From 2651372291df9106ff2c63c1e356969239fe31b4 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Wed, 9 Nov 2016 15:51:33 +0900
Subject: [PATCH 01/24] Separate some tests to make them more specific.

---
 tests/test_list.cpp | 28 ++++++++++++++++++++--------
 tests/test_map.cpp  | 18 ++++++++++++++----
 2 files changed, 34 insertions(+), 12 deletions(-)

diff --git a/tests/test_list.cpp b/tests/test_list.cpp
index e314836a..1c6d8c4c 100644
--- a/tests/test_list.cpp
+++ b/tests/test_list.cpp
@@ -32,12 +32,13 @@ using namespace TagLib;
 class TestList : public CppUnit::TestFixture
 {
   CPPUNIT_TEST_SUITE(TestList);
-  CPPUNIT_TEST(testList);
+  CPPUNIT_TEST(testAppend);
+  CPPUNIT_TEST(testDetach);
   CPPUNIT_TEST_SUITE_END();
 
 public:
 
-  void testList()
+  void testAppend()
   {
     List<int> l1;
     List<int> l2;
@@ -51,14 +52,25 @@ public:
     l3.append(2);
     l3.append(3);
     l3.append(4);
+    CPPUNIT_ASSERT_EQUAL(4U, l1.size());
     CPPUNIT_ASSERT(l1 == l3);
-
-    List<int> l4 = l1;
-    List<int>::Iterator it = l4.find(3);
-    *it = 33;
-    CPPUNIT_ASSERT_EQUAL(l1[2], 3);
-    CPPUNIT_ASSERT_EQUAL(l4[2], 33);
   }
+
+  void testDetach()
+  {
+    List<int> l1;
+    l1.append(1);
+    l1.append(2);
+    l1.append(3);
+    l1.append(4);
+
+    List<int> l2 = l1;
+    List<int>::Iterator it = l2.find(3);
+    *it = 33;
+    CPPUNIT_ASSERT_EQUAL(3,  l1[2]);
+    CPPUNIT_ASSERT_EQUAL(33, l2[2]);
+  }
+
 };
 
 CPPUNIT_TEST_SUITE_REGISTRATION(TestList);
diff --git a/tests/test_map.cpp b/tests/test_map.cpp
index 3096ff69..b5e493b6 100644
--- a/tests/test_map.cpp
+++ b/tests/test_map.cpp
@@ -34,6 +34,7 @@ class TestMap : public CppUnit::TestFixture
 {
   CPPUNIT_TEST_SUITE(TestMap);
   CPPUNIT_TEST(testInsert);
+  CPPUNIT_TEST(testDetach);
   CPPUNIT_TEST_SUITE_END();
 
 public:
@@ -42,19 +43,28 @@ public:
   {
     Map<String, int> m1;
     m1.insert("foo", 3);
+    m1.insert("bar", 5);
+    CPPUNIT_ASSERT_EQUAL(2U, m1.size());
     CPPUNIT_ASSERT_EQUAL(3, m1["foo"]);
+    CPPUNIT_ASSERT_EQUAL(5, m1["bar"]);
     m1.insert("foo", 7);
+    CPPUNIT_ASSERT_EQUAL(2U, m1.size());
     CPPUNIT_ASSERT_EQUAL(7, m1["foo"]);
+    CPPUNIT_ASSERT_EQUAL(5, m1["bar"]);
+  }
 
-    m1.insert("alice",  5);
-    m1.insert("bob",    9);
+  void testDetach()
+  {
+    Map<String, int> m1;
+    m1.insert("alice", 5);
+    m1.insert("bob", 9);
     m1.insert("carol", 11);
 
     Map<String, int> m2 = m1;
     Map<String, int>::Iterator it = m2.find("bob");
     (*it).second = 99;
-    CPPUNIT_ASSERT_EQUAL(m1["bob"], 9);
-    CPPUNIT_ASSERT_EQUAL(m2["bob"], 99);
+    CPPUNIT_ASSERT_EQUAL(9,  m1["bob"]);
+    CPPUNIT_ASSERT_EQUAL(99, m2["bob"]);
   }
 
 };

From 7b8d576bdee3b7304806b3008bfde30baa3d03a6 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Thu, 10 Nov 2016 17:09:40 +0900
Subject: [PATCH 02/24] Don't decode redundant UTF-8 sequences in Win32.

Linux and OS X are working well and won't be affected.
---
 taglib/toolkit/tstring.cpp | 2 +-
 tests/test_string.cpp      | 9 +++++++++
 2 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/taglib/toolkit/tstring.cpp b/taglib/toolkit/tstring.cpp
index a1891a45..44788249 100644
--- a/taglib/toolkit/tstring.cpp
+++ b/taglib/toolkit/tstring.cpp
@@ -88,7 +88,7 @@ namespace
 #ifdef _WIN32
 
     len = ::MultiByteToWideChar(
-      CP_UTF8, 0, src, static_cast<int>(srcLength), dst, static_cast<int>(dstLength));
+      CP_UTF8, MB_ERR_INVALID_CHARS, src, static_cast<int>(srcLength), dst, static_cast<int>(dstLength));
 
 #else
 
diff --git a/tests/test_string.cpp b/tests/test_string.cpp
index f9f1a2dd..599e002d 100644
--- a/tests/test_string.cpp
+++ b/tests/test_string.cpp
@@ -50,6 +50,7 @@ class TestString : public CppUnit::TestFixture
   CPPUNIT_TEST(testEncodeNonLatin1);
   CPPUNIT_TEST(testEncodeEmpty);
   CPPUNIT_TEST(testIterator);
+  CPPUNIT_TEST(testRedundantUTF8);
   CPPUNIT_TEST_SUITE_END();
 
 public:
@@ -321,6 +322,14 @@ public:
     CPPUNIT_ASSERT_EQUAL(L'I', *it2);
   }
 
+  void testRedundantUTF8()
+  {
+    CPPUNIT_ASSERT_EQUAL(String("/"), String(ByteVector("\x2F"), String::UTF8));
+    CPPUNIT_ASSERT(String(ByteVector("\xC0\xAF"), String::UTF8).isEmpty());
+    CPPUNIT_ASSERT(String(ByteVector("\xE0\x80\xAF"), String::UTF8).isEmpty());
+    CPPUNIT_ASSERT(String(ByteVector("\xF0\x80\x80\xAF"), String::UTF8).isEmpty());
+  }
+
 };
 
 CPPUNIT_TEST_SUITE_REGISTRATION(TestString);

From f9a747dceb694e7adcf65705f2b784ba806e7301 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Thu, 10 Nov 2016 20:02:30 +0900
Subject: [PATCH 03/24] Avoid adding fields with invalid keys to Vorbis
 Comments.

According to the spec, '\x7F' is not allowed.
---
 taglib/ogg/xiphcomment.cpp | 21 ++++++++++++++++-----
 tests/test_xiphcomment.cpp | 20 +++++++++++++++++---
 2 files changed, 33 insertions(+), 8 deletions(-)

diff --git a/taglib/ogg/xiphcomment.cpp b/taglib/ogg/xiphcomment.cpp
index db46aecb..d6b8e11c 100644
--- a/taglib/ogg/xiphcomment.cpp
+++ b/taglib/ogg/xiphcomment.cpp
@@ -261,10 +261,14 @@ bool Ogg::XiphComment::checkKey(const String &key)
 {
   if(key.size() < 1)
     return false;
-  for(String::ConstIterator it = key.begin(); it != key.end(); it++)
-      // forbid non-printable, non-ascii, '=' (#61) and '~' (#126)
-      if (*it < 32 || *it >= 128 || *it == 61 || *it == 126)
+
+  // A key may consist of ASCII 0x20 through 0x7D, 0x3D ('=') excluded.
+
+  for(String::ConstIterator it = key.begin(); it != key.end(); it++) {
+      if(*it < 0x20 || *it > 0x7D || *it == 0x3D)
         return false;
+  }
+
   return true;
 }
 
@@ -275,11 +279,18 @@ String Ogg::XiphComment::vendorID() const
 
 void Ogg::XiphComment::addField(const String &key, const String &value, bool replace)
 {
+  if(!checkKey(key)) {
+    debug("Ogg::XiphComment::addField() - Invalid key. Field not added.");
+    return;
+  }
+
+  const String upperKey = key.upper();
+
   if(replace)
-    removeFields(key.upper());
+    removeFields(upperKey);
 
   if(!key.isEmpty() && !value.isEmpty())
-    d->fieldListMap[key.upper()].append(value);
+    d->fieldListMap[upperKey].append(value);
 }
 
 void Ogg::XiphComment::removeField(const String &key, const String &value)
diff --git a/tests/test_xiphcomment.cpp b/tests/test_xiphcomment.cpp
index 699f60a5..d2a28a1f 100644
--- a/tests/test_xiphcomment.cpp
+++ b/tests/test_xiphcomment.cpp
@@ -42,7 +42,8 @@ class TestXiphComment : public CppUnit::TestFixture
   CPPUNIT_TEST(testSetYear);
   CPPUNIT_TEST(testTrack);
   CPPUNIT_TEST(testSetTrack);
-  CPPUNIT_TEST(testInvalidKeys);
+  CPPUNIT_TEST(testInvalidKeys1);
+  CPPUNIT_TEST(testInvalidKeys2);
   CPPUNIT_TEST(testClearComment);
   CPPUNIT_TEST(testRemoveFields);
   CPPUNIT_TEST(testPicture);
@@ -90,19 +91,32 @@ public:
     CPPUNIT_ASSERT_EQUAL(String("3"), cmt.fieldListMap()["TRACKNUMBER"].front());
   }
 
