Migrate from NEAR Lake framework

In this article you'll learn how to migrate your NEAR Lake Framework JavaScript indexer to Near QueryAPI, a fully managed solution to build indexer functions, extract on-chain data, store it in a database, and be able to query it using GraphQL endpoints.

Supported languages

Currently QueryAPI only supports JavaScript, so if your indexer code uses TypeScript, Python, or Rust, you'll have to re-write your indexing logic in JS before you can migrate it.

Basic migration

build/near-data-infrastructure/lake-framework/building-indexers/js-lake-indexer Let's take a basic JS indexer built with NEAR Lake Framework as an example. This indexer simply logs the Block height and the number of shards for each indexed block, using an indexer handler function handleStreamerMessage.

Migrating this basic indexer to QueryAPI is simple. You only need to migrate the code from the handleStreamerMessage function:

async function handleStreamerMessage(streamerMessage: types.StreamerMessage): Promise<void> {
    console.log(`
        Block #${streamerMessage.block.header.height}
        Shards: ${streamerMessage.shards.length}
    `);
}

Migrating to QueryAPI

1. To start the migration process, create a new indexer using QueryAPI. You should see a similar interface like this:

2. Since QueryAPI keeps a compatibility layer with Lake Framework, you don't need to change any references to streamerMessage in your indexer function. Just change the function definition to:

function handleStreamerMessage(streamerMessage) {
    // ... Lake framework indexer code
}

3. Next, add your migrated indexer function to the getBlock(...) method, and simply call your function passing block.streamerMessage as parameter:

async function getBlock(block: Block) {
  // Add your code here

  function handleStreamerMessage(streamerMessage) {
    console.log(`
        Block #${streamerMessage.block.header.height}
        Shards: ${streamerMessage.shards.length}
    `);
    }

  handleStreamerMessage(block.streamerMessage);
}

That's all! The basic Lake Framework JS indexer has been migrated to QueryAPI, and you can test it out by using Debug Mode. If you run the indexer using local debug mode, you should see an output like:

    Block #106812523
    Shards: 4

Database storage

If you want to take advantage of QueryAPI's database features, you can also store the indexer results in a Postgres DB.

1. First, create the database schema:

CREATE TABLE
  "basic" (
    "block_height" BIGINT NOT NULL,
    "shards" INTEGER NOT NULL,
    PRIMARY KEY ("block_height")
  );

2. In your indexer JavaScript code, use the context.db object to store the results:

    const basicData = {
      block_height: streamerMessage.block.header.height,
      shards: streamerMessage.shards.length,
    };

    context.db.Basic.insert(basicData);

Advanced migration

For this example, let's take the TypeScript NFT indexer built with NEAR Lake Framework as reference. This indexer is watching for nft_mint Events and prints some relevant data about minted NFTs.

As with the previous example, moving this NFT indexer to QueryAPI requires to migrate the code from the handleStreamerMessage function. But since it was done in TypeScript, it also needs some additional work as it needs to re-written in JavaScript.

Migrating to QueryAPI

To migrate the code, you can take advantage of the near-lake-primitives provided by QueryAPI, and simplify the indexer logic. For example:

• to get all Events in a Block, you can simply call block.events().
• you don't need to iterate through shards and execution outcomes, nor manually parse the EVENT_JSON logs to detect events (it's handled by QueryAPI)

Here's a JavaScript implementation of the NFT indexer using QueryAPI features, that provides the same output as the original indexer:

async function getBlock(block: Block) {
  let output = [];

  for (let ev of block.events()) {
    const r = block.actionByReceiptId(ev.relatedReceiptId);
    const createdOn = new Date(block.streamerMessage.block.header.timestamp / 1000000);

    try {
      let event = ev.rawEvent;

      if (event.standard === "nep171" && event.event === "nft_mint") {
        let nfts = [];
        let marketplace = "unknown";

        if (r.receiverId.endsWith(".paras.near"))
        {
          marketplace = "Paras";
          nfts = event.data.map(eventData => ({
                owner: eventData.owner_id,
                links: eventData.token_ids.map(
                  tokenId => `https://paras.id/token/${r.receiverId}::${tokenId.split(":")[0]}/${tokenId}`)
                })
              );
        }
        else if (r.receiverId.match(/\.mintbase\d+\.near$/))
        {
          marketplace = "Mintbase";
          nfts = event.data.map(eventData => {
            const memo = JSON.parse(eventData.memo)
            return {
              owner: eventData.owner_id,
              links: [`https://mintbase.io/thing/${memo["meta_id"]}:${r.receiverId}`]
            }
          });
        }

        output.push({
          receiptId: ev.relatedReceiptId,
          marketplace,
          createdOn,
          nfts,
        });

      }
    } catch (e) {
      console.log(e);
    }
  }

  if (output.length) {
    console.log(`We caught freshly minted NFTs!`);
    console.dir(output, { depth: 5 });
  }
}

That's all! The NFT indexer has been migrated to QueryAPI, and you can test it out by using Debug Mode. If you run the indexer using local debug mode, you should see an output like:

Block Height #66264722

We caught freshly minted NFTs!

[
  {
    "receiptId": "BAVZ92XdbkAPX4DkqW5gjCvrhLX6kGq8nD8HkhQFVt5q",
    "marketplace": "Mintbase",
    "createdOn": "2022-05-24T09:36:00.411Z",
    "nfts": [
      {
        "owner": "chiming.near",
        "links": [
          "https://mintbase.io/thing/HOTcn6LTo3qTq8bUbB7VwA1GfSDYx2fYOqvP0L_N5Es:vnartistsdao.mintbase1.near"
        ]
      }
    ]
  }
]

Database storage

If you want to take advantage of QueryAPI's database features, you can also store the indexer results in a Postgres DB.

1. First, create the database schema:

CREATE TABLE
  "nfts" (
    "id" SERIAL NOT NULL,
    "marketplace" TEXT,
    "block_height" BIGINT,
    "timestamp" DATETIME,
    "receipt_id" TEXT,
    "nft_data" TEXT,
    PRIMARY KEY ("id", "block_height")
  );

2. In your indexer JavaScript code, use the context.db object to store the results:

// ... previous code ...
        output.push({
          receiptId: ev.relatedReceiptId,
          marketplace,
          createdOn,
          nfts,
        });

        const nftMintData = {
          marketplace: marketplace,
          block_height: block.header().height,
          timestamp: createdOn,
          receipt_id: r.receiptId,
          nft_data: JSON.stringify(event.data),
        };

        context.db.Nfts.insert(nftMintData);
      }
    } catch (e) {
      console.log(e);
    }
  }
// ... code continues ...

tip

You can find the migrated NFT indexer source code by clicking here.