This causes TagLib to check to make sure that text values to-be-written as

ISO-8859-1 are in fact ISO-8859-1 values (assuming they were passed into TagLib
properly) and if not automatically switches those frames to UTF8.

FEATURE:90635
CCBUG:90635


git-svn-id: svn://anonsvn.kde.org/home/kde/trunk/kdesupport/taglib@768857 283d02a7-25f6-0310-bc7c-ecb5cbfe19da

taglib/mpeg/id3v2/frames/attachedpictureframe.cpp

@@ -23,10 +23,11 @@
  *   http://www.mozilla.org/MPL/                                           *
  ***************************************************************************/
 
-#include <tdebug.h>
-
 #include "attachedpictureframe.h"
 
+#include <tstringlist.h>
+#include <tdebug.h>
+
 using namespace TagLib;
 using namespace ID3v2;
 
@@ -145,12 +146,14 @@ ByteVector AttachedPictureFrame::renderFields() const
 {
   ByteVector data;
 
-  data.append(char(d->textEncoding));
+  String::Type encoding = checkEncoding(d->description, d->textEncoding);
+
+  data.append(char(encoding));
   data.append(d->mimeType.data(String::Latin1));
   data.append(textDelimiter(String::Latin1));
   data.append(char(d->type));
-  data.append(d->description.data(d->textEncoding));
-  data.append(textDelimiter(d->textEncoding));
+  data.append(d->description.data(encoding));
+  data.append(textDelimiter(encoding));
   data.append(d->data);
 
   return data;

taglib/mpeg/id3v2/frames/commentsframe.cpp

@@ -26,6 +26,7 @@
 #include <tbytevectorlist.h>
 #include <id3v2tag.h>
 #include <tdebug.h>
+#include <tstringlist.h>
 
 #include "commentsframe.h"
 
@@ -98,7 +99,6 @@ void CommentsFrame::setText(const String &s)
   d->text = s;
 }
 
-
 String::Type CommentsFrame::textEncoding() const
 {
   return d->textEncoding;
@@ -153,11 +153,16 @@ ByteVector CommentsFrame::renderFields() const
 {
   ByteVector v;
 
-  v.append(char(d->textEncoding));
+  String::Type encoding = d->textEncoding;
+
+  encoding = checkEncoding(d->description, encoding);
+  encoding = checkEncoding(d->text, encoding);
+
+  v.append(char(encoding));
   v.append(d->language.size() == 3 ? d->language : "XXX");
-  v.append(d->description.data(d->textEncoding));
-  v.append(textDelimiter(d->textEncoding));
-  v.append(d->text.data(d->textEncoding));
+  v.append(d->description.data(encoding));
+  v.append(textDelimiter(encoding));
+  v.append(d->text.data(encoding));
 
   return v;
 }

taglib/mpeg/id3v2/frames/textidentificationframe.cpp

@@ -139,9 +139,11 @@ void TextIdentificationFrame::parseFields(const ByteVector &data)
 
 ByteVector TextIdentificationFrame::renderFields() const
 {
+  String::Type encoding = checkEncoding(d->fieldList, d->textEncoding);
+
   ByteVector v;
 
-  v.append(char(d->textEncoding));
+  v.append(char(encoding));
 
   for(StringList::ConstIterator it = d->fieldList.begin(); it != d->fieldList.end(); it++) {
 
@@ -150,9 +152,9 @@ ByteVector TextIdentificationFrame::renderFields() const
     // encoding.
 
     if(it != d->fieldList.begin())
-      v.append(textDelimiter(d->textEncoding));
+      v.append(textDelimiter(encoding));
 
-    v.append((*it).data(d->textEncoding));
+    v.append((*it).data(encoding));
   }
 
   return v;

taglib/mpeg/id3v2/frames/textidentificationframe.h

@@ -95,6 +95,10 @@ namespace TagLib {
      * The ID3v2 Frames document gives a description of each of these formats
      * and the expected order of strings in each.  ID3v2::Header::frameID() can
      * be used to determine the frame type.
+     *
+     * \note If non-Latin1 compatible strings are used with this class, even if
+     * the text encoding is set to Latin1, the frame will be written using UTF8
+     * (with the encoding flag appropriately set in the output).
      */
 
     class TAGLIB_EXPORT TextIdentificationFrame : public Frame
@@ -108,6 +112,8 @@ namespace TagLib {
        *
        * \note In this case you must specify the text encoding as it
        * resolves the ambiguity between constructors.
+       *
+       * \note Please see the note in the class description regarding Latin1.
        */
       TextIdentificationFrame(const ByteVector &type, String::Type encoding);
 
@@ -144,6 +150,8 @@ namespace TagLib {
        * This defaults to the type that was either specified in the constructor
        * or read from the frame when parsed.
        *
+       * \note Please see the note in the class description regarding Latin1.
+       *
        * \see setTextEncoding()
        * \see render()
        */
@@ -153,6 +161,8 @@ namespace TagLib {
        * Sets the text encoding to be used when rendering this frame to
        * \a encoding.
        *
+       * \note Please see the note in the class description regarding Latin1.
+       *
        * \see textEncoding()
        * \see render()
        */

taglib/mpeg/id3v2/frames/urllinkframe.cpp

@@ -27,6 +27,7 @@
 
 #include "urllinkframe.h"
 #include <tdebug.h>
+#include <tstringlist.h>
 
 using namespace TagLib;
 using namespace ID3v2;
@@ -174,9 +175,11 @@ ByteVector UserUrlLinkFrame::renderFields() const
 {
   ByteVector v;
 
-  v.append(char(d->textEncoding));
-  v.append(d->description.data(d->textEncoding));
-  v.append(textDelimiter(d->textEncoding));
+  String::Type encoding = checkEncoding(d->description, d->textEncoding);
+
+  v.append(char(encoding));
+  v.append(d->description.data(encoding));
+  v.append(textDelimiter(encoding));
   v.append(url().data(String::Latin1));
 
   return v;

taglib/mpeg/id3v2/id3v2frame.cpp

@@ -32,6 +32,7 @@
 #endif
 
 #include <tdebug.h>
+#include <tstringlist.h>
 
 #include "id3v2frame.h"
 #include "id3v2synchdata.h"
@@ -222,6 +223,20 @@ String Frame::readStringField(const ByteVector &data, String::Type encoding, int
   return str;
 }
 
+String::Type Frame::checkEncoding(const StringList &fields, String::Type encoding) // static
+{
+  if(encoding != String::Latin1)
+    return encoding;
+
+  for(StringList::ConstIterator it = fields.begin(); it != fields.end(); ++it) {
+    if(!(*it).isLatin1()) {
+      debug("Frame::checkEncoding() -- Rendering using UTF8.");
+      return String::UTF8;
+    }
+  }
+
+  return String::Latin1;
+}
 
 ////////////////////////////////////////////////////////////////////////////////
 // Frame::Header class
@@ -523,7 +538,7 @@ ByteVector Frame::Header::render() const
   return v;
 }
 
-bool Frame::Header::frameAlterPreservation() const // deprecated
+bool Frame::Header::frameAlterPreservation() const
 {
   return fileAlterPreservation();
 }

taglib/mpeg/id3v2/id3v2frame.h

@@ -32,6 +32,8 @@
 
 namespace TagLib {
 
+  class StringList;
+
   namespace ID3v2 {
 
     class Tag;
@@ -193,6 +195,13 @@ namespace TagLib {
       String readStringField(const ByteVector &data, String::Type encoding,
                              int *positon = 0);
 
+      /*!
+       * Checks a the list of string values to see if they can be used with the
+       * specified encoding and returns the recommended encoding.
+       */
+      static String::Type checkEncoding(const StringList &fields,
+                                        String::Type encoding);
+
     private:
       Frame(const Frame &);
       Frame &operator=(const Frame &);
@@ -387,7 +396,7 @@ namespace TagLib {
       ByteVector render() const;
 
       /*!
-       * @deprecated
+       * \deprecated
        */
       bool frameAlterPreservation() const;
 

taglib/mpeg/id3v2/id3v2framefactory.cpp

@@ -51,6 +51,12 @@ public:
 
   String::Type defaultEncoding;
   bool useDefaultEncoding;
+
+  template <class T> void setTextEncoding(T *frame)
+  {
+    if(useDefaultEncoding)
+      frame->setTextEncoding(defaultEncoding);
+  }
 };
 
 FrameFactory *FrameFactory::factory = 0;
@@ -142,12 +148,12 @@ Frame *FrameFactory::createFrame(const ByteVector &origData, Header *tagHeader)
   // Text Identification (frames 4.2)
 
   if(frameID.startsWith("T")) {
+
     TextIdentificationFrame *f = frameID != "TXXX"
       ? new TextIdentificationFrame(data, header)
       : new UserTextIdentificationFrame(data, header);
 
-    if(d->useDefaultEncoding)
-      f->setTextEncoding(d->defaultEncoding);
+    d->setTextEncoding(f);
 
     if(frameID == "TCON")
       updateGenre(f);
@@ -159,8 +165,7 @@ Frame *FrameFactory::createFrame(const ByteVector &origData, Header *tagHeader)
 
   if(frameID == "COMM") {
     CommentsFrame *f = new CommentsFrame(data, header);
-    if(d->useDefaultEncoding)
-      f->setTextEncoding(d->defaultEncoding);
+    d->setTextEncoding(f);
     return f;
   }
 
@@ -168,8 +173,7 @@ Frame *FrameFactory::createFrame(const ByteVector &origData, Header *tagHeader)
 
   if(frameID == "APIC") {
     AttachedPictureFrame *f = new AttachedPictureFrame(data, header);
-    if(d->useDefaultEncoding)
-      f->setTextEncoding(d->defaultEncoding);
+    d->setTextEncoding(f);
     return f;
   }
 
@@ -185,8 +189,11 @@ Frame *FrameFactory::createFrame(const ByteVector &origData, Header *tagHeader)
 
   // General Encapsulated Object (frames 4.15)
 
-  if(frameID == "GEOB")
-    return new GeneralEncapsulatedObjectFrame(data, header);
+  if(frameID == "GEOB") {
+    GeneralEncapsulatedObjectFrame *f = new GeneralEncapsulatedObjectFrame(data, header);
+    d->setTextEncoding(f);
+    return f;
+  }
 
   // URL link (frames 4.3)
 
@@ -196,8 +203,7 @@ Frame *FrameFactory::createFrame(const ByteVector &origData, Header *tagHeader)
     }
     else {
       UserUrlLinkFrame *f = new UserUrlLinkFrame(data, header);
-      if(d->useDefaultEncoding)
-        f->setTextEncoding(d->defaultEncoding);
+      d->setTextEncoding(f);
       return f;
     }
   }