query-orchestration/cmd/docCleanRunner/main.go

// Listens For: Body Containing Document ID.
//
// Action:
// If the document is not already clean according to the collector standards, a clean is triggered.
// Note: Text extraction has been removed. The document pipeline ends after cleaning.
// See remove_texttract_and_mocks_plan.md for details.
package main

import (
	"context"
	"log/slog"
	"os"

	doccleanrunner "queryorchestration/api/docCleanRunner"
	documentclean "queryorchestration/internal/document/clean"
	"queryorchestration/internal/server/runner"
	"queryorchestration/internal/serviceconfig/build"
	"queryorchestration/internal/serviceconfig/objectstore"

	_ "github.com/lib/pq"
)

// DocCleanConfig provides configuration for the document cleaning runner.
// Note: documenttext.DocTextConfig has been removed since text extraction is deprecated.
type DocCleanConfig struct {
	runner.BaseConfig[doccleanrunner.Body]
	objectstore.ObjectStoreConfig
}

func main() {
	// Print version information before any environment checks
	build.PrintVersionInfo("docCleanRunner")

	ctx := context.Background()

	cfg := &DocCleanConfig{}

	cfg.ControllerFunc = func() runner.Controller[doccleanrunner.Body] {
		clean := documentclean.New(cfg)

		return doccleanrunner.New(&doccleanrunner.Services{
			Clean: clean,
		})
	}

	server, err := runner.New(ctx, cfg)
	if err != nil {
		slog.Error(err.Error())
		os.Exit(1)
	}

	err = cfg.SetStoreClient(ctx)
	if err != nil {
		slog.Error(err.Error())
		os.Exit(1)
	}

	server.Listen(ctx)
}