Enable use of wildcard location configurations

[buchdag]

This PR is a fix for #561

It makes the location configuration work as follow. Let say the domain is foo.bar.baz.example.tld:

• if a config file with the exact same name as your domain (foo.bar.baz.example.tld) is present, use it.
• else try every corresponding wildcard domain, from most qualified to least qualified, first ascending through the labels to the base domain (*.bar.baz.example.tld, *.baz.example.tld, *.example.tld) then descending (foo.bar.baz.example.*, foor.bar.baz.*, foo.bar.*, foo.*). The first matching config file will be used.
• if no wildcard location config file was found either, use the default location config file.

[buchdag]

plus corresponding tests

[jayhar]

Previously to achieve a wildcard SSL for all subdomains I have been using the following example. Two A records are needed to make this work. Note I do not force redirect from non-www to www.

docker run -d -e VIRTUAL_HOST=.example.com -e LETSENCRYPT_HOST=www.example.com,example.com -e LETSENCRYPT_EMAIL=example@example.com --network=exampleproxy --name example.com -v /srv/www/nginx-data/html/example.com:/usr/local/apache2/htdocs httpd:alpine

With the changes you made (thank you for your work) am I am able to to change my current syntax to -e VIRTUAL_HOST=example.com -e LETSENCRYPT_HOST=example.com to achieve the same scenario? The issue I had before is the NGINX config had to manually modified to make this work.