From 6d4cea57ede2786baabc16ffe5a0711f4058e657 Mon Sep 17 00:00:00 2001
From: Frank Tang <ftang@chromium.org>
Date: Thu, 11 Apr 2024 11:07:06 -0700
Subject: [PATCH] ICU-22727 fix addLikelySubtags for 5-8 chars lang codes

Fix C++ code, add tests for C, C++ and Java API
---
 icu4c/source/common/loclikely.cpp             |  7 +++--
 icu4c/source/test/cintltst/cloctst.c          | 31 +++++++++++++++++++
 icu4c/source/test/intltest/loctest.cpp        | 25 +++++++++++++++
 .../ibm/icu/dev/test/util/ULocaleTest.java    | 20 ++++++++++++
 4 files changed, 81 insertions(+), 2 deletions(-)

diff --git a/icu4c/source/common/loclikely.cpp b/icu4c/source/common/loclikely.cpp
index e8fd91ca79ee..5aa929dc4b15 100644
--- a/icu4c/source/common/loclikely.cpp
+++ b/icu4c/source/common/loclikely.cpp
@@ -157,14 +157,17 @@ _uloc_addLikelySubtags(const char* localeID,
         return;
     }
 
-    if (lang.length() > 3) {
-        if (lang.length() == 4 && script.isEmpty()) {
+    if (lang.length() == 4) {
+        if (script.isEmpty()) {
             script = std::move(lang);
             lang.clear();
         } else {
             err = U_ILLEGAL_ARGUMENT_ERROR;
             return;
         }
+    } else if (lang.length() > 8) {
+        err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
     }
 
     int32_t trailingLength = (int32_t)uprv_strlen(trailing);
diff --git a/icu4c/source/test/cintltst/cloctst.c b/icu4c/source/test/cintltst/cloctst.c
index 99187b70877a..77278619fee6 100644
--- a/icu4c/source/test/cintltst/cloctst.c
+++ b/icu4c/source/test/cintltst/cloctst.c
@@ -3867,6 +3867,27 @@ const char* const basic_maximize_data[][2] = {
     // so "aaaa" is a well-formed unicode_language_id
     "aaaa",
     "aaaa",
+  }, {
+    // ICU-22727
+    // unicode_language_subtag = alpha{2,3} | alpha{5,8};
+    // so "bbbbb", "cccccc", "ddddddd", "eeeeeeee" are
+    // well-formed unicode_language_subtag and therefore
+    // well-formed unicode_language_id
+    // but "fffffffff" is not.
+    "bbbbb",
+    "bbbbb",
+  }, {
+    // ICU-22727
+    "cccccc",
+    "cccccc",
+  }, {
+    // ICU-22727
+    "ddddddd",
+    "ddddddd",
+  }, {
+    // ICU-22727
+    "eeeeeeee",
+    "eeeeeeee",
   }, {
     // ICU-22546
     "und-Zzzz",
@@ -6048,6 +6069,16 @@ const errorData maximizeErrors[] = {
         "en_Latn_US_POSIX@currency=EURO",
         U_STRING_NOT_TERMINATED_WARNING,
         30
+    },
+    {
+        // ICU-22727
+        // unicode_language_subtag = alpha{2,3} | alpha{5,8};
+        // so "bbbbb", "cccccc", "ddddddd", "eeeeeeee" are
+        // well-formed unicode_language_id but "fffffffff" is not.
+        "fffffffff",
+        NULL,
+        U_ILLEGAL_ARGUMENT_ERROR,
+        0
     }
 };
 
diff --git a/icu4c/source/test/intltest/loctest.cpp b/icu4c/source/test/intltest/loctest.cpp
index 50f3d7075a51..8ba259c32bf8 100644
--- a/icu4c/source/test/intltest/loctest.cpp
+++ b/icu4c/source/test/intltest/loctest.cpp
@@ -3915,6 +3915,31 @@ LocaleTest::TestAddLikelyAndMinimizeSubtags() {
             "aaaa",
             "aaaa",
             "aaaa",
+        }, {
+            // ICU-22727
+            // unicode_language_subtag = alpha{2,3} | alpha{5,8};
+            // so "bbbbb", "cccccc", "ddddddd", "eeeeeeee" are
+            // well-formed unicode_language_subtag and therefore
+            // well-formed unicode_language_id
+            // but "fffffffff" is not.
+            "bbbbb",
+            "bbbbb",
+            "bbbbb",
+        }, {
+            // ICU-22727
+            "cccccc",
+            "cccccc",
+            "cccccc",
+        }, {
+            // ICU-22727
+            "ddddddd",
+            "ddddddd",
+            "ddddddd",
+        }, {
+            // ICU-22727
+            "eeeeeeee",
+            "eeeeeeee",
+            "eeeeeeee",
         }, {
             // ICU-22546
             "und-Zzzz",
diff --git a/icu4j/main/common_tests/src/test/java/com/ibm/icu/dev/test/util/ULocaleTest.java b/icu4j/main/common_tests/src/test/java/com/ibm/icu/dev/test/util/ULocaleTest.java
index e4376270d9b1..c7a8c1c889e0 100644
--- a/icu4j/main/common_tests/src/test/java/com/ibm/icu/dev/test/util/ULocaleTest.java
+++ b/icu4j/main/common_tests/src/test/java/com/ibm/icu/dev/test/util/ULocaleTest.java
@@ -1925,6 +1925,26 @@ public void TestAddLikelySubtags() {
                     // so "aaaa" is a well-formed unicode_language_id
                     "aaaa",
                     "aaaa",
+                }, {
+                    // ICU-22727
+                    // unicode_language_subtag = alpha{2,3} | alpha{5,8};
+                    // so "bbbbb", "cccccc", "ddddddd", "eeeeeeee" are
+                    // well-formed unicode_language_subtag and therefore
+                    // well-formed unicode_language_id
+                    "bbbbb",
+                    "bbbbb",
+                }, {
+                    // ICU-22727
+                    "cccccc",
+                    "cccccc",
+                }, {
+                    // ICU-22727
+                    "ddddddd",
+                    "ddddddd",
+                }, {
+                    // ICU-22727
+                    "eeeeeeee",
+                    "eeeeeeee",
                 }, {
                     // ICU-22546
                     "und-Zzzz",