Nhảy đến nội dung chính

Rapid Prototyping

When you change the interface of a contract and re-deploy it, you may see this error:

Cannot deserialize the contract state.

Why does this happen?

When your contract is executed, the NEAR Runtime reads the serialized state from disk and attempts to load it using current contract code. When your code changes but the serialized state stays the same, it can't figure out how to do this.

How can you avoid such errors?

When you're still in the Research & Development phase, building a prototype and deploying it locally or on testnet, you can just delete all previous contract state when you make a breaking change. See below for a couple ways to do this.

When you're ready to deploy a more stable contract, there are a couple of production strategies that will help you update the contract state without deleting it all. And once your contract graduates from "trusted mode" (when maintainers control a Full Access key) to community-governed mode (no more Full Access keys), you can set up your contract to upgrade itself.

Rapid Prototyping: Delete Everything All The Time

There are two ways to delete all account state:

  1. Deploying on a new account each time
  2. Deleting & recreating contract account

For both cases, let's consider the following example.

The rust-status-message example contract has the following structure:

src/lib.rs
loading...

Let's say you deploy this contract to testnet, then call it with:

near call [contract] set_status '{"message": "lol"}' --accountId you.testnet
near view [contract] get_status '{"account_id": "you.testnet"}'

This will return the message that you set with the call to set_status, in this case "lol".

At this point the contract is deployed and has some state.

Now let's say you change the contract to store two kinds of data for each account:

#[near_bindgen]
#[derive(BorshDeserialize, BorshSerialize)]
pub struct StatusMessage {
taglines: LookupMap<AccountId, String>,
bios: LookupMap<AccountId, String>,
}

impl Default for StatusMessage {
fn default() -> Self {
Self {
taglines: LookupMap::new(b"r"),
bios: LookupMap::new(b"b"),
}
}
}

#[near_bindgen]
impl StatusMessage {
pub fn set_tagline(&mut self, message: String) {
let account_id = env::signer_account_id();
self.taglines.insert(&account_id, &message);
}

pub fn get_tagline(&self, account_id: AccountId) -> Option<String> {
return self.taglines.get(&account_id);
}

pub fn set_bio(&mut self, message: String) {
let account_id = env::signer_account_id();
self.bios.insert(&account_id, &message);
}

pub fn get_bio(&self, account_id: AccountId) -> Option<String> {
return self.bios.get(&account_id);
}
}

You build & deploy the contract again, thinking that maybe because the new taglines LookupMap has the same prefix as the old records LookupMap (the prefix is r, set by LookupMap::new(b"r".to_vec())), the tagline for you.testnet should be "lol". But when you near view the contract, you get the "Cannot deserialize" message. What to do?

1. Deploying on a new account each time

When first getting started with a new project, the fastest way to deploy a contract is creating an account and deploying the contract into it using NEAR CLI.

  near create-account <account-id> --useFaucet
near deploy <account-id> ./path/to/compiled.wasm

This does a few things:

  1. Creates a new testnet account pre-funded with 10N from the faucet
  2. Stores the private key for this account in the ~/.near-credentials folder
  3. Deploys your contract code to this account

2. Deleting & Recreating Contract Account

Another option to start from scratch is to delete the account and recreate it.

Delete sub-account
near delete app-name.you.testnet you.testnet

This sends all funds still on the app-name.you.testnet account to you.testnet and deletes the contract that had been deployed to it, including all contract state.

Now you create the sub-account and deploy to it again using the commands above, and it will have empty state like it did the first time you deployed it.

Was this page helpful?