feat: Add nexojornal.com.br custom parser

[jocmp]

Adds a custom Mercury parser for Nexo Jornal, a Brazilian Portuguese news outlet, to support jocmp/capyreader#1914.

Notes

Nexo Jornal is a Next.js single-page app: the article body is rendered client-side from the __NEXT_DATA__ JSON blob, and Mercury strips <script> tags before extraction runs. As a result, the parser relies entirely on the head meta tags that Mercury normalizes (property -> name, content -> value):

• title: og:title (the title cleaner strips the " - Nexo Jornal" suffix)
• dek: og:description
• date_published: articlePublishTime
• lead_image_url: og:image
• author: null (the only available editor meta tag is the section editor, not the article author)
• content: #__next selector (empty server-side, so Mercury returns null with fallback: false)

This still gives Capyreader clean title, description, image, and date for items in the RSS feed.

Test plan

• npx jest src/extractors/custom/www.nexojornal.com.br/index.test.js passes