Compressed NFT Augmenting API

How to use

Setup

1. Create a postgres database
2. Copy the server, ctl binaries and the .env.example file to a directory
3. Change the POSTGRES_URI variable in your .env to point to the database
4. Change the PORT as needed
5. Change the ADMIN_* credentials to be used. Ideally, you should generate random ones
6. Change the DEPTH as needed. The maximum number of items can be calculated as 2^DEPTH. So DEPTH = 20 will allow you to have at most 1048576 items in your collection. This API will require changes to its code if you want to have DEPTH > 30
7. Change the DATA_DIR as needed. A small number of .json files will be stored there (1 constantly and 1 per each update)
8. Change TONCENTER_URI as needed. That means removing testnet. if you want to deploy your collection to mainnet
9. cd to the directory where ctl and .env are located
10. Create a file containing the addresses of owners of your items, one address per line. Empty lines will be ignored. The first one will get item index 0, and so on. We will assume that this file is named owners.txt and is located in the same directory
11. Run ./ctl migrate. This will create the necessary tables in the database
12. Run ./ctl add owners.txt. This will add the addresses to the database
13. Host your collection metadata and items' metadata with formats as outlined in Token Data Standard. The items' metadata files must all have a pattern of some-common-uri-part + '/' + item-index + '.json'. Other patterns are possible but will require changes to the API's code
14. Run ./server in a way that prevents it from closing when your SSH (or any other kind of session) closes. You can do that using the screen utility for example. Make sure that the assigned PORT is visible to the public Internet on some endpoint
15. Navigate to api-uri + '/admin/rediscover'. Use your ADMIN_* credentials. If all went well, you should see the string ok and a file should appear under DATA_DIR + '/upd/1.json' (perhaps after some time if the number of items is large)
16. Run ./ctl genupd path-to-update-file collection-owner collection-meta item-meta-prefix royalty-base royalty-factor royalty-recipient api-uri-including-v1 where path-to-update-file is the path to the file mentioned in step 15, collection-owner is the intended collection owner, collection-meta is the full URI to collection metadata file, item-meta-prefix is the common item metadata file prefix (for example, if your item 0 has its metadata hosted at https://example.com/0.json, then you should use https://example.com/ here), royalty-base is the royalty numerator, royalty-factor is the royalty denominator (base = 1 and factor = 100 give 1% royalty), royalty-recipient is the address which will get royalties (you can just use the collection-owner here), and api-uri-including-v1 is the publicly visible API URI with the /v1 postfix (so if you used https://example.com/admin/rediscover to create the update file, you should put https://example.com/v1 here. Using localhost or similar here will not allow users to claim your items, but for testing purposes that's fine)
17. Invoke the ton:// deeplink that appears
18. Navigate to api-uri + '/admin/setaddr/' + collection-address using the address that you saw after step 16
19. Wait for a commited state message in server logs
20. Done

Updating

1. Prepare a list of owners to be newly added as described in step 11 of the Setup section. If you previously added 100 owners, then these new owners will have items starting with index 100 and so on. We will assume that this file is named new-owners.txt and is located in the same directory as the ctl binary
2. Run ./ctl add new-owners.txt
3. Navigate to api-uri + '/admin/rediscover'
4. Locate the newly created update file under DATA_DIR + '/upd'. If your latest applied update was update 1 (as after setup), then the newly created one will have the name 2.json
5. Run ./ctl genupd path-to-update-file collection-address where path-to-update-file is the path to the file mentioned in step 4, and collection-address is the address of the deployed collection
6. Invoke the ton:// deeplink that appears
7. Wait for a commited state message in server logs
8. Done

NOTE: During the brief period when the onchain transaction to update the collection has happened, but the API has not detected it yet, all generated proofs will be invalid and therefore claim requests generated during this period will fail. Therefore, we do not recommend updating your collection under large traffic (or often). Instead, try updating your collection with large batches and when under little traffic.

License

MIT