Parse headers outside of constructor

In several cases, the parser would parse a header
(say, a track header, or the segment info header)
in the constructor for the type.  The parser had
assumed (incorrectly) that the file was well-formed,
but this turned out to be an incorrect assumption.

The parse errors triggered some assertion failures,
but this is not acceptable in a production system.

The parser solved that problem by separating the
construction of the header object from the parsing
of the file.  There is now a separate parse method
to be called after construction of the object,
which returns a status whether the parse was
successful.

This change fixes the bugs from the following
tracker issues:

http://code.google.com/p/webm/issues/detail?id=399
http://code.google.com/p/webm/issues/detail?id=400

Change-Id: Idb09154ae7008429d8613ce3b3e8294f5a12de86

libwebm_2008.sln

@@ -13,8 +13,6 @@ Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "sample_muxer", "sample_muxe
 		{7B1F12CA-0724-430B-B61A-1D357C912CBA} = {7B1F12CA-0724-430B-B61A-1D357C912CBA}
 	EndProjectSection
 EndProject
-Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "libwebvtt", "libwebvtt.vcproj", "{9545310B-A6E6-4B42-BB00-655D66B6E3D5}"
-EndProject
 Global
 	GlobalSection(SolutionConfigurationPlatforms) = preSolution
 		Debug|Win32 = Debug|Win32
@@ -33,10 +31,6 @@ Global
 		{B407561F-1F5E-4798-B9C2-81AB09CFBC16}.Debug|Win32.Build.0 = Debug|Win32
 		{B407561F-1F5E-4798-B9C2-81AB09CFBC16}.Release|Win32.ActiveCfg = Release|Win32
 		{B407561F-1F5E-4798-B9C2-81AB09CFBC16}.Release|Win32.Build.0 = Release|Win32
-		{9545310B-A6E6-4B42-BB00-655D66B6E3D5}.Debug|Win32.ActiveCfg = Debug|Win32
-		{9545310B-A6E6-4B42-BB00-655D66B6E3D5}.Debug|Win32.Build.0 = Debug|Win32
-		{9545310B-A6E6-4B42-BB00-655D66B6E3D5}.Release|Win32.ActiveCfg = Release|Win32
-		{9545310B-A6E6-4B42-BB00-655D66B6E3D5}.Release|Win32.Build.0 = Release|Win32
 	EndGlobalSection
 	GlobalSection(SolutionProperties) = preSolution
 		HideSolutionNode = FALSE

libwebm_2008.vcproj

@@ -48,6 +48,7 @@
 				UsePrecompiledHeader="0"
 				WarningLevel="4"
 				DebugInformationFormat="1"
+				DisableSpecificWarnings="4996"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"

