fix(#377): migrate url_import to firecrawl-py 2.x API

[NoveliaYuki]

Purpose

URL imports fail at runtime with AttributeError: 'Firecrawl' object has no attribute 'crawl_url'. The plugin was written against firecrawl-py 0.x; the dependency currently in the library-manager container is firecrawl-py 4.25.2, which renamed the entry points and reshaped the response objects.

Changes

• Replace app.crawl_url(url, params={...}, poll_interval=5) with app.crawl(url, **kwargs, scrape_options=ScrapeOptions(formats=["markdown"]), poll_interval=5) — keyword arguments now carry limit, max_discovery_depth, crawl_entire_domain, allow_external_links directly.
• Per-page metadata extraction handles both the new Pydantic DocumentMetadata (attribute access on source_url/title) and the legacy dict form (camelCase sourceURL) for forward compatibility.
• No behavior change to the zero-pages fallback, progress reporting, or surrounding plugin scaffolding.

The migration is dependency-only — no public API of the plugin changes, no other call sites needed updates.

Verification

• library-manager/tests/ — 52/52 passing against the patched plugin.
• Live smoke against a self-hosted Firecrawl instance: queue an import of https://example.com, item reaches status ready, content endpoint returns non-empty markdown.

Related

• Closes fix(#377): migrate url_import to firecrawl-py 2.x API #377