-  void testInvalidKeys()
+  void testInvalidKeys1()
   {
     PropertyMap map;
     map[""] = String("invalid key: empty string");
     map["A=B"] = String("invalid key: contains '='");
     map["A~B"] = String("invalid key: contains '~'");
+    map["A\x7F\B"] = String("invalid key: contains '\x7F'");
+    map[L"A\x3456\B"] = String("invalid key: Unicode");
 
     Ogg::XiphComment cmt;
     PropertyMap unsuccessful = cmt.setProperties(map);
-    CPPUNIT_ASSERT_EQUAL((unsigned int)3, unsuccessful.size());
+    CPPUNIT_ASSERT_EQUAL((unsigned int)5, unsuccessful.size());
     CPPUNIT_ASSERT(cmt.properties().isEmpty());
   }
 
+  void testInvalidKeys2()
+  {
+    Ogg::XiphComment cmt;
+    cmt.addField("", "invalid key: empty string");
+    cmt.addField("A=B", "invalid key: contains '='");
+    cmt.addField("A~B", "invalid key: contains '~'");
+    cmt.addField("A\x7F\B", "invalid key: contains '\x7F'");
+    cmt.addField(L"A\x3456\B", "invalid key: Unicode");
+    CPPUNIT_ASSERT_EQUAL(0U, cmt.fieldCount());
+  }
+
   void testClearComment()
   {
     ScopedFileCopy copy("empty", ".ogg");

From b98a984b66f69cacfcbf259ac3b2d77ea3c9415f Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Fri, 11 Nov 2016 00:07:32 +0900
Subject: [PATCH 04/24] Fix handling of lowercase 'metadata_block_picture'
 fields in Vorbis comments.

Also refactored some redundant code for parsing pictures.
---
 taglib/ogg/xiphcomment.cpp      | 122 ++++++++++++++------------------
 tests/data/lowercase-fields.ogg | Bin 0 -> 4477 bytes
 tests/test_xiphcomment.cpp      |  18 +++++
 3 files changed, 71 insertions(+), 69 deletions(-)
 create mode 100644 tests/data/lowercase-fields.ogg

diff --git a/taglib/ogg/xiphcomment.cpp b/taglib/ogg/xiphcomment.cpp
index d6b8e11c..f56bf810 100644
--- a/taglib/ogg/xiphcomment.cpp
+++ b/taglib/ogg/xiphcomment.cpp
@@ -447,85 +447,69 @@ void Ogg::XiphComment::parse(const ByteVector &data)
     const unsigned int commentLength = data.toUInt(pos, false);
     pos += 4;
 
-    ByteVector entry = data.mid(pos, commentLength);
-
+    const ByteVector entry = data.mid(pos, commentLength);
     pos += commentLength;
 
     // Don't go past data end
+
     if(pos > data.size())
       break;
 
-    // Handle Pictures separately
-    if(entry.startsWith("METADATA_BLOCK_PICTURE=")) {
-
-      // We need base64 encoded data including padding
-      if((entry.size() - 23) > 3 && ((entry.size() - 23) % 4) == 0) {
-
-        // Decode base64 picture data
-        ByteVector picturedata = ByteVector::fromBase64(entry.mid(23));
-        if(picturedata.size()) {
-
-          // Decode Flac Picture
-          FLAC::Picture * picture = new FLAC::Picture();
-          if(picture->parse(picturedata)) {
-
-            d->pictureList.append(picture);
-
-            // continue to next field
-            continue;
-          }
-          else {
-            delete picture;
-            debug("Failed to decode FlacPicture block");
-          }
-        }
-        else {
-          debug("Failed to decode base64 encoded data");
-        }
-      }
-      else {
-        debug("Invalid base64 encoded data");
-      }
-    }
-
-    // Handle old picture standard
-    if(entry.startsWith("COVERART=")) {
-
-      if((entry.size() - 9) > 3 && ((entry.size() - 9) % 4) == 0) {
-
-        // Decode base64 picture data
-        ByteVector picturedata = ByteVector::fromBase64(entry.mid(9));
-        if (picturedata.size()) {
-
-          // Assume it's some type of image file
-          FLAC::Picture * picture = new FLAC::Picture();
-          picture->setData(picturedata);
-          picture->setMimeType("image/");
-          picture->setType(FLAC::Picture::Other);
-          d->pictureList.append(picture);
-
-          // continue to next field
-          continue;
-        }
-        else {
-          debug("Failed to decode base64 encoded data");
-        }
-      }
-      else {
-        debug("Invalid base64 encoded data");
-      }
-    }
-
     // Check for field separator
-    int sep = entry.find('=');
+
+    const int sep = entry.find('=');
     if(sep < 1) {
-      debug("Discarding invalid comment field.");
+      debug("Ogg::XiphComment::parse() - Discarding a field. Separator not found.");
       continue;
     }
 
-    // Parse key and value
-    String key = String(entry.mid(0, sep), String::UTF8);
-    String value = String(entry.mid(sep + 1), String::UTF8);
-    addField(key, value, false);
+    // Parse the key
+
+    const String key = String(entry.mid(0, sep), String::UTF8).upper();
+    if(!checkKey(key)) {
+      debug("Ogg::XiphComment::parse() - Discarding a field. Invalid key.");
+      continue;
+    }
+
+    if(key == "METADATA_BLOCK_PICTURE" || key == "COVERART") {
+
+      // Handle Pictures separately
+
+      const ByteVector picturedata = ByteVector::fromBase64(entry.mid(sep + 1));
+      if(picturedata.isEmpty()) {
+        debug("Ogg::XiphComment::parse() - Discarding a field. Invalid base64 data");
+        continue;
+      }
+
+      if(key[0] == L'M') {
+
+        // Decode FLAC Picture
+
+        FLAC::Picture * picture = new FLAC::Picture();
+        if(picture->parse(picturedata)) {
+          d->pictureList.append(picture);
+        }
+        else {
+          delete picture;
+          debug("Ogg::XiphComment::parse() - Failed to decode FLAC Picture block");
+        }
+      }
+      else {
+
+        // Assume it's some type of image file
+
+        FLAC::Picture * picture = new FLAC::Picture();
+        picture->setData(picturedata);
+        picture->setMimeType("image/");
+        picture->setType(FLAC::Picture::Other);
+        d->pictureList.append(picture);
+      }
+    }
+    else {
+
+      // Parse the text
+
+      addField(key, String(entry.mid(sep + 1), String::UTF8), false);
+    }
   }
 }
diff --git a/tests/data/lowercase-fields.ogg b/tests/data/lowercase-fields.ogg
new file mode 100644
index 0000000000000000000000000000000000000000..0ddd49357a906c70d6f3e5f222f0fa5404e4e2cb
GIT binary patch
literal 4477
zcmeHLeNa<Z7Qcaj0TEvy*g#{O;EV9m-~<d^VsSTl5+KnOcu9~C*~&+R2)0OI7k2AR
zB*CVEGHe>tHl+?=l0qu3&|<ASJL_7dDi-;u=+<p(OBe8CcHH%kI{VM=c`x|U{ii$Q
z{?XalJNMn3d+xdC-1|HCop){uYHRf<3@w-A*Z%MvO|B@rwTtx>>lZIIRhca`BP{0#
z{ezHojK%z(U=`3a9~62fJ@Qrimjm3>|NMSg5gyXC76eMnjddFenrej&=Bi>wzfdF0
zkjXM-8)fov5Qu}PvZ>i@Y2Iv5>kUE`WzgsiwZPqDYObuQY_5Evs^O*T`WG6_)y*wU
zrp+oSB;BYQ8>_bGzF4ML?5)l#ev#&E*`C=@(?*8)MKs4qH+giUrW;i?-Dvd<WG>Z4
z)|Zx0br_gxHa9nzLiP<B!!|W5BD4cBxuuwelnCJ#z9iRgS!2hN$il>V0XbQJJCWSR
zZ5L#0E1wWsdPKfH(ovqef2zlb=n1c=T6uP-iX*a%H<!3<DVu?ED1c|Cc^G-WQ7h_T
zRK;0qpO<CD*#44rPheLoN-*Fp3GJ)N6JFa91qkDr`GvMf#gf2&TAJ!?KO;M_#?H=6
z?P~v4cJ4FVO48xAX)^C6*=tp;YeF_~Rep~xQgx1R&sD(+A{8fi_6gNsl3k+UJZ%@s
zYZ6coLVW|N&VkfuENL_vokIvy7V$pWD>(lf!TAxvO3td%2!vRO19JIJUw-$u`7XTF
zC1&cvt3m0+C{r_->G=p17R9Ld@>VeOY7pY|WjXyNPJf+q(b3JZA2<?*tO&8ejC_ke
zJI|eS$zZ&4e9yH<YPrXh$V>+eu^IEOp-6`;JTa2-11!VJb;{0I?}KdRYhwqZh817+
z&i9qZdbJ~pwY~XcZLYVSA1M}lEtBULdebL?8t5&J@q*l9@0q^R=n$1M(l+Il(7e5q
zH+J@N$AMZu+}7iLZ49W0VRN^4>j+T!AGJ|$Z>4z_S{2jWJbOdX8&7u!fB%u^?Wwas
z6^!5bvKM<G!pJco!+w?6f!8YkOmJov>x8-7s-%n?s_&L$QAxI5^5*nXozK0ABR;s7
z(x!v}4?{|hsDnr4-*2iMq67S<3&gv9?OMsu;6PW#g!KK3Otb0sk@sVti%CC&ZFy$+
z45i<bIagBuSH+^&_Cax3{qPv2CH!tGn!!&9*prcoH9bYN<|+sAZf{exWH1D$kai71
z1%DSzbJsqse-m8dii_f)rp;z#&g#|17Sk0^OX-LE%CB}*e7EllsP|ebOFX91@72)O
zA<_upu!lB6jPorQ6-&4qyT46vC~D;8G3QM)if^64Kb?^l+?5x^Z!Y4gXXVD}go>h{
z>?wNIG@DtWf40I{yl1;-Px<kdYj2pYeYfwcM{30iW{VI_Ho4Xju4KYBpVf_Ply&DM
zNEu1Lk<M>YyXUi9xW*+myK37!gKv3nymd$tWJC}~i_paxXmt*?ItN?3mmIFTw>*PG
z-e)cZZl64Q>xb$}=78f6P4K)XJdeWF6c7x$))*RH!N=*ix=A`QBM<FyI({<%y1h9R
zG?pEJP|W#byoqUnE>3)zN;B%yE>r5udgBij-K<;}55PvD9v14s^l541-l$HzYG(I)
zebq74ZU0EVq;~AX3+WyFv7aZs;x58|J%m>~RTtBIQ^fk=IRn|j^SxOuNuXjWa9#((
zysYbu0_CU_DuWC{J@Qpk*{hGqS1<Q*UAVL$Tz-j;<BwaEGjiqKY^ckMz|fs+?wuU&
za$kBWTs|G1T?EnNvCF^PItI3KxKr7W*aAb@)BlZqJ@#V%ui$?j0mwkG>+tLp@sdW9
zks}^8=EvbtF7~J4`JM=!E$~M<%Vko{_L*h6hgje+pqdprkp>2)<c{QrLoOKTGh=G2
z4+iRz9=t;!aY-(6V(K^7pW3K%B<FAgQ4eQ5N<n1iH^K*8)*tg=O<7J83t8v0JdhJ~
z&zG!$<jx~@BDAyAn+k#c@yhoN#vw--YD@)#pNvSagGK!ODDp&1Wm^L**bNA$(yKeB
z(pT-ZLf{WWMF>Tn9g*td)h?geo8rVZv)Khavv;xWkool8Y<k-RbzV$eB-gzeFb5?v
zoiPFOmFqO@w!oQhP{)BIY($13fwQXAyU=zh#Tgj#hP0@s<vP7l;|jDLf@I0;4FpG}
zy1Uu*C(2+nb}X}l!JA;HfCtjX%Pn8i8o*-k=1IortU+gNsSr8?W^XVUJi0QcfCoA&
zuC(l5e09Ja48C`A1E!<ZwN#Ww(kl$!x{#qXf~Jb?9u__2*2&SBLgN(}_m~jflTHBY
zu>#1}yZT=5l5`%Mwy#{R>=!)~W=%!t6hFYzUy!XA`LLvM+|DD{i~TcHQ0Sv5BEfFZ
zO2miV)NaCvQN)ltnv{qSQV3n$9}Z7PVcj_`60Sy8SuXw2em>d9tDaP(a4Bc{V(;tP
z^aV(RG(;kN%@=)cCgq{DG*3bJnY_nMX(3a3UQ3AWld2TIdj|5Oc!UfL+K10Au-B;=
zxG5Hj>8_^X2BbVV<E>P0V7(^ffZ+N2cpd~d($EJbR}vsE&>4*tqf!RTn<r&7Yye9L
zw38be%4yV0iN}Y*BY@eCD=nr!dG<?xcl7q>qX^w$uRwiaPZsB85lJhh@)|Myy;SmT
zJ&=-{Ffv+XcT*gAD;_!iaVQ^-eAmbvolwmwYoP(kdXecqY$$s4VfR0x(jYlwGPFmU
z2=otQ7P|MdtH(B-)=Zb|1dp<;aNJMv{d`LdN%##M@IbSntaF5Pq%Y#W&&nFkS|(Mg
zqJwi)(nFS5k|53&R~Hd>iqfVXFX`0wA9hjN;majm=|jG`vd-c0sxA%j!DnTi>1)<C
z=!V9w?tX8CUtKRr`+#5Fb(Zr&O+#AX^WCe)IqS-$X_w3DW#jxW$~*nXy`AYlz4tjK
zO`qF=&@XM_D5w3kjYsfz61pRwNE(epfSVA5Ljq?8PKA%^QoyN*B{}^D;4zfQ-9?`s
z=Ke!zfyYoi)XilI(-QB{Gw3T4;kPlPm^Rh#<1OEqhJ0?CqDa^+IEs-I`yj)md7OT5
zo|fo<X?{0snefMHqr~<&iXi+vEk^`{z%yv0`t5OWIBD!bftCR83bG(M=BrW?jcY0o
zM?LXWD5}&z&NJYqh%|M3vY+qcZ_Q1x6qY#$n_{R>hp0*60`767+ug4lG8TAb40%(~
zvW<vLJIE);`M#@7Tl+1s5EWI(YVAQNoE@<dow|;`wqVU+?7SzDRhYBOk@GRx*jgBc
zarG=^RXe!>8A^8eDt0K}U~O3X=4iNxE#j!8v3*vypttek_?KEwuX=upkhBOvFG8!(
z9a!KF#N|$3tgNYPXl!nM*=l!m^+14xl89c0o2yuvlOwnA6O*14tbOWfp(uGB4q-g>
sCxFn3*z8b7MMrQpo4Mt)PyH$s&*@t}P3`*azVd$7kG-Ji*dKTK7v71uJOBUy

