perf(scraper): use direct HTTP for chapter text fetching, bypass Browserless

novelfire.net chapter content is server-rendered, so Browserless is not
needed. Add a dedicated chapterClient (always StrategyDirect) to Scraper
and use it in ScrapeChapterText, removing the now-irrelevant WaitFor /
RejectResourceTypes / GotoOptions fields from the ContentRequest.

scraper/cmd/scraper/main.go

@@ -12,7 +12,7 @@
 //
 //	BROWSERLESS_URL             Browserless base URL           (default: http://localhost:3030)
 //	BROWSERLESS_TOKEN           Browserless API token          (default: "")
-//	BROWSERLESS_STRATEGY        content | scrape | cdp         (default: content)
+//	BROWSERLESS_STRATEGY        content | scrape | cdp | direct (default: direct; chapters always use direct HTTP)
 //	BROWSERLESS_MAX_CONCURRENT  Max simultaneous browser sessions  (default: 5)
 //	SCRAPER_WORKERS             Chapter goroutine count        (default: NumCPU)
 //	SCRAPER_HTTP_ADDR           HTTP listen address            (default: :8080)
@@ -95,6 +95,8 @@ func run(log *slog.Logger) error {
 	urlStrategy := browser.Strategy(strings.ToLower(envOr("BROWSERLESS_URL_STRATEGY", string(browser.StrategyContent))))
 	bc := newBrowserClient(strategy, browserCfg)
 	urlClient := newBrowserClient(urlStrategy, browserCfg)
+	// Chapter text is server-rendered on novelfire.net — direct HTTP is faster and avoids Browserless.
+	chapterClient := browser.NewDirectHTTPClient(browserCfg)
 
 	// ── Storage backends ────────────────────────────────────────────────────
 	minioCfg := storage.MinioConfig{
@@ -119,7 +121,7 @@ func run(log *slog.Logger) error {
 		return fmt.Errorf("storage init failed: %w", err)
 	}
 
-	nf := novelfire.New(bc, log, urlClient, store)
+	nf := novelfire.New(bc, log, urlClient, chapterClient, store)
 
 	workers := 0
 	if s := os.Getenv("SCRAPER_WORKERS"); s != "" {