opentextsummaries/app/Jobs/ImportBookFromGutendex.php

<?php

namespace App\Jobs;

use App\Models\Author;
use App\Models\Book;
use App\Models\Category;
use App\Services\GutendexClient;
use Illuminate\Bus\Queueable;
use Illuminate\Contracts\Queue\ShouldQueue;
use Illuminate\Foundation\Bus\Dispatchable;
use Illuminate\Queue\InteractsWithQueue;
use Illuminate\Queue\SerializesModels;
use Illuminate\Support\Facades\DB;
use Illuminate\Support\Facades\Log;
use Illuminate\Support\Str;

class ImportBookFromGutendex implements ShouldQueue
{
    use Dispatchable, InteractsWithQueue, Queueable, SerializesModels;

    public int $gutenbergId;
    public bool $generateSummary;

    /**
     * The number of times the job may be attempted.
     *
     * @var int
     */
    public $tries = 3;

    /**
     * The number of seconds to wait before retrying the job.
     *
     * @var int
     */
    public $backoff = 60;

    /**
     * Create a new job instance.
     */
    public function __construct(int $gutenbergId, bool $generateSummary = true)
    {
        $this->gutenbergId = $gutenbergId;
        $this->generateSummary = $generateSummary;
        $this->onQueue('gutendex-import');
    }

    /**
     * Execute the job.
     */
    public function handle(GutendexClient $gutendexClient): void
    {
        Log::info('ImportBookFromGutendex: Starting import', [
            'gutenberg_id' => $this->gutenbergId,
        ]);

        try {
            // Check if book already exists
            $existingBook = Book::where('gutenberg_id', $this->gutenbergId)->first();
            if ($existingBook) {
                Log::info('ImportBookFromGutendex: Book already exists, skipping', [
                    'gutenberg_id' => $this->gutenbergId,
                    'book_id' => $existingBook->id,
                ]);
                return;
            }

            // Fetch book metadata from Gutendex
            $gutendexBook = $gutendexClient->getBook($this->gutenbergId);

            if (!$gutendexBook) {
                Log::error('ImportBookFromGutendex: Failed to fetch book from Gutendex', [
                    'gutenberg_id' => $this->gutenbergId,
                ]);
                $this->fail(new \Exception("Failed to fetch book {$this->gutenbergId} from Gutendex"));
                return;
            }

            // Only import public domain books
            if (!$gutendexBook->isPublicDomain()) {
                Log::warning('ImportBookFromGutendex: Book is not public domain, skipping', [
                    'gutenberg_id' => $this->gutenbergId,
                    'copyright' => $gutendexBook->copyright,
                ]);
                return;
            }

            // Check if full text is available
            if (!$gutendexBook->getPlainTextUrl()) {
                Log::warning('ImportBookFromGutendex: No plain text URL available, skipping', [
                    'gutenberg_id' => $this->gutenbergId,
                ]);
                return;
            }

            DB::beginTransaction();

            try {
                // Find or create category
                $category = $this->findOrCreateCategory($gutendexBook->suggestCategory());

                // Create book record
                $bookData = $gutendexBook->toBookArray();
                $bookData['category_id'] = $category->id;
                $bookData['slug'] = $this->generateUniqueSlug($gutendexBook->title);

                $book = Book::create($bookData);

                // Map authors to the authors table
                $this->syncAuthors($book, $gutendexBook->getAuthorNames());

                DB::commit();

                Log::info('ImportBookFromGutendex: Book imported successfully', [
                    'gutenberg_id' => $this->gutenbergId,
                    'book_id' => $book->id,
                    'title' => $book->title,
                ]);

                // Dispatch summary generation job
                if ($this->generateSummary) {
                    GenerateSummary::dispatch($book->id)
                        ->delay(now()->addSeconds(5)); // Small delay to spread load
                }
            } catch (\Exception $e) {
                DB::rollBack();
                throw $e;
            }
        } catch (\Exception $e) {
            Log::error('ImportBookFromGutendex: Import failed', [
                'gutenberg_id' => $this->gutenbergId,
                'error' => $e->getMessage(),
                'trace' => $e->getTraceAsString(),
            ]);
            throw $e;
        }
    }

    /**
     * Find or create a category.
     *
     * @param string|null $categorySlug
     * @return Category
     */
    protected function findOrCreateCategory(?string $categorySlug): Category
    {
        if (!$categorySlug) {
            $categorySlug = 'uncategorized';
        }

        $category = Category::where('slug', $categorySlug)->first();

        if (!$category) {
            $categoryName = Str::title(str_replace('-', ' ', $categorySlug));
            $category = Category::create([
                'name' => $categoryName,
                'slug' => $categorySlug,
                'description' => "Books in the {$categoryName} category",
            ]);

            Log::info('ImportBookFromGutendex: Created new category', [
                'category_id' => $category->id,
                'slug' => $categorySlug,
            ]);
        }

        return $category;
    }

    /**
     * Generate a unique slug for the book.
     *
     * @param string $title
     * @return string
     */
    protected function generateUniqueSlug(string $title): string
    {
        $slug = Str::slug($title);
        $originalSlug = $slug;
        $counter = 1;

        while (Book::where('slug', $slug)->exists()) {
            $slug = $originalSlug . '-' . $counter;
            $counter++;
        }

        return $slug;
    }

    /**
     * Sync authors to the book.
     *
     * @param Book $book
     * @param array $authorNames
     * @return void
     */
    protected function syncAuthors(Book $book, array $authorNames): void
    {
        $authorIds = [];

        foreach ($authorNames as $authorName) {
            $author = Author::firstOrCreate(
                ['name' => $authorName],
                ['slug' => Str::slug($authorName)]
            );
            $authorIds[] = $author->id;
        }

        $book->authors()->sync($authorIds);

        Log::info('ImportBookFromGutendex: Authors synced', [
            'book_id' => $book->id,
            'author_count' => count($authorIds),
        ]);
    }

    /**
     * Handle a job failure.
     */
    public function failed(\Throwable $exception): void
    {
        Log::error('ImportBookFromGutendex: Job failed permanently', [
            'gutenberg_id' => $this->gutenbergId,
            'error' => $exception->getMessage(),
        ]);
    }
}