libwebvtt.vcproj

@@ -1,157 +0,0 @@
-<?xml version="1.0" encoding="Windows-1252"?>
-<VisualStudioProject
-	ProjectType="Visual C++"
-	Version="9.00"
-	Name="libwebvtt_2008"
-	ProjectGUID="{9545310B-A6E6-4B42-BB00-655D66B6E3D5}"
-	RootNamespace="libwebvtt"
-	Keyword="Win32Proj"
-	TargetFrameworkVersion="196613"
-	>
-	<Platforms>
-		<Platform
-			Name="Win32"
-		/>
-	</Platforms>
-	<ToolFiles>
-	</ToolFiles>
-	<Configurations>
-		<Configuration
-			Name="Debug|Win32"
-			OutputDirectory="$(SolutionDir)..\lib\$(SolutionName)\$(ProjectName)\$(ConfigurationName)"
-			IntermediateDirectory="$(SolutionDir)..\obj\$(SolutionName)\$(ProjectName)\$(ConfigurationName)"
-			ConfigurationType="4"
-			CharacterSet="1"
-			>
-			<Tool
-				Name="VCPreBuildEventTool"
-			/>
-			<Tool
-				Name="VCCustomBuildTool"
-			/>
-			<Tool
-				Name="VCXMLDataGeneratorTool"
-			/>
-			<Tool
-				Name="VCWebServiceProxyGeneratorTool"
-			/>
-			<Tool
-				Name="VCMIDLTool"
-			/>
-			<Tool
-				Name="VCCLCompilerTool"
-				Optimization="0"
-				PreprocessorDefinitions="WIN32;_DEBUG;_LIB"
-				MinimalRebuild="false"
-				BasicRuntimeChecks="3"
-				RuntimeLibrary="1"
-				UsePrecompiledHeader="0"
-				WarningLevel="4"
-				DebugInformationFormat="1"
-			/>
-			<Tool
-				Name="VCManagedResourceCompilerTool"
-			/>
-			<Tool
-				Name="VCResourceCompilerTool"
-			/>
-			<Tool
-				Name="VCPreLinkEventTool"
-			/>
-			<Tool
-				Name="VCLibrarianTool"
-			/>
-			<Tool
-				Name="VCALinkTool"
-			/>
-			<Tool
-				Name="VCXDCMakeTool"
-			/>
-			<Tool
-				Name="VCBscMakeTool"
-			/>
-			<Tool
-				Name="VCFxCopTool"
-			/>
-			<Tool
-				Name="VCPostBuildEventTool"
-			/>
-		</Configuration>
-		<Configuration
-			Name="Release|Win32"
-			OutputDirectory="$(SolutionDir)$(ConfigurationName)"
-			IntermediateDirectory="$(ConfigurationName)"
-			ConfigurationType="4"
-			CharacterSet="1"
-			WholeProgramOptimization="1"
-			>
-			<Tool
-				Name="VCPreBuildEventTool"
-			/>
-			<Tool
-				Name="VCCustomBuildTool"
-			/>
-			<Tool
-				Name="VCXMLDataGeneratorTool"
-			/>
-			<Tool
-				Name="VCWebServiceProxyGeneratorTool"
-			/>
-			<Tool
-				Name="VCMIDLTool"
-			/>
-			<Tool
-				Name="VCCLCompilerTool"
-				Optimization="2"
-				EnableIntrinsicFunctions="true"
-				PreprocessorDefinitions="WIN32;NDEBUG;_LIB"
-				RuntimeLibrary="2"
-				EnableFunctionLevelLinking="true"
-				UsePrecompiledHeader="0"
-				WarningLevel="3"
-				DebugInformationFormat="3"
-			/>
-			<Tool
-				Name="VCManagedResourceCompilerTool"
-			/>
-			<Tool
-				Name="VCResourceCompilerTool"
-			/>
-			<Tool
-				Name="VCPreLinkEventTool"
-			/>
-			<Tool
-				Name="VCLibrarianTool"
-			/>
-			<Tool
-				Name="VCALinkTool"
-			/>
-			<Tool
-				Name="VCXDCMakeTool"
-			/>
-			<Tool
-				Name="VCBscMakeTool"
-			/>
-			<Tool
-				Name="VCFxCopTool"
-			/>
-			<Tool
-				Name="VCPostBuildEventTool"
-			/>
-		</Configuration>
-	</Configurations>
-	<References>
-	</References>
-	<Files>
-		<File
-			RelativePath=".\webvttparser.cpp"
-			>
-		</File>
-		<File
-			RelativePath=".\webvttparser.hpp"
-			>
-		</File>
-	</Files>
-	<Globals>
-	</Globals>
-</VisualStudioProject>

mkvmuxer.cpp

@@ -830,7 +830,7 @@ bool Tracks::AddTrack(Track* track, int32 number) {
   if (number < 0)
     return false;
 
-  int32 track_num = number;
+  uint32 track_num = number;
 
   // Check to make sure a track does not already have |track_num|.
   for (uint32 i = 0; i < track_entries_size_; ++i) {

mkvparser.hpp

@@ -30,23 +30,26 @@ protected:
 
 long long GetUIntLength(IMkvReader*, long long, long&);
 long long ReadUInt(IMkvReader*, long long, long&);
-long long SyncReadUInt(IMkvReader*, long long pos, long long stop, long&);
 long long UnserializeUInt(IMkvReader*, long long pos, long long size);
-float Unserialize4Float(IMkvReader*, long long);
-double Unserialize8Double(IMkvReader*, long long);
 
-#if 0
+long UnserializeFloat(IMkvReader*, long long pos, long long size, double&);
-short Unserialize2SInt(IMkvReader*, long long);
-signed char Unserialize1SInt(IMkvReader*, long long);
-#else
 long UnserializeInt(IMkvReader*, long long pos, long len, long long& result);
-#endif
+
+long UnserializeString(
+        IMkvReader*,
+        long long pos,
+        long long size,
+        char*& str);
+
+long ParseElementHeader(
+    IMkvReader* pReader,
+    long long& pos,  //consume id and size fields
+    long long stop,  //if you know size of element's parent
+    long long& id,
+    long long& size);
 
 bool Match(IMkvReader*, long long&, unsigned long, long long&);
-bool Match(IMkvReader*, long long&, unsigned long, char*&);
 bool Match(IMkvReader*, long long&, unsigned long, unsigned char*&, size_t&);
-bool Match(IMkvReader*, long long&, unsigned long, double&);
-bool Match(IMkvReader*, long long&, unsigned long, short&);
 
 void GetVersion(int& major, int& minor, int& build, int& revision);
 
