Make URL part joining aware of absolute URLs.

Previously, with RELATIVE_URLS disabled, when both SITEURL and
STATIC_URL were absolute, the final generate data URLs looked wrong like
this (two absolute URLs joined by `/`):

    http://your.site/http://static.your.site/image.png

With this patch, the data URLs are correctly:

    http://static.your.site/image.png

This also applies to all *_URL configuration options (for example,
ability to have pages and articles on different domains) and behaves
like one expects even with URLs starting with just `//`, thanks to
making use of urllib.parse.urljoin().

However, when RELATIVE_URLS are enabled, urllib.parse.urljoin() doesn't
handle the relative base correctly. In that case, simple os.path.join()
is used. That, however, breaks the above case, but as RELATIVE_URLS are
meant for local development (thus no data scattered across multiple
domains), I don't see any problem.

Just to clarify, this is a fully backwards-compatible change, it only
enables new use cases that were impossible before.

pelican/contents.py

@@ -11,7 +11,7 @@ import sys
 import pytz
 
 import six
-from six.moves.urllib.parse import urlparse, urlunparse
+from six.moves.urllib.parse import urljoin, urlparse, urlunparse
 
 from pelican import signals
 from pelican.settings import DEFAULT_CONFIG
@@ -234,6 +234,25 @@ class Content(object):
         path = value.path
         origin = m.group('path')
 
+        # urllib.parse.urljoin() produces `a.html` for urljoin("..", "a.html")
+        # so if RELATIVE_URLS are enabled, we fall back to os.path.join() to
+        # properly get `../a.html`. However, os.path.join() produces
+        # `baz/http://foo/bar.html` for join("baz", "http://foo/bar.html")
+        # instead of correct "http://foo/bar.html", so one has to pick a side
+        # as there is no silver bullet.
+        if self.settings['RELATIVE_URLS']:
+            joiner = os.path.join
+        else:
+            joiner = urljoin
+
+            # However, it's not *that* simple: urljoin("blog", "index.html")
+            # produces just `index.html` instead of `blog/index.html` (unlike
+            # os.path.join()), so in order to get a correct answer one needs to
+            # append a trailing slash to siteurl in that case. This also makes
+            # the new behavior fully compatible with Pelican 3.7.1.
+            if not siteurl.endswith('/'):
+                siteurl += '/'
+
         # XXX Put this in a different location.
         if what in {'filename', 'attach'}:
             if path.startswith('/'):
@@ -260,7 +279,7 @@ class Content(object):
                             "%s used {attach} link syntax on a "
                             "non-static file. Use {filename} instead.",
                             self.get_relative_source_path())
-                origin = '/'.join((siteurl, linked_content.url))
+                origin = joiner(siteurl, linked_content.url)
                 origin = origin.replace('\\', '/')  # for Windows paths.
             else:
                 logger.warning(
@@ -269,13 +288,13 @@ class Content(object):
                         'limit_msg': ("Other resources were not found "
                                       "and their urls not replaced")})
         elif what == 'category':
-            origin = '/'.join((siteurl, Category(path, self.settings).url))
+            origin = joiner(siteurl, Category(path, self.settings).url)
         elif what == 'tag':
-            origin = '/'.join((siteurl, Tag(path, self.settings).url))
+            origin = joiner(siteurl, Tag(path, self.settings).url)
         elif what == 'index':
-            origin = '/'.join((siteurl, self.settings['INDEX_SAVE_AS']))
+            origin = joiner(siteurl, self.settings['INDEX_SAVE_AS'])
         elif what == 'author':
-            origin = '/'.join((siteurl, Author(path, self.settings).url))
+            origin = joiner(siteurl, Author(path, self.settings).url)
         else:
             logger.warning(
                 "Replacement Indicator '%s' not recognized, "