Issue 2139 - Allow nested translations for tag attributes.

cms/converters.py

 # coding: utf-8
 
 # This file is part of the Adblock Plus web scripts,
 # Copyright (C) 2006-2015 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,
@@ skipping 125 matching lines @@
       result = localedata[name].strip()
     else:
       result = default
 
     # Insert fixed strings
     for i in range(len(fixed_strings)):
       result = re.sub(r"\{%d\}" % (i + 1), fixed_strings[i], result, 1)
 
     # Insert attributes
     result = escape(result)
+    def stringify_attribute((name, value)):
+      return '%s="%s"' % (
+        escape(name),
+        escape(self.insert_localized_strings(value, escapes))
+      )
+
     for tag in self.whitelist:
       saved = saved_attributes.get(tag, [])
       for attrs in saved:
-        attrs = map(lambda (name, value): '%s="%s"' % (escape(name), escape(value)), attrs)
+        attrs = map(stringify_attribute, attrs)
         result = re.sub(
           r"%s([^<>]*?)%s" % (re_escape("<%s>" % tag), re_escape("</%s>" % tag)),
           r'<%s%s>\1</%s>' % (tag, " " + " ".join(attrs) if attrs else "", tag),
           result, 1, flags=re.S
         )
       result = re.sub(
         r"%s([^<>]*?)%s" % (re_escape("<%s>" % tag), re_escape("</%s>" % tag)),
         r"<%s>\1</%s>" % (tag, tag),
         result, flags=re.S
       )
     return result
 
   def insert_localized_strings(self, text, escapes, to_html=lambda s: s):
     def lookup_string(match):
       name, comment, default = match.groups()
       default = to_html(default).strip()
 
       # Note: We currently ignore the comment, it is only relevant when
       # generating the master translation.
       return self.localize_string(name, default, self._params["localedata"], escapes)
 
     return re.sub(
-      r"\{\{\s*([\w\-]+)(?:\[(.*?)\])?\s+(.*?)\}\}",
+      r"\{\{\s*"
+      r"([\w\-]+)" # String ID
+      r"(?:\[(.*?)\])?" # Optional comment
+      r"\s+"
+      r"((?:[^\{]|\{(?!\{)|\{\{(?:[^\}]|\}(?!\}))*?\}\})*?)" # Translatable text

[Wladimir Palant, 2015/04/22 13:52:24]

Yes, this line isn't very readable now. How about splitting it up:

r"((?:(?!{{).|"  # Translatable text
  r"{{(?:(?!}}).)*}}" # Nested translation
r")*?)"

Note that the quantifier for the nested translation doesn't need to be greedy
any more, so I removed the question mark there. Also, I realized that { isn't a
special character in regular expressions so we probably shouldn't escape it.
Finally, it seems that (?:[^{]|{(?!{)| can be replaced by (?:(?!{{).| which is
slightly simpler.

[kzar, 2015/04/22 14:46:06]

Nice, that looks much better and still seems to work fine. Done.

+      r"\}\}",
       lookup_string,
       text,
       flags=re.S
     )
 
   def process_links(self, text):
     def process_link(match):
       pre, attr, url, post = match.groups()
       url = jinja2.Markup(url).unescape()
 
@@ skipping 199 matching lines @@
         stack.pop()
       stack[-1]["subitems"].append(item)
       stack.append(item)
     return structured
 
 converters = {
   "html": RawConverter,
   "md": MarkdownConverter,
   "tmpl": TemplateConverter,
 }