@@ -289,13 +292,15 @@ class Track
     Track& operator=(const Track&);
 
 public:
+    enum Type { kVideo = 1, kAudio = 2 };
+
     Segment* const m_pSegment;
     const long long m_element_start;
     const long long m_element_size;
     virtual ~Track();
 
-    long long GetType() const;
+    long GetType() const;
-    long long GetNumber() const;
+    long GetNumber() const;
     unsigned long long GetUid() const;
     const char* GetNameAsUTF8() const;
     const char* GetCodecNameAsUTF8() const;
@@ -311,21 +316,29 @@ public:
         long long size;
     };
 
-    struct Info
+    class Info
     {
-        long long type;
+    public:
-        long long number;
+        Info();
+        ~Info();
+        int Copy(Info&) const;
+        void Clear();
+    private:
+        Info(const Info&);
+        Info& operator=(const Info&);
+    public:
+        long type;
+        long number;
         unsigned long long uid;
         char* nameAsUTF8;
         char* codecId;
+        char* codecNameAsUTF8;
         unsigned char* codecPrivate;
         size_t codecPrivateSize;
-        char* codecNameAsUTF8;
         bool lacing;
         Settings settings;
-
+    private:
-        Info();
+        int CopyStr(char* Info::*str, Info&) const;
-        void Clear();
     };
 
     long GetFirst(const BlockEntry*&) const;
@@ -341,10 +354,10 @@ public:
 protected:
     Track(
         Segment*,
-        const Info&,
         long long element_start,
         long long element_size);
-    const Info m_info;
+
+    Info m_info;
 
     class EOSBlock : public BlockEntry
     {
@@ -368,12 +381,19 @@ class VideoTrack : public Track
     VideoTrack(const VideoTrack&);
     VideoTrack& operator=(const VideoTrack&);
 
-public:
     VideoTrack(
+        Segment*,
+        long long element_start,
+        long long element_size);
+
+public:
+    static long Parse(
         Segment*,
         const Info&,
         long long element_start,
-        long long element_size);
+        long long element_size,
+        VideoTrack*&);
+
     long long GetWidth() const;
     long long GetHeight() const;
     double GetFrameRate() const;
@@ -394,12 +414,18 @@ class AudioTrack : public Track
     AudioTrack(const AudioTrack&);
     AudioTrack& operator=(const AudioTrack&);
 
-public:
     AudioTrack(
+        Segment*,
+        long long element_start,
+        long long element_size);
+public:
+    static long Parse(
         Segment*,
         const Info&,
         long long element_start,
-        long long element_size);
+        long long element_size,
+        AudioTrack*&);
+
     double GetSamplingRate() const;
     long long GetChannels() const;
     long long GetBitDepth() const;
@@ -431,24 +457,27 @@ public:
         long long size,
         long long element_start,
         long long element_size);
-    virtual ~Tracks();
 
-    const Track* GetTrackByNumber(unsigned long tn) const;
+    ~Tracks();
+
+    long Parse();
+
+    unsigned long GetTracksCount() const;
+
+    const Track* GetTrackByNumber(long tn) const;
     const Track* GetTrackByIndex(unsigned long idx) const;
 
 private:
     Track** m_trackEntries;
     Track** m_trackEntriesEnd;
 
-    void ParseTrackEntry(
+    long ParseTrackEntry(
-        long long,
+        long long payload_start,
-        long long,
+        long long payload_size,
-        Track*&,
         long long element_start,
-        long long element_size);
+        long long element_size,
+        Track*&) const;
 
-public:
-    unsigned long GetTracksCount() const;
 };
 
 
@@ -473,6 +502,8 @@ public:
 
     ~SegmentInfo();
 
+    long Parse();
+
     long long GetTimeCodeScale() const;
     long long GetDuration() const;  //scaled
     const char* GetMuxingAppAsUTF8() const;
@@ -509,6 +540,8 @@ public:
 
     ~SeekHead();
 
+    long Parse();
+
     struct Entry
     {
         //the SeekHead entry payload

sample.cpp

@@ -11,6 +11,7 @@
 
 #include "mkvreader.hpp"
 #include "mkvparser.hpp"
+#include <memory>
 
 #ifdef _MSC_VER
 // Silences these warnings:
@@ -83,15 +84,18 @@ int main(int argc, char* argv[])
     printf("\t\tDoc Type\t\t: %s\n", ebmlHeader.m_docType);
     printf("\t\tPos\t\t\t: %lld\n", pos);
 
