Using NBSP with colons in the punctuation spacing fixup

[HarmfulBreeze]

Describe the problem

Opening a feature request as requested in #9470

Some guides like the Lexique des règles typographiques en usage à l'Imprimerie nationale, often used as a reference in France, specify that there should be a regular non-breaking space before the colon character, but not for other "two-part" characters, such as semicolons, where a NNBSP should be used instead. Some organizations and media however prefer to use NNBSP with colons.

Are there plans to add a toggle (of sorts) to allow the use of NBSP with colons?

Thanks!

Describe the solution you would like

A toggle allowing to switch between NBSP and NNBSP with colons would be useful.

[nijel]

There could be an additional flag for this which would configure desired spacing. Something like punctuation-spacing:nbsp could work.

[github-actions]

This issue seems to be a good fit for newbie contributors. You are welcome to contribute to Weblate! Don't hesitate to ask any questions you would have while implementing this.

You can learn about how to get started in our contributors documentation.

[ashsgelb]

Can I claim this?

[Brian3015]

I'm thinking of working on this. How would you go about testing if you added this feature? Also, where would you add it?

[nijel]

Anybody can start working on an issue. The code belongs to the existing check and fixup implementation:

weblate/weblate/trans/autofixes/chars.py

Lines 80 to 101 in ec2105f

	class PunctuationSpacing(AutoFix):
	"""Ensures French and Breton use correct punctuation spacing."""
	fix_id = "punctuation-spacing"
	name = gettext_lazy("Punctuation spacing")
	@staticmethod
	def get_related_checks():
	return [PunctuationSpacingCheck()]
	def fix_single_target(self, target, source, unit):
	if (
	unit.translation.language.is_base(("fr", "br"))
	and unit.translation.language.code != "fr_CA"
	and "ignore-punctuation-spacing" not in unit.all_flags
	):
	# Fix existing
	new_target = re.sub(FRENCH_PUNCTUATION_FIXUP_RE, "\u202f\\2", target)
	# Do not add missing as that is likely to trigger issues with other content
	# such as URLs or Markdown syntax.
	return new_target, new_target != target
	return target, False

weblate/weblate/checks/chars.py

Lines 418 to 469 in ec2105f

	class PunctuationSpacingCheck(TargetCheck):
	check_id = "punctuation_spacing"
	name = gettext_lazy("Punctuation spacing")
	description = gettext_lazy(
	"Missing non breakable space before double punctuation sign"
	)
	def check_single(self, source, target, unit) -> bool:
	if (
	not unit.translation.language.is_base(("fr", "br"))
	or unit.translation.language.code == "fr_CA"
	):
	return False
	# Remove XML/HTML entities to simplify parsing
	target = strip_entities(target)
	whitespace = {" ", "\u00a0", "\u202f", "\u2009"}
	total = len(target)
	for i, char in enumerate(target):
	if char in FRENCH_PUNCTUATION:
	if i == 0:
	# Trigger if punctionation at beginning of the string
	return True
	if (
	i + 1 < total
	and unicodedata.category(target[i + 1])
	not in FRENCH_PUNCTUATION_SPACING
	):
	# Ignore when not followed by space or open/close bracket
	continue
	prev_char = target[i - 1]
	if prev_char not in whitespace and prev_char not in FRENCH_PUNCTUATION:
	return True
	return False
	def get_fixup(self, unit):
	return [
	# First fix possibly wrong whitespace
	(
	FRENCH_PUNCTUATION_FIXUP_RE,
	"\u202f$2",
	"gu",
	),
	# Then add missing ones
	(
	FRENCH_PUNCTUATION_MISSING_RE,
	"$1\u202f$2",
	"gu",
	),
	]