fix: Allow non-ASCII values in all attributes

PR #1017 allowed Unicord characters everywhere.
This fixes a bug that xml2rfc was guarding against non-ASCII characters
in attribute values.
With `--warn-bare-unicode`, xml2rfc will warn if non-ASCII characters are
present in attribute values.

Fixes #1105

xml2rfc/util/unicode.py

@@ -51,15 +51,6 @@
  ('xref', 'derivedContent'),
 }
 
-latinscript_attributes = {
- ('author', 'asciiFullname'),
- ('author', 'asciiSurname'),
- ('author', 'asciiInitials'),
- ('contact', 'asciiFullname'),
- ('contact', 'asciiSurname'),
- ('contact', 'asciiInitials'),
-}
-
 # These unicode tags don't require an ascii attribute
 bare_unicode_tags = set([
  'artwork',

xml2rfc/writers/preptool.py

@@ -42,7 +42,7 @@
 from xml2rfc.util.num import ol_style_formatter
 from xml2rfc.util.unicode import (
  unicode_content_tags, unicode_attributes, expand_unicode_element,
- isascii, latinscript_attributes, is_svg)
+ isascii, is_svg)
 from xml2rfc.utils import build_dataurl, namespaces, sdict, clean_text
 from xml2rfc.writers.base import default_options, BaseV3Writer, RfcWriterError
 
@@ -436,12 +436,9 @@ def check_attribute_values(self, e, p):
  for a in c.attrib:
  v = c.get(a)
  if not (c.tag, a) in unicode_attributes:
- if (c.tag, a) in latinscript_attributes:
- if not is_script(v, 'Latin'):
- self.err(c, 'Found non-Latin-script content in <%s> attribute value %s="%s"' % (c.tag, a, v))
- else:
+ if self.options.warn_bare_unicode:
  if not isascii(v):
- self.err(c, 'Found non-ASCII content in <%s> attribute value %s="%s"' % (c.tag, a, v))
+ self.warn(c, f'Found non-ASCII content in {c.tag} attribute value {a}="{v}" that should be inspected to ensure they are intentional.')
  if not (c.tag, a) in space_attributes:
  vv = v.strip()
  if vv != v: