RFC: Manipulate the generated file

[Hebilicious]

Summary

It would be great if we could manipulate the generated file in some ways, such as appending/or preprending code, or executing code after the generation with some kind of hook.

Use case :

The introspected file has

/* eslint-disable */
/* prettier-ignore */

appended to it, but I'm currently using another linter (dprint) that needs dprint-ignore. Currently I'm patching the internal package to add stuff there, but that's not ideal.

Proposed Solution

I haven't thought too much about it, but there's 2 things that could work :

1. Add a pre and post settings in the schemas property of the plugin, so we can pass abitrary code there.
   This would be good for a simple use case. For example a linter use case you'd do something like this :

    "plugins": [
      {
        "name": "gql.tada/ts-plugin",
        "schemas": [
          {
            "name": "my-api",
            "schema": "./src/graphql/my-api.graphql",
            "tadaOutputLocation": "./src/graphql/my-api_env.d.ts",
            "tadaTurboLocation": "./src/graphql/my-api_cache_env.d.ts",
			"pre": "// dprint-ignore-file \n"
          }
        ]
      }

2. Add a beforeGeneration and afterGeneration hook, that will run a command with child_process.
   This would be very flexible, for the linter use case it would allow to run a linter after each generation.

    "plugins": [
      {
        "name": "gql.tada/ts-plugin",
        "schemas": [
          {
            "name": "my-api",
            "schema": "./src/graphql/my-api.graphql",
            "tadaOutputLocation": "./src/graphql/my-api_env.d.ts",
            "tadaTurboLocation": "./src/graphql/my-api_cache_env.d.ts",
			"afterGeneration": "bun run ./scripts/lint-gql-tada-output.ts"
          }
        ]
      }

Requirements