Issue 5333 - Allow cms to generate relative pages

cms/converters.py

 # This file is part of the Adblock Plus web scripts,
 # Copyright (C) 2006-present eyeo GmbH
 #
 # Adblock Plus is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License version 3 as
 # published by the Free Software Foundation.
 #
 # Adblock Plus is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
@@ skipping 280 matching lines @@
             r'{{(?:(?!}}).)*}}'  # Nested translation
             r')*?)'
             r')?'
             r'}}',
             lookup_string,
             text,
             flags=re.S,
         )
 
     def process_links(self, text):
+        def make_relative(source, target):

[Vasily Kuznetsov, 2018/11/02 10:45:35]

It would probably be good to rename `source` to `base_url` to use the
terminology from the relevant RFC. The code would be easier to follow.

[rhowell, 2018/11/06 03:34:41]

Done.

+            if not target.startswith('/'):
+                # Links to an external resource
+                return target
+            if source == target:
+                target = re.split('/', target)

[Vasily Kuznetsov, 2018/11/02 10:45:35]

If this better than target.split('/')?

[rhowell, 2018/11/06 03:34:41]

Done.

+                return target[-1]
+            return os.path.relpath(target, source)

[Vasily Kuznetsov, 2018/11/02 10:45:35]

Here we need to take the part of `source` (or `base_url`) up to the last slash.
Something like `source.rsplit('/', 1)[0]`.

[rhowell, 2018/11/06 03:34:41]

Or, maybe it's cleaner to just pass the base path into this function in the
first place, and not include the current page when we first make the
current_page variable on line 325-326?

[Vasily Kuznetsov, 2018/11/06 12:45:16]

This is not quite right. First, "base URL" has specific meaning in the RFC and
it is the URL of the original page (unless the page overrides it, but we presume
it doesn't). Second, in your implementation (that means "containing folder" by
"base URL") the base URL is equal to "/" + locale, which is not the right
containing folder for pages that are in a folder (for example is the source is
"pages/foo/bar", it will be rendered to "/en/foo/bar" and the containing folder
would be "/en/foo").

[rhowell, 2018/11/07 16:16:58]

Ok, I think I get it now. I've tried a different way to get the current page
(using the pagedata info, and grabbing whatever comes after `/pages/`), and then
splitting it as you suggested when passing it to `relpath()`. Is this more what
you had in mind?

+
         def process_link(match):
             pre, attr, url, post = match.groups()
             url = jinja2.Markup(url).unescape()
 
             locale, new_url = (
                 self._params['source']
                 .resolve_link(url, self._params['locale']))
 
             if new_url is not None:
                 url = new_url
                 if attr == 'href':
                     post += ' hreflang="{}"'\
                         .format(jinja2.Markup.escape(locale))
 
+            if self._params['relative']:
+                current_page = '/{}/{}'.format(self._params['locale'],
+                                               self._params['page'])
+                url = make_relative(current_page, url)
+
             return ''.join((pre, jinja2.Markup.escape(url), post))
 
         text = re.sub(r'(<a\s[^<>]*\b(href)=\")([^<>\"]+)(\")',
                       process_link, text)
         text = re.sub(r'(<img\s[^<>]*\b(src)=\")([^<>\"]+)(\")',
                       process_link, text)
         return text
 
     include_start_regex = '<'
     include_end_regex = '>'
@@ skipping 100 matching lines @@
             'linkify': self.linkify,
             'toclist': self.toclist,
         }
 
         globals = {
             'get_string': self.get_string,
             'has_string': self.has_string,
             'get_page_content': self.get_page_content,
             'get_pages_metadata': self.get_pages_metadata,
             'get_canonical_url': self.get_canonical_url,
+            'relative': self._params['relative'],
         }
 
         for dirname, dictionary in [('filters', filters),
                                     ('globals', globals)]:
             for filename in self._params['source'].list_files(dirname):
                 root, ext = os.path.splitext(filename)
                 if ext.lower() != '.py':
                     continue
 
                 path = os.path.join(dirname, filename)
@@ skipping 132 matching lines @@
             stack[-1]['subitems'].append(item)
             stack.append(item)
         return structured
 
 
 converters = {
     'html': RawConverter,
     'md': MarkdownConverter,
     'tmpl': TemplateConverter,
 }