literal 0
HcmV?d00001

diff --git a/tests/test_xiphcomment.cpp b/tests/test_xiphcomment.cpp
index d2a28a1f..64ae0b21 100644
--- a/tests/test_xiphcomment.cpp
+++ b/tests/test_xiphcomment.cpp
@@ -47,6 +47,7 @@ class TestXiphComment : public CppUnit::TestFixture
   CPPUNIT_TEST(testClearComment);
   CPPUNIT_TEST(testRemoveFields);
   CPPUNIT_TEST(testPicture);
+  CPPUNIT_TEST(testLowercaseFields);
   CPPUNIT_TEST_SUITE_END();
 
 public:
@@ -192,6 +193,23 @@ public:
     }
   }
 
+  void testLowercaseFields()
+  {
+    const ScopedFileCopy copy("lowercase-fields", ".ogg");
+    {
+      Vorbis::File f(copy.fileName().c_str());
+      List<FLAC::Picture *> lst = f.tag()->pictureList();
+      CPPUNIT_ASSERT_EQUAL(String("TEST TITLE"), f.tag()->title());
+      CPPUNIT_ASSERT_EQUAL(String("TEST ARTIST"), f.tag()->artist());
+      CPPUNIT_ASSERT_EQUAL((unsigned int)1, lst.size());
+      f.save();
+    }
+    {
+      Vorbis::File f(copy.fileName().c_str());
+      CPPUNIT_ASSERT(f.find("METADATA_BLOCK_PICTURE") > 0);
+    }
+  }
+
 };
 
 CPPUNIT_TEST_SUITE_REGISTRATION(TestXiphComment);

From d5b9d7b8a79ba77e7b803ffa2a04781b8c06de86 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Fri, 18 Nov 2016 13:55:43 +0900
Subject: [PATCH 05/24] Update NEWS.

---
 NEWS | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/NEWS b/NEWS
index 3c74bc45..1a185342 100644
--- a/NEWS
+++ b/NEWS
@@ -4,6 +4,8 @@
  * Added support for classical music tags of iTunes 12.5.
  * Dropped support for Windows 9x and NT 4.0 or older.
  * Fixed reading MP4 atoms with zero length.
+ * Fixed handling of redundant UTF-8 sequences in Win32.
+ * Fixed handling of lowercase field names in Vorbis Comments.
  * Several smaller bug fixes and performance improvements.
 
 TagLib 1.11.1 (Oct 24, 2016)

From eff28c55bf6cc3437dcfbc96449e3d2904370a74 Mon Sep 17 00:00:00 2001
From: mathbunnyru <mathbunnyru@gmail.com>
Date: Tue, 22 Nov 2016 01:10:28 +0300
Subject: [PATCH 06/24] Increment fixes

---
 taglib/toolkit/tstring.cpp | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/taglib/toolkit/tstring.cpp b/taglib/toolkit/tstring.cpp
index 44788249..b81b084e 100644
--- a/taglib/toolkit/tstring.cpp
+++ b/taglib/toolkit/tstring.cpp
@@ -508,7 +508,7 @@ ByteVector String::data(Type t) const
       ByteVector v(size(), 0);
       char *p = v.data();
 
-      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); it++)
+      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); ++it)
         *p++ = static_cast<char>(*it);
 
       return v;
@@ -537,7 +537,7 @@ ByteVector String::data(Type t) const
       *p++ = '\xff';
       *p++ = '\xfe';
 
-      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); it++) {
+      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); ++it) {
         *p++ = static_cast<char>(*it & 0xff);
         *p++ = static_cast<char>(*it >> 8);
       }
@@ -549,7 +549,7 @@ ByteVector String::data(Type t) const
       ByteVector v(size() * 2, 0);
       char *p = v.data();
 
-      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); it++) {
+      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); ++it) {
         *p++ = static_cast<char>(*it >> 8);
         *p++ = static_cast<char>(*it & 0xff);
       }
@@ -561,7 +561,7 @@ ByteVector String::data(Type t) const
       ByteVector v(size() * 2, 0);
       char *p = v.data();
 
-      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); it++) {
+      for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); ++it) {
         *p++ = static_cast<char>(*it & 0xff);
         *p++ = static_cast<char>(*it >> 8);
       }
@@ -611,7 +611,7 @@ String String::stripWhiteSpace() const
 
 bool String::isLatin1() const
 {
-  for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); it++) {
+  for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); ++it) {
     if(*it >= 256)
       return false;
   }
@@ -620,7 +620,7 @@ bool String::isLatin1() const
 
 bool String::isAscii() const
 {
-  for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); it++) {
+  for(wstring::const_iterator it = d->data.begin(); it != d->data.end(); ++it) {
     if(*it >= 128)
       return false;
   }

