address is a public key hash encoded in base58check)
Pay to public key scripts are a simplified form of the p2pkh, but aren't commonly used in new transactions anymore, because p2pkh scripts are more secure (the public key is not revealed until the output is spent).
Multisig outputs allow to share control of bitcoins between several keys. When creating the script, one specifies the public keys that control the funds, and how many of those keys are required to sign off spending transactions to be valid. An output with N public keys of which M are required is called an m-of-n output (For example, 2-of-3, 3-of-5, 4-of-4, etc.)
Pay to script hash outputs are scripts that contain the hash of another script, called `redeemScript`. To spend bitcoins sent in a p2sh output, the spending transaction must provide a script matching the script hash and data which makes the script evaluate to true. This allows to defer revealing the spending conditions to the moment of spending. It also makes it possible for the receiver to set the conditions to spend those bitcoins.
Data outputs are used to push data into the blockchain. Up to 40 bytes can be pushed in a standard way, but more data can be used, if a miner decides to accept the transaction.
To create a custom `Script` instance, you must rely on the lower-level methods `add` and `prepend`. Both methods accept the same parameter types, and insert an opcode or data at the beginning (`prepend`) or end (`add`) of the `Script`.
`Script` has an easy interface to parse raw scripts from the network or bitcoind, and to extract useful information. An illustrative example (for more options check the API reference)
var raw_script = new Buffer('5221022df8750480ad5b26950b25c7ba79d3e37d75f640f8e5d9bcd5b150a0f85014da2103e3818b65bcc73a7d64064106a859cc1a5a728c4345ff0b641209fba0d90de6e921021f2f6e1e50cb6a953935c3601284925decd3fd21bc445712576873fb8c6ebc1853ae', 'hex');
To validate a transaction, the bitcoin network validates all of its inputs and outputs. To validate an input, the input's script is concatenated with the referenced output script, and the result is executed. If at the end of execution the stack contains a 'true' value, then the transaction is valid. You can do this in `bitcore` by using the `Interpreter` class. The entry point (and probably the only interface you'll need for most applications) is the method `Interpreter#verify()`.
var outputScript = Script('OP_15 OP_ADD OP_16 OP_EQUAL');
var verified = Interpreter().verify(inputScript, outputScript);
// verified will be true
```
Note that `verify` expects two scripts: one is the input script (scriptSig) and the other is the output script (scriptPubkey). This is because different conditions are checked for each.
It also accepts some optional parameters, assuming defaults if not provided: