Invoking the generalized "call-block-logs" transformation strategy twice in a row to filter for different events would cause them all to be written into the same file

[TimDaub]

• e.g. with this blueprint we've filtering all event logs from singleton music nft contracts: https://github.com/neume-network/blueprints/blob/cf8eacac6199c7c379420f62e0e57caa11fb3931/src/filter-mint-in-singleton-music-nft-contracts/step.mjs
• but with another blueprint https://github.com/neume-network/blueprints/blob/cf8eacac6199c7c379420f62e0e57caa11fb3931/src/soundxyz-filter-created-artist/step.mjs we're filtering out CreatedArtist events from soundxyz.
• And if we'd run both of these one after each other, all their results would go to "datadir/call-block-logs-transformation"
• What has changed is that the strategies are now general purpose retrieval functions that can be controlled with their arguments, whereas blueprints are more opinionated and feature actual platform business logic
• So a blueprint should be allowed to e.g. control the file name a transformer can write to, e.g. in the crawl path

{
  name: "call-block-logs",
  transformer: {
    args: ["a", 1],
    outputFile: "hello-world"
  }
}