(?!.) ftw

unicode-org · Dec 12, 2024 · 938ef97 · 938ef97
1 parent d428c8d
commit 938ef97
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 20 deletions.
diff --git a/icu4c/source/test/intltest/rbbitst.cpp b/icu4c/source/test/intltest/rbbitst.cpp
@@ -2906,20 +2906,16 @@ RBBILineMonkey::RBBILineMonkey() :
 
     std::list<std::pair<std::string, UnicodeSet>> partition;
 
-    // TODO(egg): The following two workarounds for what seems to be ICU bugs;
-    // with UREGEX_DOTALL (but not UREGEX_MULTILINE):
-    // 1. /.*\u000A/ does not match CR LF;
-    // 2. /$/ matches ( BK | CR | LF | NL ) eot.
+    // TODO(egg): The following is a workaround for what seems to be an ICU bug:
+    // with UREGEX_DOTALL (but not UREGEX_MULTILINE), /.*\u000A/ does not match
+    // CR LF.
     rules.push_back(std::make_unique<RegexRule>(uR"(CR LF ÷)", uR"(\u000D\u000A)", u'÷', uR"()"));
-    rules.push_back(std::make_unique<RegexRule>(
-        uR"([^ BK CR LF NL ] × [ BK CR LF NL ] eot)",
-        uR"([^ \p{lb=BK} \p{lb=CR} \p{lb=LF} \p{lb=NL} ])",
-        u'×',
-        uR"([ \p{lb=BK} \p{lb=CR} \p{lb=LF} \p{lb=NL} ] $)"));
 
     rules.push_back(std::make_unique<RegexRule>(uR"(sot ÷ contra LB2)", uR"(^)", u'÷', uR"()"));
     // This one could be part of the rules.
-    rules.push_back(std::make_unique<RegexRule>(uR"(LB3 ÷ eot)", uR"()", u'÷', uR"($)"));
+    // Note that /$/ matches ( BK | CR | LF | NL ) eot, so we use (?!.) instead.
+    // The generated rules use the same (?!.).
+    rules.push_back(std::make_unique<RegexRule>(uR"(LB3 ÷ eot)", uR"()", u'÷', uR"((?!.))"));
 
     // --- NOLI ME TANGERE ---
     // Generated by GenerateBreakTest.java in the Unicode tools.

diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/rbbi/RBBITestMonkey.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/rbbi/RBBITestMonkey.java
@@ -701,18 +701,11 @@ class NamedSet {
             List<NamedSet> partition = new ArrayList<>();
             rules = new ArrayList<>();
 
-            // /$/ matches ( BK | CR | LF | NL ) eot, so in this case we need to apply LB6 before
-            // LB3 gets incorrectly applied.
-            rules.add(new RegexRule(
-                "[^ BK CR LF NL ] × [ BK CR LF NL ] eot",
-                "[^ \\p{lb=BK} \\p{lb=CR} \\p{lb=LF} \\p{lb=NL} ]",
-                Resolution.NO_BREAK,
-                "[ \\p{lb=BK} \\p{lb=CR} \\p{lb=LF} \\p{lb=NL} ] $"));
-
             rules.add(new RegexRule("sot ÷ contra LB2", "^", Resolution.BREAK, ""));
             // This one could be part of the rules.
-            rules.add(new RegexRule("LB3 ÷ eot", "", Resolution.BREAK, "$"));
-
+            // Note that /$/ matches ( BK | CR | LF | NL ) eot, so we use (?!.) instead.
+            // The generated rules use the same (?!.).
+            rules.add(new RegexRule("LB3 ÷ eot", "", Resolution.BREAK, "(?!.)"));
 
             // --- NOLI ME TANGERE ---
             // Generated by GenerateBreakTest.java in the Unicode tools.