User-agent: *
Allow: /

# The CMS. The stylesheets and scripts under it are what the site's own pages
# load, so they stay crawlable - blocking them makes Google render the site
# unstyled and judge the mobile experience on that.
Allow: /admin/css/
Allow: /admin/js/
Allow: /admin/editor/css/
Allow: /admin/editor/js/
Allow: /admin/editor/jeditable/
Allow: /admin/plugins/deedsofficeadmin/instantsearch/js/
Disallow: /admin/
Disallow: /scripts/

# Order flows and one-visitor pages. Each of these carries a canonical pointing
# at the product's /search-info/ page and a noindex, so this is belt and braces
# rather than the mechanism - but it keeps crawl budget on the content pages.
Disallow: /payment/
Disallow: /payment-successful/
Disallow: /payment-failed/
Disallow: /payment-cancelled/
Disallow: /banking-details/
Disallow: /cancelsearch/
Disallow: /instant-search-results/

# The map step of an order. Already noindex and already dropped from the index,
# so blocking it now only saves the crawl budget it was spending on six
# parameter variants of the same picker.
Disallow: /map/

# Form endpoints. Not pages, and they answer to POST; they already return
# X-Robots-Tag: noindex (ajax/.htaccess), which is what keeps them out of the
# index - this only stops them being fetched.
Disallow: /ajax/

# Paid search results are written here. Nothing under it is public.
Disallow: /files/

# Not blocked on purpose:
#
#   /samples/           the sample documents carry X-Robots-Tag: noindex, and a
#                       blocked file's noindex is never read, so the copies
#                       already indexed would never come out
#   /instant-search/    the order flows canonicalise to /search-info/ and carry
#                       noindex; Google has to fetch them to see either
#   /contact/thank-you/ noindex on the page record, same reason

Sitemap: https://www.deedsonline.co.za/sitemap.xml