[libcxx-commits] [PATCH] D149672: [libc++][format] Fixes UTF-8 continuation.

[Mark de Wever via Phabricator via libcxx-commits]

Mordante created this revision.
Mordante added reviewers: ldionne, tahonermann.
Herald added a project: All.
Mordante requested review of this revision.
Herald added a project: libc++.
Herald added a subscriber: libcxx-commits.
Herald added a reviewer: libc++.

The mask used to check whether a code unit is a valid continuation was
incorrect and accepts non-continuation code points. This fixes the
issue.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D149672

Files:
  libcxx/include/__format/unicode.h
  libcxx/test/std/utilities/format/format.functions/escaped_output.unicode.pass.cpp
  libcxx/test/std/utilities/format/format.functions/unicode.pass.cpp


Index: libcxx/test/std/utilities/format/format.functions/unicode.pass.cpp
===================================================================
--- libcxx/test/std/utilities/format/format.functions/unicode.pass.cpp
+++ libcxx/test/std/utilities/format/format.functions/unicode.pass.cpp
@@ -273,6 +273,31 @@
     check(SV("*ZZZZ\xefZZZZ*"), SV("{:*^11}"), SV("ZZZZ\xefZZZZ"));
     check(SV("*ZZZZ\xffZZZZ*"), SV("{:*^11}"), SV("ZZZZ\xffZZZZ"));
 
+    // Invalid continuations
+    check(SV("\xc2\x00"), SV("{}"), SV("\xc2\x00")); // 0b0000'0000
+    check(SV("\xc2\x40"), SV("{}"), SV("\xc2\x40")); // 0b0100'0000
+    check(SV("\xc2\xc0"), SV("{}"), SV("\xc2\xc0")); // 0b1100'0000
+
+    check(SV("\xe0\x00\x80"), SV("{}"), SV("\xe0\x00\x80"));
+    check(SV("\xe0\x40\x80"), SV("{}"), SV("\xe0\x40\x80"));
+    check(SV("\xe0\xc0\x80"), SV("{}"), SV("\xe0\xc0\x80"));
+
+    check(SV("\xe0\x80\x00"), SV("{}"), SV("\xe0\x80\x00"));
+    check(SV("\xe0\x80\x40"), SV("{}"), SV("\xe0\x80\x40"));
+    check(SV("\xe0\x80\xc0"), SV("{}"), SV("\xe0\x80\xc0"));
+
+    check(SV("\xf0\x80\x80\x00"), SV("{}"), SV("\xf0\x80\x80\x00"));
+    check(SV("\xf0\x80\x80\x40"), SV("{}"), SV("\xf0\x80\x80\x40"));
+    check(SV("\xf0\x80\x80\xc0"), SV("{}"), SV("\xf0\x80\x80\xc0"));
+
+    check(SV("\xf0\x80\x00\x80"), SV("{}"), SV("\xf0\x80\x00\x80"));
+    check(SV("\xf0\x80\x40\x80"), SV("{}"), SV("\xf0\x80\x40\x80"));
+    check(SV("\xf0\x80\xc0\x80"), SV("{}"), SV("\xf0\x80\xc0\x80"));
+
+    check(SV("\xf0\x00\x80\x80"), SV("{}"), SV("\xf0\x00\x80\x80"));
+    check(SV("\xf0\x40\x80\x80"), SV("{}"), SV("\xf0\x40\x80\x80"));
+    check(SV("\xf0\xc0\x80\x80"), SV("{}"), SV("\xf0\xc0\x80\x80"));
+
     // Premature end.
     check(SV("*ZZZZ\xef\xf5*"), SV("{:*^8}"), SV("ZZZZ\xef\xf5"));
     check(SV("*ZZZZ\xef\xf5ZZZZ*"), SV("{:*^12}"), SV("ZZZZ\xef\xf5ZZZZ"));
Index: libcxx/test/std/utilities/format/format.functions/escaped_output.unicode.pass.cpp
===================================================================
--- libcxx/test/std/utilities/format/format.functions/escaped_output.unicode.pass.cpp
+++ libcxx/test/std/utilities/format/format.functions/escaped_output.unicode.pass.cpp
@@ -505,6 +505,13 @@
               "\xf7\xbf\xbf"
               "a");
 
+  // http://unicode.org/review/pr-121.html
+  test_format(R"("a\x{f1}\x{80}\x{80}\x{e1}\x{80}\x{c2}b")"sv,
+              "{:?}",
+              "a"
+              "\xf1\x80\x80\xe1\x80\xc2"
+              "b");
+
   // Code unit out of range
   test_format(R"("\u{10ffff}")"sv, "{:?}", "\xf4\x8f\xbf\xbf");               // last valid code point
   test_format(R"("\x{f4}\x{90}\x{80}\x{80}")"sv, "{:?}", "\xf4\x90\x80\x80"); // first invalid code point
Index: libcxx/include/__format/unicode.h
===================================================================
--- libcxx/include/__format/unicode.h
+++ libcxx/include/__format/unicode.h
@@ -105,7 +105,7 @@
   requires same_as<iter_value_t<_Iterator>, char>
 _LIBCPP_HIDE_FROM_ABI constexpr bool __is_continuation(_Iterator __char, int __count) {
   do {
-    if ((*__char & 0b1000'0000) != 0b1000'0000)
+    if ((*__char & 0b1100'0000) != 0b1000'0000)
       return false;
     --__count;
     ++__char;


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D149672.518800.patch
Type: text/x-patch
Size: 3225 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/libcxx-commits/attachments/20230502/9c58824f/attachment.bin>