fix(core): break a label before a left parenthesis, not after it (#4553)

src/mln/text/shaping.cpp

@@ -385,7 +385,10 @@ std::set<std::size_t> determineLineBreaks(const TaggedString& logicalInput,
         // often appear without surrounding spaces.
         if (i < logicalInput.length() - 1) {
             const bool allowsIdeographicBreak = util::i18n::allowsIdeographicBreaking(codePoint);
-            if (section.imageID || allowsIdeographicBreak || util::i18n::allowsWordBreaking(codePoint)) {
+            const bool allowsBreakBeforeNext = i != logicalInput.length() - 2 &&
+                                               util::i18n::allowsWordBreakingBefore(logicalInput.getCharCodeAt(i + 1));
+            if (section.imageID || allowsIdeographicBreak || util::i18n::allowsWordBreaking(codePoint) ||
+                allowsBreakBeforeNext) {
                 const bool penalizableIdeographicBreak = allowsIdeographicBreak && hasServerSuggestedBreaks;
                 const std::size_t nextIndex = i + 1;
                 potentialBreaks.push_back(evaluateBreak(

src/mln/util/i18n.cpp

@@ -352,7 +352,6 @@ bool allowsWordBreaking(char16_t chr) {
     return (chr == 0x0a      /* newline */
             || chr == 0x20   /* space */
             || chr == 0x26   /* ampersand */
-            || chr == 0x28   /* open parenthesis */
             || chr == 0x29   /* close parenthesis */
             || chr == 0x2b   /* plus sign */
             || chr == 0x2d   /* hyphen-minus */
@@ -365,6 +364,10 @@ bool allowsWordBreaking(char16_t chr) {
     );
 }
 
+bool allowsWordBreakingBefore(char16_t chr) {
+    return chr == 0x28; /* open parenthesis */
+}
+
 bool charAllowsLetterSpacing(char16_t chr) {
     return !(isInArabic(chr) || isInArabicSupplement(chr) || isInArabicExtendedA(chr) ||
              isInArabicPresentationFormsA(chr) || isInArabicPresentationFormsB(chr));

src/mln/util/i18n.hpp

@@ -10,6 +10,10 @@ namespace i18n {
     by the given Unicode codepoint due to word breaking. */
 bool allowsWordBreaking(char16_t chr);
 
+/** Returns whether a line break can be inserted before the character indicated
+    by the given Unicode codepoint due to word breaking. */
+bool allowsWordBreakingBefore(char16_t chr);
+
 /** Returns whether the given string can be displayed with letter-spacing.
     False for Arabic scripts, where letter-spacing will break ligatures. */
 bool allowsLetterSpacing(const std::u16string& string);