From 5e1d9fad31b65bd798381b41be7268dea919a920 Mon Sep 17 00:00:00 2001
From: mathbunnyru <mathbunnyru@gmail.com>
Date: Thu, 24 Nov 2016 02:02:38 +0300
Subject: [PATCH 07/24] Small fixes

---
 taglib/ape/apetag.cpp                             | 4 ++--
 taglib/asf/asfattribute.h                         | 2 +-
 taglib/mp4/mp4tag.cpp                             | 4 ++--
 taglib/mpeg/id3v2/frames/tableofcontentsframe.cpp | 4 +++-
 taglib/mpeg/id3v2/id3v2tag.cpp                    | 5 +++--
 taglib/xm/xmfile.cpp                              | 2 +-
 6 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/taglib/ape/apetag.cpp b/taglib/ape/apetag.cpp
index c31237eb..92ad536f 100644
--- a/taglib/ape/apetag.cpp
+++ b/taglib/ape/apetag.cpp
@@ -191,7 +191,7 @@ void APE::Tag::setGenre(const String &s)
 
 void APE::Tag::setYear(unsigned int i)
 {
-  if(i <= 0)
+  if(i == 0)
     removeItem("YEAR");
   else
     addValue("YEAR", String::number(i), true);
@@ -199,7 +199,7 @@ void APE::Tag::setYear(unsigned int i)
 
 void APE::Tag::setTrack(unsigned int i)
 {
-  if(i <= 0)
+  if(i == 0)
     removeItem("TRACK");
   else
     addValue("TRACK", String::number(i), true);
diff --git a/taglib/asf/asfattribute.h b/taglib/asf/asfattribute.h
index 64979216..1738eb45 100644
--- a/taglib/asf/asfattribute.h
+++ b/taglib/asf/asfattribute.h
@@ -113,7 +113,7 @@ namespace TagLib
       /*!
        * Copies the contents of \a other into this item.
        */
-      ASF::Attribute &operator=(const Attribute &other);
+      Attribute &operator=(const Attribute &other);
 
       /*!
        * Exchanges the content of the Attribute by the content of \a other.
diff --git a/taglib/mp4/mp4tag.cpp b/taglib/mp4/mp4tag.cpp
index 0c2e5cbc..a7a473a2 100644
--- a/taglib/mp4/mp4tag.cpp
+++ b/taglib/mp4/mp4tag.cpp
@@ -948,10 +948,10 @@ PropertyMap MP4::Tag::setProperties(const PropertyMap &props)
     if(reverseKeyMap.contains(it->first)) {
       String name = reverseKeyMap[it->first];
       if((it->first == "TRACKNUMBER" || it->first == "DISCNUMBER") && !it->second.isEmpty()) {
-        int first = 0, second = 0;
         StringList parts = StringList::split(it->second.front(), "/");
         if(!parts.isEmpty()) {
-          first = parts[0].toInt();
+          int first = parts[0].toInt();
+          int second = 0;
           if(parts.size() > 1) {
             second = parts[1].toInt();
           }
diff --git a/taglib/mpeg/id3v2/frames/tableofcontentsframe.cpp b/taglib/mpeg/id3v2/frames/tableofcontentsframe.cpp
index f7971250..b554d2cd 100644
--- a/taglib/mpeg/id3v2/frames/tableofcontentsframe.cpp
+++ b/taglib/mpeg/id3v2/frames/tableofcontentsframe.cpp
@@ -36,7 +36,9 @@ class TableOfContentsFrame::TableOfContentsFramePrivate
 {
 public:
   TableOfContentsFramePrivate() :
-    tagHeader(0)
+    tagHeader(0),
+    isTopLevel(false),
+    isOrdered(false)
   {
     embeddedFrameList.setAutoDelete(true);
   }
diff --git a/taglib/mpeg/id3v2/id3v2tag.cpp b/taglib/mpeg/id3v2/id3v2tag.cpp
index 4c00ab6f..810b2aef 100644
--- a/taglib/mpeg/id3v2/id3v2tag.cpp
+++ b/taglib/mpeg/id3v2/id3v2tag.cpp
@@ -60,6 +60,7 @@ class ID3v2::Tag::TagPrivate
 {
 public:
   TagPrivate() :
+    factory(0),
     file(0),
     tagOffset(0),
     extendedHeader(0),
@@ -286,7 +287,7 @@ void ID3v2::Tag::setGenre(const String &s)
 
 void ID3v2::Tag::setYear(unsigned int i)
 {
-  if(i <= 0) {
+  if(i == 0) {
     removeFrames("TDRC");
     return;
   }
@@ -295,7 +296,7 @@ void ID3v2::Tag::setYear(unsigned int i)
 
 void ID3v2::Tag::setTrack(unsigned int i)
 {
-  if(i <= 0) {
+  if(i == 0) {
     removeFrames("TRCK");
     return;
   }
diff --git a/taglib/xm/xmfile.cpp b/taglib/xm/xmfile.cpp
index 9192e9bf..216dfd8c 100644
--- a/taglib/xm/xmfile.cpp
+++ b/taglib/xm/xmfile.cpp
@@ -586,9 +586,9 @@ void XM::File::read(bool)
     unsigned int count = 4 + instrument.read(*this, instrumentHeaderSize - 4U);
     READ_ASSERT(count == std::min(instrumentHeaderSize, (unsigned long)instrument.size() + 4));
 
-    unsigned long sampleHeaderSize = 0;
     long offset = 0;
     if(sampleCount > 0) {
+      unsigned long sampleHeaderSize = 0;
       sumSampleCount += sampleCount;
       // wouldn't know which header size to assume otherwise:
       READ_ASSERT(instrumentHeaderSize >= count + 4 && readU32L(sampleHeaderSize));

From ad075a56f924b9b40c09ddff68a940655dc97c33 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Thu, 24 Nov 2016 14:45:22 +0900
Subject: [PATCH 08/24] Suppress MSVC warnings in test.

---
 tests/test_xiphcomment.cpp | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/tests/test_xiphcomment.cpp b/tests/test_xiphcomment.cpp
index 64ae0b21..386a3e67 100644
--- a/tests/test_xiphcomment.cpp
+++ b/tests/test_xiphcomment.cpp
@@ -98,8 +98,8 @@ public:
     map[""] = String("invalid key: empty string");
     map["A=B"] = String("invalid key: contains '='");
     map["A~B"] = String("invalid key: contains '~'");
-    map["A\x7F\B"] = String("invalid key: contains '\x7F'");
-    map[L"A\x3456\B"] = String("invalid key: Unicode");
+    map["A\x7F" "B"] = String("invalid key: contains '\x7F'");
+    map[L"A\x3456" "B"] = String("invalid key: Unicode");
 
     Ogg::XiphComment cmt;
     PropertyMap unsuccessful = cmt.setProperties(map);
@@ -113,8 +113,8 @@ public:
     cmt.addField("", "invalid key: empty string");
     cmt.addField("A=B", "invalid key: contains '='");
     cmt.addField("A~B", "invalid key: contains '~'");
-    cmt.addField("A\x7F\B", "invalid key: contains '\x7F'");
-    cmt.addField(L"A\x3456\B", "invalid key: Unicode");
+    cmt.addField("A\x7F" "B", "invalid key: contains '\x7F'");
+    cmt.addField(L"A\x3456" "B", "invalid key: Unicode");
     CPPUNIT_ASSERT_EQUAL(0U, cmt.fieldCount());
   }
 

From 6cfb11bb12bc2a83d5c36151538ad080ba6b7bde Mon Sep 17 00:00:00 2001
From: Martin Flaska <martin.flaska@legrand.us>
Date: Fri, 25 Nov 2016 12:58:25 +0100
Subject: [PATCH 09/24] test_string: Make 'stripWhiteSpace' test more complex

No string without leading/trailing spaces was used in the test.

Signed-off-by: Martin Flaska <martin.flaska@legrand.us>
---
 tests/test_string.cpp | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/tests/test_string.cpp b/tests/test_string.cpp
index 599e002d..b99c3f30 100644
--- a/tests/test_string.cpp
+++ b/tests/test_string.cpp
@@ -112,6 +112,9 @@ public:
     CPPUNIT_ASSERT(String("  foo  ").stripWhiteSpace() == String("foo"));
     CPPUNIT_ASSERT(String("foo    ").stripWhiteSpace() == String("foo"));
     CPPUNIT_ASSERT(String("    foo").stripWhiteSpace() == String("foo"));
+    CPPUNIT_ASSERT(String("foo").stripWhiteSpace() == String("foo"));
+    CPPUNIT_ASSERT(String("f o o").stripWhiteSpace() == String("f o o"));
+    CPPUNIT_ASSERT(String(" f o o ").stripWhiteSpace() == String("f o o"));
 
     CPPUNIT_ASSERT(memcmp(String("foo").data(String::Latin1).data(), "foo", 3) == 0);
     CPPUNIT_ASSERT(memcmp(String("f").data(String::Latin1).data(), "f", 1) == 0);

From c9a0754e3b4eff05491ac82ecd934aa86bf665d5 Mon Sep 17 00:00:00 2001
From: Martin Flaska <martin.flaska@legrand.us>
Date: Fri, 25 Nov 2016 15:32:26 +0100
Subject: [PATCH 10/24] tstring: String::substr optimization when returning
 itself as a substring

Use copy ctor to return in a case whole string is being returned.

The intention was to optimize String::stripWhiteSpace for no-strip case
(without any leading or trailing white space removal).

copyFromUTF16 was used in any case previously and allocated duplicate
buffer for the same string - no implicit sharing.

