Fixed a bug regarding using custom port for confluence

[wooyeong]

When a confluence server is hosted on custom web port, Alfred Confluence gives quick search results well, but navigates to wrong URL for found pages.

If the confluence address is https://wiki.example.com:8443, Alfred Confluence trys to connect https://wiki.example.com/dosearchsite.action?queryString=[query], :8443 was gone.

Here are screenshots.

---

• Pages found

• Port number has gone on navigation, so unable to see content.

---

To fix this, this pull request changes a line of code, replacing urlparse.hostname to urlparse.netloc. I think there's no side effect on this change, here's a simple test.

# cat test.py
from urlparse import urlparse

urls = ['https://c.mins.site:8443/', "https://a.b.c/", "http://confluence.example.com", "http://d.e.f:223"]

for url in urls:

    parsed = urlparse(url)

    print(parsed)
    print(parsed.hostname)
    print(parsed.netloc)

ParseResult(scheme='https', netloc='c.mins.site:8443', path='/', params='', query='', fragment='')
c.mins.site
c.mins.site:8443
ParseResult(scheme='https', netloc='a.b.c', path='/', params='', query='', fragment='')
a.b.c
a.b.c
ParseResult(scheme='http', netloc='confluence.example.com', path='', params='', query='', fragment='')
confluence.example.com
confluence.example.com
ParseResult(scheme='http', netloc='d.e.f:223', path='', params='', query='', fragment='')
d.e.f
d.e.f:223

I've tested on both single instance configuration and multiple instances configuration using .alfred-confluence.json.

[skleinei]

Thanks for that. I've manually made this change.