rules.go 821 B

12345678910111213141516171819202122
  1. // Copyright 2017 Frédéric Guillot. All rights reserved.
  2. // Use of this source code is governed by the Apache 2.0
  3. // license that can be found in the LICENSE file.
  4. package scraper
  5. // List of predefined scraper rules (alphabetically sorted)
  6. // domain => CSS selectors
  7. var predefinedRules = map[string]string{
  8. "github.com": "article.entry-content",
  9. "igen.fr": "section.corps",
  10. "lemonde.fr": "div#articleBody",
  11. "lesjoiesducode.fr": ".blog-post-content img",
  12. "linux.com": "div.content, div[property]",
  13. "medium.com": ".section-content",
  14. "opensource.com": "div[property]",
  15. "osnews.com": "div.newscontent1",
  16. "phoronix.com": "div.content",
  17. "techcrunch.com": "div.article-entry",
  18. "theregister.co.uk": "#body",
  19. "wired.com": "main figure, article",
  20. }