Signed-off-by: Martin Flaska <martin.flaska@legrand.us>
---
 taglib/toolkit/tstring.cpp | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/taglib/toolkit/tstring.cpp b/taglib/toolkit/tstring.cpp
index b81b084e..ee8d04d1 100644
--- a/taglib/toolkit/tstring.cpp
+++ b/taglib/toolkit/tstring.cpp
@@ -447,7 +447,10 @@ bool String::startsWith(const String &s) const
 
 String String::substr(unsigned int position, unsigned int n) const
 {
-  return String(d->data.substr(position, n));
+  if(position == 0 && n == size())
+    return *this;
+  else
+    return String(d->data.substr(position, n));
 }
 
 String &String::append(const String &s)

From d3062f3af4c4c0274ee7e886ece333a71814021b Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Sat, 26 Nov 2016 13:05:14 +0900
Subject: [PATCH 11/24] A bit more tolerant check to return itself in
 String::substr().

---
 taglib/toolkit/tstring.cpp | 2 +-
 tests/test_string.cpp      | 2 ++
 2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/taglib/toolkit/tstring.cpp b/taglib/toolkit/tstring.cpp
index ee8d04d1..1639d7c5 100644
--- a/taglib/toolkit/tstring.cpp
+++ b/taglib/toolkit/tstring.cpp
@@ -447,7 +447,7 @@ bool String::startsWith(const String &s) const
 
 String String::substr(unsigned int position, unsigned int n) const
 {
-  if(position == 0 && n == size())
+  if(position == 0 && n >= size())
     return *this;
   else
     return String(d->data.substr(position, n));
diff --git a/tests/test_string.cpp b/tests/test_string.cpp
index b99c3f30..4e3a1ac5 100644
--- a/tests/test_string.cpp
+++ b/tests/test_string.cpp
@@ -257,6 +257,8 @@ public:
     CPPUNIT_ASSERT_EQUAL(String("01"), String("0123456").substr(0, 2));
     CPPUNIT_ASSERT_EQUAL(String("12"), String("0123456").substr(1, 2));
     CPPUNIT_ASSERT_EQUAL(String("123456"), String("0123456").substr(1, 200));
+    CPPUNIT_ASSERT_EQUAL(String("0123456"), String("0123456").substr(0, 7));
+    CPPUNIT_ASSERT_EQUAL(String("0123456"), String("0123456").substr(0, 200));
   }
 
   void testNewline()

From 2786aa7463700aeefa358cbbfeccb9ff4839eadd Mon Sep 17 00:00:00 2001
From: Michael Helmling <michaelhelmling@posteo.de>
Date: Sun, 27 Nov 2016 19:17:13 +0100
Subject: [PATCH 12/24] Fixes #743 by not overwriting existing Xiph comment in
 FLAC::File::save

---
 taglib/flac/flacfile.cpp | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/taglib/flac/flacfile.cpp b/taglib/flac/flacfile.cpp
index b31cc65e..dba54d58 100644
--- a/taglib/flac/flacfile.cpp
+++ b/taglib/flac/flacfile.cpp
@@ -168,8 +168,9 @@ bool FLAC::File::save()
   }
 
   // Create new vorbis comments
-
-  Tag::duplicate(&d->tag, xiphComment(true), false);
+  if (!hasXiphComment())
+    Tag::duplicate(&d->tag, xiphComment(true), false);
+  
 
   d->xiphCommentData = xiphComment()->render(false);
 

From 6df61cf2af02189d37f783b63745efca00d1bd34 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Tue, 29 Nov 2016 10:38:11 +0900
Subject: [PATCH 13/24] Small fix in style.

---
 taglib/flac/flacfile.cpp | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/taglib/flac/flacfile.cpp b/taglib/flac/flacfile.cpp
index dba54d58..93a0e952 100644
--- a/taglib/flac/flacfile.cpp
+++ b/taglib/flac/flacfile.cpp
@@ -168,9 +168,8 @@ bool FLAC::File::save()
   }
 
   // Create new vorbis comments
-  if (!hasXiphComment())
+  if(!hasXiphComment())
     Tag::duplicate(&d->tag, xiphComment(true), false);
-  
 
   d->xiphCommentData = xiphComment()->render(false);
 

From 4381bd75f3a41f1507b13922dade6efce0da7ec2 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Tue, 29 Nov 2016 10:53:33 +0900
Subject: [PATCH 14/24] Add a test for #743/#779.

---
 tests/test_flac.cpp | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/tests/test_flac.cpp b/tests/test_flac.cpp
index 6a306981..f2065e09 100644
--- a/tests/test_flac.cpp
+++ b/tests/test_flac.cpp
@@ -62,6 +62,7 @@ class TestFLAC : public CppUnit::TestFixture
   CPPUNIT_TEST(testUpdateID3v2);
   CPPUNIT_TEST(testEmptyID3v2);
   CPPUNIT_TEST(testStripTags);
+  CPPUNIT_TEST(testRemoveXiphField);
   CPPUNIT_TEST_SUITE_END();
 
 public:
@@ -493,6 +494,28 @@ public:
     }
   }
 
+  void testRemoveXiphField()
+  {
+    ScopedFileCopy copy("silence-44-s", ".flac");
+
+    {
+      FLAC::File f(copy.fileName().c_str());
+      f.xiphComment(true)->setTitle("XiphComment Title");
+      f.ID3v2Tag(true)->setTitle("ID3v2 Title");
+      f.save();
+    }
+    {
+      FLAC::File f(copy.fileName().c_str());
+      CPPUNIT_ASSERT_EQUAL(String("XiphComment Title"), f.xiphComment()->title());
+      f.xiphComment()->removeFields("TITLE");
+      f.save();
+    }
+    {
+      FLAC::File f(copy.fileName().c_str());
+      CPPUNIT_ASSERT_EQUAL(String(), f.xiphComment()->title());
+    }
+  }
+
 };
 
 CPPUNIT_TEST_SUITE_REGISTRATION(TestFLAC);

From 046c98230f5f458932fe24bc3e936ed4a47c3ce7 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Tue, 29 Nov 2016 14:58:39 +0900
Subject: [PATCH 15/24] Remove Utils::floatByteOrder() and use
 systemByteOrder() instead.

We can safely assume that the integer and float byte orders are the same on IEEE754 compliant systems.
---
 taglib/toolkit/tbytevector.cpp |  4 ++--
 taglib/toolkit/tutils.h        | 22 +---------------------
 2 files changed, 3 insertions(+), 23 deletions(-)

diff --git a/taglib/toolkit/tbytevector.cpp b/taglib/toolkit/tbytevector.cpp
index a2258bc5..3d6daed0 100644
--- a/taglib/toolkit/tbytevector.cpp
+++ b/taglib/toolkit/tbytevector.cpp
@@ -176,7 +176,7 @@ TFloat toFloat(const ByteVector &v, size_t offset)
   } tmp;
   ::memcpy(&tmp, v.data() + offset, sizeof(TInt));
 
-  if(ENDIAN != Utils::floatByteOrder())
+  if(ENDIAN != Utils::systemByteOrder())
     tmp.i = Utils::byteSwap(tmp.i);
 
   return tmp.f;
@@ -191,7 +191,7 @@ ByteVector fromFloat(TFloat value)
   } tmp;
   tmp.f = value;
 
-  if(ENDIAN != Utils::floatByteOrder())
+  if(ENDIAN != Utils::systemByteOrder())
     tmp.i = Utils::byteSwap(tmp.i);
 
   return ByteVector(reinterpret_cast<char *>(&tmp), sizeof(TInt));
diff --git a/taglib/toolkit/tutils.h b/taglib/toolkit/tutils.h
index d4051f64..b114619e 100644
--- a/taglib/toolkit/tutils.h
+++ b/taglib/toolkit/tutils.h
@@ -233,7 +233,7 @@ namespace TagLib
       };
 
       /*!
-       * Returns the integer byte order of the system.
+       * Returns the byte order of the system.
        */
       inline ByteOrder systemByteOrder()
       {
@@ -248,26 +248,6 @@ namespace TagLib
         else
           return BigEndian;
       }
-
-      /*!
-       * Returns the IEEE754 byte order of the system.
-       */
-      inline ByteOrder floatByteOrder()
-      {
-        union {
-          double d;
-          char   c;
-        } u;
-
-        // 1.0 is stored in memory like 0x3FF0000000000000 in canonical form.
-        // So the first byte is zero if little endian.
-
-        u.d = 1.0;
-        if(u.c == 0)
-          return LittleEndian;
-        else
-          return BigEndian;
-      }
     }
   }
 }

From cfbaf3459776e45cdbf2954de57bde8b689264fa Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Thu, 1 Dec 2016 10:50:30 +0900
Subject: [PATCH 16/24] Prevent the segment table of Ogg pages from exceeding
 the size limit.

