Add structured data extractor (JSON-LD, OpenGraph, Twitter Card)

[justinhsu1477]

Have you searched if there an existing feature request for this?

• I have searched the existing requests

Feature description

Most modern sites embed structured metadata (<script type="application/ld+json">, <meta property="og:*">, Twitter Cards, microdata) for SEO and social sharing. This data is intentionally stable across UI redesigns, which aligns well with Scrapling's adaptive-by-default philosophy.

Today users have to parse it manually:

import json

scripts = page.css('script[type="application/ld+json"]::text').getall()
data = []
for s in scripts:
    try:
        parsed = json.loads(s)
        if isinstance(parsed, list):
            data.extend(parsed)
        elif "@graph" in parsed:
            data.extend(parsed["@graph"])
        else:
            data.append(parsed)
    except json.JSONDecodeError:
        pass

og = {
    m.attrib["property"].replace("og:", ""): m.attrib.get("content")
    for m in page.css('meta[property^="og:"]')
}

Proposed first-class API on Selector:

page.json_ld()          # list[dict]; flattens @graph; tolerant of malformed JSON
page.opengraph()        # dict of og:* meta
page.twitter_card()     # dict of twitter:* meta
page.microdata()        # list[dict] parsed from itemscope/itemprop
page.structured_data()  # everything above, grouped by source
page.metadata()         # normalized summary: {title, description, image, type, ...} fused across sources

Why this fits Scrapling:

• Aligns with the existing ROADMAP item "Add the ability to auto-detect schemas in pages and manipulate them".
• No new heavy deps (json is stdlib; microdata can be done with the existing lxml).
• Pure addition on Selector — doesn't touch fetchers or the adaptive parser.

Reference implementations in the ecosystem: extruct, metascraper.

Happy to send a PR against dev if there's interest — would love early feedback on naming / scope before implementing.