Charset: Latin1: when encoding, substitute '?' instead of truncating

String::data(Latin1) keeps only the low byte of each code point, so the
fast path wrote garbage for characters ISO 8859-1 cannot hold,
instead of '?' as documented in tcharset.h:
     * When encoding into a legacy code page, characters which cannot be
     * represented by the code page are replaced by '?'.

taglib/toolkit/tcharset.cpp

@@ -73,9 +73,18 @@ ByteVector TagLib::Charset::encode(const String &s, Type type)
   if(type == Type::UTF8)
     return s.data(String::UTF8);
 
-  // Fast path for likely Latin-1 case
-  if(type == Type::Latin1)
-    return s.data(String::Latin1);
+  if(type == Type::Latin1) {
+    // String::data(Latin1) casts every code point to char, keeping only the
+    // low byte, so the '?' substitution the header promises is done here.
+    ByteVector data(s.size(), '?');
+    char *p = data.data();
+    for(const wchar_t c : s) {
+      if(c < 256)
+        *p = static_cast<char>(c);
+      ++p;
+    }
+    return data;
+  }
 
   const auto *entry = detail::entryForCodePage(static_cast<unsigned int>(type));
   if(!entry)

tests/test_id3v1.cpp

@@ -28,6 +28,7 @@
 
 #include "taglib_config.h"
 #include "tstring.h"
+#include "tcharset.h"
 #include "mpegfile.h"
 #include "id3v1tag.h"
 #include "id3v1genres.h"
@@ -64,6 +65,7 @@ class TestID3v1 : public CppUnit::TestFixture
   CPPUNIT_TEST(testStripWhiteSpace);
   CPPUNIT_TEST(testGenres);
   CPPUNIT_TEST(testRenamedGenres);
+  CPPUNIT_TEST(testRenderLatin1Substitutes);
 #ifdef TAGLIB_WITH_CHARSETTABLES
   CPPUNIT_TEST(testAssumedCodePageDefaultIsLatin1);
   CPPUNIT_TEST(testAssumedCodePage);
@@ -116,6 +118,22 @@ public:
     CPPUNIT_ASSERT_EQUAL(129U, tag.genreNumber());
   }
 
+  // Encoding into Latin1 has to substitute '?' for characters it cannot hold;
+  // truncating them to the low byte would write garbage into a file.
+  void testRenderLatin1Substitutes()
+  {
+    const ByteVector expected("A ? B", 5);
+
+    CPPUNIT_ASSERT_EQUAL(expected,
+      Charset::encode(kCyrillicArtistText, Charset::Type::Latin1));
+
+    // A fresh tag has no code page stated and is Latin1, so this is what
+    // render() writes for the artist field at offset 3 + 30.
+    ID3v1::Tag tag;
+    tag.setArtist(kCyrillicArtistText);
+    CPPUNIT_ASSERT_EQUAL(expected, tag.render().mid(33, 5));
+  }
+
 #ifdef TAGLIB_WITH_CHARSETTABLES
   //! A temporary MP3 file which is automatically deleted when it goes out of scope.
   class ScopedMp3FileCopyWithId3v1Tag : public ScopedFileCopy