---
 taglib/ogg/oggpage.cpp |  8 ++++----
 tests/test_ogg.cpp     | 32 ++++++++++++++++++++++++++++++--
 2 files changed, 34 insertions(+), 6 deletions(-)

diff --git a/taglib/ogg/oggpage.cpp b/taglib/ogg/oggpage.cpp
index 75aea22a..414d3d53 100644
--- a/taglib/ogg/oggpage.cpp
+++ b/taglib/ogg/oggpage.cpp
@@ -208,15 +208,15 @@ List<Ogg::Page *> Ogg::Page::paginate(const ByteVectorList &packets,
 
   static const unsigned int SplitSize = 32 * 255;
 
-  // Force repagination if the packets are too large for a page.
+  // Force repagination if the segment table will exceed the size limit.
 
   if(strategy != Repaginate) {
 
-    size_t totalSize = packets.size();
+    size_t tableSize = 0;
     for(ByteVectorList::ConstIterator it = packets.begin(); it != packets.end(); ++it)
-      totalSize += it->size();
+      tableSize += it->size() / 255 + 1;
 
-    if(totalSize > 255 * 255)
+    if(tableSize > 255)
       strategy = Repaginate;
   }
 
diff --git a/tests/test_ogg.cpp b/tests/test_ogg.cpp
index af2d5b94..5569e59c 100644
--- a/tests/test_ogg.cpp
+++ b/tests/test_ogg.cpp
@@ -42,7 +42,8 @@ class TestOGG : public CppUnit::TestFixture
 {
   CPPUNIT_TEST_SUITE(TestOGG);
   CPPUNIT_TEST(testSimple);
-  CPPUNIT_TEST(testSplitPackets);
+  CPPUNIT_TEST(testSplitPackets1);
+  CPPUNIT_TEST(testSplitPackets2);
   CPPUNIT_TEST(testDictInterface1);
   CPPUNIT_TEST(testDictInterface2);
   CPPUNIT_TEST(testAudioProperties);
@@ -67,7 +68,7 @@ public:
     }
   }
 
-  void testSplitPackets()
+  void testSplitPackets1()
   {
     ScopedFileCopy copy("empty", ".ogg");
     string newname = copy.fileName();
@@ -110,6 +111,33 @@ public:
     }
   }
 
+  void testSplitPackets2()
+  {
+    ScopedFileCopy copy("empty", ".ogg");
+    string newname = copy.fileName();
+
+    const String text = longText(60890, true);
+
+    {
+      Vorbis::File f(newname.c_str());
+      f.tag()->setTitle(text);
+      f.save();
+    }
+    {
+      Vorbis::File f(newname.c_str());
+      CPPUNIT_ASSERT(f.isValid());
+      CPPUNIT_ASSERT_EQUAL(text, f.tag()->title());
+
+      f.tag()->setTitle("ABCDE");
+      f.save();
+    }
+    {
+      Vorbis::File f(newname.c_str());
+      CPPUNIT_ASSERT(f.isValid());
+      CPPUNIT_ASSERT_EQUAL(String("ABCDE"), f.tag()->title());
+    }
+  }
+
   void testDictInterface1()
   {
     ScopedFileCopy copy("empty", ".ogg");

From 9336c82da3a04552168f208cd7a5fa4646701ea4 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Thu, 1 Dec 2016 11:32:01 +0900
Subject: [PATCH 17/24] Fix possible Ogg packet losses.

---
 taglib/ogg/oggfile.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/taglib/ogg/oggfile.cpp b/taglib/ogg/oggfile.cpp
index 86b0b076..c36e4d46 100644
--- a/taglib/ogg/oggfile.cpp
+++ b/taglib/ogg/oggfile.cpp
@@ -253,7 +253,7 @@ void Ogg::File::writePacket(unsigned int i, const ByteVector &packet)
   ByteVectorList packets = firstPage->packets();
   packets[i - firstPage->firstPacketIndex()] = packet;
 
-  if(firstPage != lastPage && lastPage->packetCount() > 2) {
+  if(firstPage != lastPage && lastPage->packetCount() > 1) {
     ByteVectorList lastPagePackets = lastPage->packets();
     lastPagePackets.erase(lastPagePackets.begin());
     packets.append(lastPagePackets);

From 489e2e6cbbce49712ce149208b057e1e4788fb9f Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Thu, 1 Dec 2016 15:25:30 +0900
Subject: [PATCH 18/24] Update NEWS.

---
 NEWS | 1 +
 1 file changed, 1 insertion(+)

diff --git a/NEWS b/NEWS
index 1a185342..774e6532 100644
--- a/NEWS
+++ b/NEWS
@@ -6,6 +6,7 @@
  * Fixed reading MP4 atoms with zero length.
  * Fixed handling of redundant UTF-8 sequences in Win32.
  * Fixed handling of lowercase field names in Vorbis Comments.
+ * Fixed possible file corruptions when saving Ogg files.
  * Several smaller bug fixes and performance improvements.
 
 TagLib 1.11.1 (Oct 24, 2016)

From f6a604f54b27b5f8c21f0d7ccda756029ed35ac6 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Fri, 2 Dec 2016 17:26:43 +0900
Subject: [PATCH 19/24] #include guards in CMake generated headers.

---
 config.h.cmake               | 5 +++++
 taglib/taglib_config.h.cmake | 5 +++++
 2 files changed, 10 insertions(+)

diff --git a/config.h.cmake b/config.h.cmake
index 7eb5993f..90ae2eba 100644
--- a/config.h.cmake
+++ b/config.h.cmake
@@ -1,5 +1,8 @@
 /* config.h. Generated by cmake from config.h.cmake */
 
+#ifndef TAGLIB_CONFIG_H
+#define TAGLIB_CONFIG_H
+
 /* Defined if your compiler supports some byte swap functions */
 #cmakedefine   HAVE_BOOST_BYTESWAP 1
 #cmakedefine   HAVE_GCC_BYTESWAP 1
@@ -31,3 +34,5 @@
 #cmakedefine   TRACE_IN_RELEASE 1
 
 #cmakedefine TESTS_DIR "@TESTS_DIR@"
+
+#endif
diff --git a/taglib/taglib_config.h.cmake b/taglib/taglib_config.h.cmake
index 5f0ee6cf..915f130a 100644
--- a/taglib/taglib_config.h.cmake
+++ b/taglib/taglib_config.h.cmake
@@ -1,6 +1,11 @@
 /* taglib_config.h. Generated by cmake from taglib_config.h.cmake */
 
+#ifndef TAGLIB_TAGLIB_CONFIG_H
+#define TAGLIB_TAGLIB_CONFIG_H
+
 /* These values are no longer used. This file is present only for compatibility reasons. */
 
 #define   TAGLIB_WITH_ASF 1
 #define   TAGLIB_WITH_MP4 1
+
+#endif

From b00a5c1aab4603819cce2a2d4946c978c60d8fd0 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Mon, 5 Dec 2016 10:15:26 +0900
Subject: [PATCH 20/24] Add a test to check if ByteVector is detached correctly
 when being replaced.

---
 tests/test_bytevector.cpp | 39 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)

diff --git a/tests/test_bytevector.cpp b/tests/test_bytevector.cpp
index 1e063b71..428e795c 100644
--- a/tests/test_bytevector.cpp
+++ b/tests/test_bytevector.cpp
@@ -46,6 +46,7 @@ class TestByteVector : public CppUnit::TestFixture
   CPPUNIT_TEST(testIntegerConversion);
   CPPUNIT_TEST(testFloatingPointConversion);
   CPPUNIT_TEST(testReplace);
+  CPPUNIT_TEST(testReplaceAndDetach);
   CPPUNIT_TEST(testIterator);
   CPPUNIT_TEST(testResize);
   CPPUNIT_TEST(testAppend1);
@@ -312,6 +313,44 @@ public:
       CPPUNIT_ASSERT_EQUAL(ByteVector("abcdaba"), a);
     }
   }
+  void testReplaceAndDetach()
+  {
+    {
+      ByteVector a("abcdabf");
+      ByteVector b = a;
+      a.replace(ByteVector("a"), ByteVector("x"));
+      CPPUNIT_ASSERT_EQUAL(ByteVector("xbcdxbf"), a);
+      CPPUNIT_ASSERT_EQUAL(ByteVector("abcdabf"), b);
+    }
+    {
+      ByteVector a("abcdabf");
+      ByteVector b = a;
+      a.replace('a', 'x');
+      CPPUNIT_ASSERT_EQUAL(ByteVector("xbcdxbf"), a);
+      CPPUNIT_ASSERT_EQUAL(ByteVector("abcdabf"), b);
+    }
+    {
+      ByteVector a("abcdabf");
+      ByteVector b = a;
+      a.replace(ByteVector("ab"), ByteVector("xy"));
+      CPPUNIT_ASSERT_EQUAL(ByteVector("xycdxyf"), a);
+      CPPUNIT_ASSERT_EQUAL(ByteVector("abcdabf"), b);
+    }
+    {
+      ByteVector a("abcdabf");
+      ByteVector b = a;
+      a.replace(ByteVector("a"), ByteVector("<a>"));
+      CPPUNIT_ASSERT_EQUAL(ByteVector("<a>bcd<a>bf"), a);
+      CPPUNIT_ASSERT_EQUAL(ByteVector("abcdabf"), b);
+    }
+    {
+      ByteVector a("ab<c>dab<c>");
+      ByteVector b = a;
+      a.replace(ByteVector("<c>"), ByteVector("c"));
+      CPPUNIT_ASSERT_EQUAL(ByteVector("abcdabc"), a);
+      CPPUNIT_ASSERT_EQUAL(ByteVector("ab<c>dab<c>"), b);
+    }
+  }
 
   void testIterator()
   {

From 36ccad2bd4b5b8aec1e547faef3bfe0269316ae9 Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Fri, 2 Dec 2016 16:31:27 +0900
Subject: [PATCH 21/24] Rewrite ByteVector::replace() to run in O(n) time.

---
 taglib/toolkit/tbytevector.cpp | 77 ++++++++++++++++++++--------------
 tests/test_bytevector.cpp      |  1 +
 2 files changed, 46 insertions(+), 32 deletions(-)

diff --git a/taglib/toolkit/tbytevector.cpp b/taglib/toolkit/tbytevector.cpp
index 3d6daed0..d272057f 100644
--- a/taglib/toolkit/tbytevector.cpp
+++ b/taglib/toolkit/tbytevector.cpp
@@ -29,7 +29,6 @@
 #include <cmath>
 #include <cstdio>
 #include <cstring>
-#include <cstddef>
 
 #include <tstring.h>
 #include <tdebug.h>
@@ -485,46 +484,60 @@ ByteVector &ByteVector::replace(char oldByte, char newByte)
 
 ByteVector &ByteVector::replace(const ByteVector &pattern, const ByteVector &with)
 {
-  // TODO: This takes O(n!) time in the worst case. Rewrite it to run in O(n) time.
-
-  if(pattern.size() == 0 || pattern.size() > size())
-    return *this;
-
   if(pattern.size() == 1 && with.size() == 1)
     return replace(pattern[0], with[0]);
 
-  const unsigned int withSize    = with.size();
-  const unsigned int patternSize = pattern.size();
-  const int diff = withSize - patternSize;
+  // Check if there is at least one occurrence of the pattern.
 
