science.alt.dataset.storageS3

{
  "id": "science.alt.dataset.storageS3",
  "defs": {
    "main": {
      "type": "object",
      "required": [
        "bucket",
        "shards"
      ],
      "properties": {
        "bucket": {
          "type": "string",
          "maxLength": 255,
          "description": "S3 bucket name"
        },
        "region": {
          "type": "string",
          "maxLength": 50,
          "description": "AWS region (e.g., 'us-east-1'). Optional for S3-compatible services."
        },
        "shards": {
          "type": "array",
          "items": {
            "ref": "#shardEntry",
            "type": "ref"
          },
          "minLength": 1,
          "description": "Array of shard entries with object key and integrity checksum"
        },
        "endpoint": {
          "type": "string",
          "format": "uri",
          "maxLength": 500,
          "description": "Custom S3-compatible endpoint URL (e.g., for MinIO, Cloudflare R2). Omit for standard AWS S3."
        }
      },
      "description": "S3 or S3-compatible storage for WebDataset tar archives. Supports custom endpoints for MinIO, Cloudflare R2, and other S3-compatible services."
    },
    "shardEntry": {
      "type": "object",
      "required": [
        "key",
        "checksum"
      ],
      "properties": {
        "key": {
          "type": "string",
          "maxLength": 1024,
          "description": "S3 object key for this WebDataset tar shard"
        },
        "checksum": {
          "ref": "science.alt.dataset.entry#shardChecksum",
          "type": "ref",
          "description": "Content hash for integrity verification"
        }
      },
      "description": "A single S3 object shard with integrity checksum"
    }
  },
  "$type": "com.atproto.lexicon.schema",
  "lexicon": 1
}