// Package documentclean provides document cleaning functionality.
// Note: Text extraction pipeline has been removed. The document processing
// pipeline now ends after cleaning. See remove_texttract_and_mocks_plan.md.
package documentclean

import (
	"context"
	"fmt"

	"github.com/google/uuid"
)

// CleanResult contains the result of the clean operation.
//
// Fields:
//   - Passed: true if the document passed all validation checks
//   - FailReason: the specific failure reason if Passed is false, nil otherwise
type CleanResult struct {
	Passed     bool
	FailReason *string
}

// Clean processes a document by cleaning it. Previously this also triggered
// text extraction, but that feature has been removed.
//
// Parameters:
//   - ctx: request context
//   - documentId: the document UUID to clean
//
// Returns:
//   - *CleanResult: result indicating whether validation passed and any failure reason
//   - error: only for infrastructure errors (DB, S3). Validation failures are
//     returned via CleanResult.Passed=false, not as errors.
func (s *Service) Clean(ctx context.Context, documentId uuid.UUID) (*CleanResult, error) {
	isclean, err := s.cfg.GetDBQueries().HasDocumentCleanEntry(ctx, documentId)
	if err != nil {
		return nil, fmt.Errorf("unable to verify if document has been cleaned: %w", err)
	}

	if !isclean {
		result, err := s.clean(ctx, documentId)
		if err != nil {
			return nil, fmt.Errorf("unable to clean document: %w", err)
		}
		return result, nil
	}

	// Note: Text extraction has been removed. Document pipeline ends here.
	return &CleanResult{Passed: true}, nil
}