-  unsigned int offset = 0;
-  while (true) {
-    offset = find(pattern, offset);
-    if(offset == static_cast<unsigned int>(-1))
-      break;
+  int offset = find(pattern, 0);
+  if(offset == -1)
+    return *this;
+
+  if(pattern.size() == with.size()) {
+
+    // We think this case might be common enough to optimize it.
 
     detach();
+    do
+    {
+      ::memcpy(data() + offset, with.data(), with.size());
+      offset = find(pattern, offset + pattern.size());
+    } while(offset != -1);
+  }
+  else {
 
-    if(diff < 0) {
-      ::memmove(
-        data() + offset + withSize,
-        data() + offset + patternSize,
-        size() - offset - patternSize);
-      resize(size() + diff);
-    }
-    else if(diff > 0) {
-      resize(size() + diff);
-      ::memmove(
-        data() + offset + withSize,
-        data() + offset + patternSize,
-        size() - diff - offset - patternSize);
+    // Loop once to calculate the result size.
+
+    unsigned int dstSize = size();
+    do
+    {
+      dstSize += with.size() - pattern.size();
+      offset = find(pattern, offset + pattern.size());
+    } while(offset != -1);
+
+    // Loop again to copy modified data to the new vector.
+
+    ByteVector dst(dstSize);
+    int dstOffset = 0;
+
+    offset = 0;
+    while(true) {
+      const int next = find(pattern, offset);
+      if(next == -1) {
+        ::memcpy(dst.data() + dstOffset, data() + offset, size() - offset);
+        break;
+      }
+
+      ::memcpy(dst.data() + dstOffset, data() + offset, next - offset);
+      dstOffset += next - offset;
+
+      ::memcpy(dst.data() + dstOffset, with.data(), with.size());
+      dstOffset += with.size();
+
+      offset = next + pattern.size();
     }
 
-    ::memcpy(data() + offset, with.data(), with.size());
-
-    offset += withSize;
-    if(offset > size() - patternSize)
-      break;
+    swap(dst);
   }
 
   return *this;
diff --git a/tests/test_bytevector.cpp b/tests/test_bytevector.cpp
index 428e795c..f03dda1a 100644
--- a/tests/test_bytevector.cpp
+++ b/tests/test_bytevector.cpp
@@ -313,6 +313,7 @@ public:
       CPPUNIT_ASSERT_EQUAL(ByteVector("abcdaba"), a);
     }
   }
+
   void testReplaceAndDetach()
   {
     {

From b5115e3497328d80a9accf8c1c303374ec36e884 Mon Sep 17 00:00:00 2001
From: Hao Xi <kevin@picturekeeper.com>
Date: Thu, 8 Dec 2016 12:53:40 +0800
Subject: [PATCH 22/24] Fix #667: Compiled TagLib framework for OS X fails at
 codesign.

---
 CMakeLists.txt        | 13 ++++++++-----
 ConfigureChecks.cmake | 15 ++++++++++-----
 taglib/CMakeLists.txt |  8 +++++++-
 3 files changed, 25 insertions(+), 11 deletions(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index a59efc9d..bbd7d7cc 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -13,6 +13,14 @@ if(DEFINED ENABLE_STATIC)
 endif()
 
 option(BUILD_SHARED_LIBS "Build shared libraries" OFF)
+if(APPLE)
+  option(BUILD_FRAMEWORK "Build an OS X framework" OFF)
+  if(BUILD_FRAMEWORK)
+    set(BUILD_SHARED_LIBS ON)
+    #set(CMAKE_MACOSX_RPATH 1)
+    set(FRAMEWORK_INSTALL_DIR "/Library/Frameworks" CACHE STRING "Directory to install frameworks to.")
+  endif()
+endif()
 if(NOT BUILD_SHARED_LIBS)
   add_definitions(-DTAGLIB_STATIC)
 endif()
@@ -48,11 +56,6 @@ set(BIN_INSTALL_DIR "${EXEC_INSTALL_PREFIX}/bin" CACHE PATH "The subdirectory to
 set(LIB_INSTALL_DIR "${EXEC_INSTALL_PREFIX}/lib${LIB_SUFFIX}" CACHE PATH "The subdirectory relative to the install prefix where libraries will be installed (default is /lib${LIB_SUFFIX})")
 set(INCLUDE_INSTALL_DIR "${CMAKE_INSTALL_PREFIX}/include" CACHE PATH "The subdirectory to the header prefix")
 
-if(APPLE)
-  option(BUILD_FRAMEWORK "Build an OS X framework" OFF)
-  set(FRAMEWORK_INSTALL_DIR "/Library/Frameworks" CACHE STRING "Directory to install frameworks to.")
-endif()
-
 if(CMAKE_COMPILER_IS_GNUCC OR CMAKE_COMPILER_IS_GNUCXX)
   set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Wall")
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall")
diff --git a/ConfigureChecks.cmake b/ConfigureChecks.cmake
index ee4fdc2e..fe365711 100644
--- a/ConfigureChecks.cmake
+++ b/ConfigureChecks.cmake
@@ -54,11 +54,16 @@ check_cxx_source_compiles("
 " HAVE_STD_ATOMIC)
 
 if(NOT HAVE_STD_ATOMIC)
-  find_package(Boost COMPONENTS atomic)
-  if(Boost_ATOMIC_FOUND)
-    set(HAVE_BOOST_ATOMIC 1)
-  else()
-    set(HAVE_BOOST_ATOMIC 0)
+  
+  # We will not find BOOST_ATOMIC on macOS when BUILD_FRAMEWORK is set, since we don't want to link
+  # to `libboost_atomic-mt.dylib` within `tag.framework`. 
+  if(NOT BUILD_FRAMEWORK)
+    find_package(Boost COMPONENTS atomic)
+    if(Boost_ATOMIC_FOUND)
+      set(HAVE_BOOST_ATOMIC 1)
+    else()
+      set(HAVE_BOOST_ATOMIC 0)
+    endif()
   endif()
 
   if(NOT HAVE_BOOST_ATOMIC)
diff --git a/taglib/CMakeLists.txt b/taglib/CMakeLists.txt
index 000f7937..70ff8d1b 100644
--- a/taglib/CMakeLists.txt
+++ b/taglib/CMakeLists.txt
@@ -365,7 +365,13 @@ set_target_properties(tag PROPERTIES
   PUBLIC_HEADER "${tag_HDRS}"
 )
 if(BUILD_FRAMEWORK)
-  set_target_properties(tag PROPERTIES FRAMEWORK TRUE)
+  unset(INSTALL_NAME_DIR)
+  set_target_properties(tag PROPERTIES 
+      FRAMEWORK TRUE
+      MACOSX_RPATH 1
+      VERSION "A"
+      SOVERSION "A"
+      )
 endif()
 
 install(TARGETS tag

From 250c59f783f11e5b23eea4cb73063b6ecb28453e Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Fri, 9 Dec 2016 09:42:29 +0900
Subject: [PATCH 23/24] Remove optional dependencies on Boost's dynamic
 libraries.

Using precompiled Boost libraries can lead to depending on external dynamic libraries.
---
 ConfigureChecks.cmake          | 75 ++++++++++++----------------------
 config.h.cmake                 |  2 -
 taglib/CMakeLists.txt          | 12 +-----
 taglib/toolkit/trefcounter.cpp |  5 ---
 taglib/toolkit/tzlib.cpp       | 46 +++------------------
 5 files changed, 33 insertions(+), 107 deletions(-)

diff --git a/ConfigureChecks.cmake b/ConfigureChecks.cmake
index fe365711..253af335 100644
--- a/ConfigureChecks.cmake
+++ b/ConfigureChecks.cmake
@@ -54,61 +54,47 @@ check_cxx_source_compiles("
 " HAVE_STD_ATOMIC)
 
 if(NOT HAVE_STD_ATOMIC)
-  
-  # We will not find BOOST_ATOMIC on macOS when BUILD_FRAMEWORK is set, since we don't want to link
-  # to `libboost_atomic-mt.dylib` within `tag.framework`. 
-  if(NOT BUILD_FRAMEWORK)
-    find_package(Boost COMPONENTS atomic)
-    if(Boost_ATOMIC_FOUND)
-      set(HAVE_BOOST_ATOMIC 1)
-    else()
-      set(HAVE_BOOST_ATOMIC 0)
-    endif()
-  endif()
+  check_cxx_source_compiles("
+    int main() {
+      volatile int x;
+      __sync_add_and_fetch(&x, 1);
+      int y = __sync_sub_and_fetch(&x, 1);
+      return 0;
+    }
+  " HAVE_GCC_ATOMIC)
 
-  if(NOT HAVE_BOOST_ATOMIC)
+  if(NOT HAVE_GCC_ATOMIC)
     check_cxx_source_compiles("
+      #include <libkern/OSAtomic.h>
       int main() {
-        volatile int x;
-        __sync_add_and_fetch(&x, 1);
-        int y = __sync_sub_and_fetch(&x, 1);
+        volatile int32_t x;
+        OSAtomicIncrement32Barrier(&x);
+        int32_t y = OSAtomicDecrement32Barrier(&x);
         return 0;
       }
-    " HAVE_GCC_ATOMIC)
+    " HAVE_MAC_ATOMIC)
 
