Add limit 1 to domain_exists

[spacedmonkey]

domain_exists uses to get_sites to looks if site exists. But it only uses the first result. The param limit should be added to get_sites, limiting the query.

[danieltj]

Adds number parameter to limit by 1

[danieltj]

Added a parameter to fetch only one results back by adding number to arguments in 42072.diff.

[flixos90]

We should consider simplifying the function further, by internally using the new get_site_by(). See https://core.trac.wordpress.org/ticket/40180#comment:24

• The advantage would be less code to maintain.
• The disadvantage would be that get_site_by() returns a full object, while we only need an ID here. Not sure that's a real issue though, since it only queries for one result anyway, so it shouldn't have a significant impact on performance.

[jeremyfelt]

I think we're okay to use get_site_by() here. While a full object is returned, domain_exists() is not widely used and because only one result is being returned, the performance difference is negligible. See 42072.2.diff​. Tests exist for domain_exists() and those continue to pass.

[jeremyfelt]

42072.3.diff​ ensures we're passing an ID to the filter as expected.

[jeremyfelt]

In 42072.4.diff​:

• Ensure the URL passed to get_site_by() is formatted with expected slashes.
• Remove the www.wordpress.org site from tests. Site creation was failing because domain_exists() would return true after get_site_by( 'url', 'www.wordpress.org/' ); found wordpress.org. We don't support the co-existence of www and non-www domains, so it should be okay to apply this restriction to domain_exists() now.

[jeremyfelt]

I'm not feeling confident enough in domain_exists() changing behavior and treating www/non-www the same. For 4.9, let's fix the bug by just adding the count of 1.

[jeremyfelt]

In 41736:

Multisite: Only query for one site in domain_exists().

get_sites() queries for a maximum of 100 records by default. In domain_exists(), we only use one.

Props danieltj, spacedmonkey.
Fixes #42072.

[jeremyfelt]

[41736] is the commit and fix for #42073, get_blog_id_from_url(), not domain_exists().

[jeremyfelt]

In 41738:

Multisite: Only query for one site in domain_exists().

get_sites() queries for a maximum of 100 records by default. In domain_exists(), we only use one.

A previous commit, [41736], has this same commit message but applies to get_blog_id_from_url() and #42073 instead.

Props danieltj, spacedmonkey.
Fixes #42072.