-    if(NOT HAVE_GCC_ATOMIC)
+    if(NOT HAVE_MAC_ATOMIC)
       check_cxx_source_compiles("
-        #include <libkern/OSAtomic.h>
+        #include <windows.h>
         int main() {
-          volatile int32_t x;
-          OSAtomicIncrement32Barrier(&x);
-          int32_t y = OSAtomicDecrement32Barrier(&x);
+          volatile LONG x;
+          InterlockedIncrement(&x);
+          LONG y = InterlockedDecrement(&x);
           return 0;
         }
-      " HAVE_MAC_ATOMIC)
+      " HAVE_WIN_ATOMIC)
 
-      if(NOT HAVE_MAC_ATOMIC)
+      if(NOT HAVE_WIN_ATOMIC)
         check_cxx_source_compiles("
-          #include <windows.h>
+          #include <ia64intrin.h>
           int main() {
-            volatile LONG x;
-            InterlockedIncrement(&x);
-            LONG y = InterlockedDecrement(&x);
+            volatile int x;
+            __sync_add_and_fetch(&x, 1);
+            int y = __sync_sub_and_fetch(&x, 1);
             return 0;
           }
-        " HAVE_WIN_ATOMIC)
-
-        if(NOT HAVE_WIN_ATOMIC)
-          check_cxx_source_compiles("
-            #include <ia64intrin.h>
-            int main() {
-              volatile int x;
-              __sync_add_and_fetch(&x, 1);
-              int y = __sync_sub_and_fetch(&x, 1);
-              return 0;
-            }
-          " HAVE_IA64_ATOMIC)
-        endif()
+        " HAVE_IA64_ATOMIC)
       endif()
     endif()
   endif()
@@ -230,15 +216,6 @@ if(NOT ZLIB_SOURCE)
   else()
     set(HAVE_ZLIB 0)
   endif()
-
-  if(NOT HAVE_ZLIB)
-    find_package(Boost COMPONENTS iostreams zlib)
-    if(Boost_IOSTREAMS_FOUND AND Boost_ZLIB_FOUND)
-      set(HAVE_BOOST_ZLIB 1)
-    else()
-      set(HAVE_BOOST_ZLIB 0)
-    endif()
-  endif()
 endif()
 
 # Determine whether CppUnit is installed.
diff --git a/config.h.cmake b/config.h.cmake
index 90ae2eba..fd716134 100644
--- a/config.h.cmake
+++ b/config.h.cmake
@@ -13,7 +13,6 @@
 
 /* Defined if your compiler supports some atomic operations */
 #cmakedefine   HAVE_STD_ATOMIC 1
-#cmakedefine   HAVE_BOOST_ATOMIC 1
 #cmakedefine   HAVE_GCC_ATOMIC 1
 #cmakedefine   HAVE_MAC_ATOMIC 1
 #cmakedefine   HAVE_WIN_ATOMIC 1
@@ -28,7 +27,6 @@
 
 /* Defined if zlib is installed */
 #cmakedefine   HAVE_ZLIB 1
-#cmakedefine   HAVE_BOOST_ZLIB 1
 
 /* Indicates whether debug messages are shown even in release mode */
 #cmakedefine   TRACE_IN_RELEASE 1
diff --git a/taglib/CMakeLists.txt b/taglib/CMakeLists.txt
index 70ff8d1b..4ea54304 100644
--- a/taglib/CMakeLists.txt
+++ b/taglib/CMakeLists.txt
@@ -32,7 +32,7 @@ elseif(HAVE_ZLIB_SOURCE)
   include_directories(${ZLIB_SOURCE})
 endif()
 
-if(HAVE_BOOST_BYTESWAP OR HAVE_BOOST_ATOMIC OR HAVE_BOOST_ZLIB)
+if(Boost_FOUND)
   include_directories(${Boost_INCLUDE_DIR})
 endif()
 
@@ -344,18 +344,10 @@ set(tag_LIB_SRCS
 
 add_library(tag ${tag_LIB_SRCS} ${tag_HDRS})
 
-if(ZLIB_FOUND)
+if(HAVE_ZLIB AND NOT HAVE_ZLIB_SOURCE)
   target_link_libraries(tag ${ZLIB_LIBRARIES})
 endif()
 
-if(HAVE_BOOST_ATOMIC)
-  target_link_libraries(tag ${Boost_ATOMIC_LIBRARY})
-endif()
-
-if(HAVE_BOOST_ZLIB)
-  target_link_libraries(tag ${Boost_IOSTREAMS_LIBRARY} ${Boost_ZLIB_LIBRARY})
-endif()
-
 set_target_properties(tag PROPERTIES
   VERSION ${TAGLIB_SOVERSION_MAJOR}.${TAGLIB_SOVERSION_MINOR}.${TAGLIB_SOVERSION_PATCH}
   SOVERSION ${TAGLIB_SOVERSION_MAJOR}
diff --git a/taglib/toolkit/trefcounter.cpp b/taglib/toolkit/trefcounter.cpp
index 27d17b83..eb2aa69f 100644
--- a/taglib/toolkit/trefcounter.cpp
+++ b/taglib/toolkit/trefcounter.cpp
@@ -34,11 +34,6 @@
 # define ATOMIC_INT std::atomic<unsigned int>
 # define ATOMIC_INC(x) x.fetch_add(1)
 # define ATOMIC_DEC(x) (x.fetch_sub(1) - 1)
-#elif defined(HAVE_BOOST_ATOMIC)
-# include <boost/atomic.hpp>
-# define ATOMIC_INT boost::atomic<unsigned int>
-# define ATOMIC_INC(x) x.fetch_add(1)
-# define ATOMIC_DEC(x) (x.fetch_sub(1) - 1)
 #elif defined(HAVE_GCC_ATOMIC)
 # define ATOMIC_INT int
 # define ATOMIC_INC(x) __sync_add_and_fetch(&x, 1)
diff --git a/taglib/toolkit/tzlib.cpp b/taglib/toolkit/tzlib.cpp
index 40158fd2..6d07ba3d 100644
--- a/taglib/toolkit/tzlib.cpp
+++ b/taglib/toolkit/tzlib.cpp
@@ -27,23 +27,19 @@
 # include <config.h>
 #endif
 
-#if defined(HAVE_ZLIB)
+#ifdef HAVE_ZLIB
 # include <zlib.h>
-#elif defined(HAVE_BOOST_ZLIB)
-# include <boost/iostreams/filtering_streambuf.hpp>
-# include <boost/iostreams/filter/zlib.hpp>
+# include <tstring.h>
+# include <tdebug.h>
 #endif
 
-#include <tstring.h>
-#include <tdebug.h>
-
 #include "tzlib.h"
 
 using namespace TagLib;
 
 bool zlib::isAvailable()
 {
-#if defined(HAVE_ZLIB) || defined(HAVE_BOOST_ZLIB)
+#ifdef HAVE_ZLIB
 
   return true;
 
@@ -56,7 +52,7 @@ bool zlib::isAvailable()
 
 ByteVector zlib::decompress(const ByteVector &data)
 {
-#if defined(HAVE_ZLIB)
+#ifdef HAVE_ZLIB
 
   z_stream stream = {};
 
@@ -102,38 +98,6 @@ ByteVector zlib::decompress(const ByteVector &data)
 
   return outData;
 
-#elif defined(HAVE_BOOST_ZLIB)
-
-  using namespace boost::iostreams;
-
-  struct : public sink
-  {
-    ByteVector data;
-
-    typedef char     char_type;
-    typedef sink_tag category;
-
-    std::streamsize write(char const* s, std::streamsize n)
-    {
-      const unsigned int originalSize = data.size();
-
-      data.resize(static_cast<unsigned int>(originalSize + n));
-      ::memcpy(data.data() + originalSize, s, static_cast<size_t>(n));
-
-      return n;
-    }
-  } sink;
-
-  try {
-    zlib_decompressor().write(sink, data.data(), data.size());
-  }
-  catch(const zlib_error &) {
-    debug("zlib::decompress() - Error reading compressed stream.");
-    return ByteVector();
-  }
-
-  return sink.data;
-
 #else
 
   return ByteVector();

From a19a623d4bfaa7d26ece05ac73c4716e5710d34f Mon Sep 17 00:00:00 2001
From: Tsuda Kageyu <tsuda.kageyu@gmail.com>
Date: Fri, 9 Dec 2016 09:56:37 +0900
Subject: [PATCH 24/24] Make use of increment/decrement operators of
 std::atomic.

---
 ConfigureChecks.cmake          | 6 +++---
 taglib/toolkit/trefcounter.cpp | 6 +++---
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/ConfigureChecks.cmake b/ConfigureChecks.cmake
index 253af335..4b55f273 100644
--- a/ConfigureChecks.cmake
+++ b/ConfigureChecks.cmake
@@ -46,9 +46,9 @@ endif()
 check_cxx_source_compiles("
   #include <atomic>
   int main() {
-    std::atomic<unsigned int> x;
-    x.fetch_add(1);
-    x.fetch_sub(1);
+    std::atomic_int x;
+    ++x;
+    --x;
     return 0;
   }
 " HAVE_STD_ATOMIC)
diff --git a/taglib/toolkit/trefcounter.cpp b/taglib/toolkit/trefcounter.cpp
index eb2aa69f..6638fcaa 100644
--- a/taglib/toolkit/trefcounter.cpp
+++ b/taglib/toolkit/trefcounter.cpp
@@ -31,9 +31,9 @@
 
 #if defined(HAVE_STD_ATOMIC)
 # include <atomic>
-# define ATOMIC_INT std::atomic<unsigned int>
-# define ATOMIC_INC(x) x.fetch_add(1)
-# define ATOMIC_DEC(x) (x.fetch_sub(1) - 1)
+# define ATOMIC_INT std::atomic_int
+# define ATOMIC_INC(x) (++x)
+# define ATOMIC_DEC(x) (--x)
 #elif defined(HAVE_GCC_ATOMIC)
 # define ATOMIC_INT int
 # define ATOMIC_INC(x) __sync_add_and_fetch(&x, 1)