-    mkvparser::Segment* pSegment;
+    typedef mkvparser::Segment seg_t;
+    seg_t* pSegment_;
 
-    long long ret = mkvparser::Segment::CreateInstance(&reader, pos, pSegment);
+    long long ret = seg_t::CreateInstance(&reader, pos, pSegment_);
     if (ret)
     {
         printf("\n Segment::CreateInstance() failed.");
         return -1;
     }
 
+    const std::auto_ptr<seg_t> pSegment(pSegment_);
+
     ret  = pSegment->Load();
     if (ret < 0)
     {
@@ -157,8 +161,6 @@ int main(int argc, char* argv[])
     unsigned long i = 0;
     const unsigned long j = pTracks->GetTracksCount();
 
-    enum { VIDEO_TRACK = 1, AUDIO_TRACK = 2 };
-
     printf("\n\t\t\t   Track Info\n");
 
     while (i != j)
@@ -168,13 +170,13 @@ int main(int argc, char* argv[])
         if (pTrack == NULL)
             continue;
 
-        const long long trackType = pTrack->GetType();
+        const long trackType = pTrack->GetType();
-        const long long trackNumber = pTrack->GetNumber();
+        const long trackNumber = pTrack->GetNumber();
         const unsigned long long trackUid = pTrack->GetUid();
         const wchar_t* const pTrackName = utf8towcs(pTrack->GetNameAsUTF8());
 
-        printf("\t\tTrack Type\t\t: %lld\n", trackType);
+        printf("\t\tTrack Type\t\t: %ld\n", trackType);
-        printf("\t\tTrack Number\t\t: %lld\n", trackNumber);
+        printf("\t\tTrack Number\t\t: %ld\n", trackNumber);
         printf("\t\tTrack Uid\t\t: %lld\n", trackUid);
 
         if (pTrackName == NULL)
@@ -203,7 +205,7 @@ int main(int argc, char* argv[])
             delete [] pCodecName;
         }
 
-        if (trackType == VIDEO_TRACK)
+        if (trackType == mkvparser::Track::kVideo)
         {
             const VideoTrack* const pVideoTrack =
                 static_cast<const VideoTrack*>(pTrack);
@@ -218,7 +220,7 @@ int main(int argc, char* argv[])
             printf("\t\tVideo Rate\t\t: %f\n",rate);
         }
 
-        if (trackType == AUDIO_TRACK)
+        if (trackType == mkvparser::Track::kAudio)
         {
             const AudioTrack* const pAudioTrack =
                 static_cast<const AudioTrack*>(pTrack);
@@ -242,7 +244,6 @@ int main(int argc, char* argv[])
     if (clusterCount == 0)
     {
         printf("\t\tSegment has no clusters.\n");
-        delete pSegment;
         return -1;
     }
 
@@ -264,7 +265,7 @@ int main(int argc, char* argv[])
         {
             printf("\t\tError parsing first block of cluster\n");
             fflush(stdout);
-            goto done;
+            return -1;
         }
 
         while ((pBlockEntry != NULL) && !pBlockEntry->EOS())
@@ -273,21 +274,27 @@ int main(int argc, char* argv[])
             const long long trackNum = pBlock->GetTrackNumber();
             const unsigned long tn = static_cast<unsigned long>(trackNum);
             const Track* const pTrack = pTracks->GetTrackByNumber(tn);
-            const long long trackType = pTrack->GetType();
-            const int frameCount = pBlock->GetFrameCount();
-            const long long time_ns = pBlock->GetTime(pCluster);
 
-            printf("\t\t\tBlock\t\t:%s,%s,%15lld\n",
+            if (pTrack == NULL)
-                   (trackType == VIDEO_TRACK) ? "V" : "A",
+                printf("\t\t\tBlock\t\t:UNKNOWN TRACK TYPE\n");
-                   pBlock->IsKey() ? "I" : "P",
+            else
-                   time_ns);
-
-            for (int i = 0; i < frameCount; ++i)
             {
-                const Block::Frame& theFrame = pBlock->GetFrame(i);
+                const long long trackType = pTrack->GetType();
-                const long size = theFrame.len;
+                const int frameCount = pBlock->GetFrameCount();
-                const long long offset = theFrame.pos;
+                const long long time_ns = pBlock->GetTime(pCluster);
-                printf("\t\t\t %15ld,%15llx\n", size, offset);
+
+                printf("\t\t\tBlock\t\t:%s,%s,%15lld\n",
+                       (trackType == mkvparser::Track::kVideo) ? "V" : "A",
+                       pBlock->IsKey() ? "I" : "P",
+                       time_ns);
+
+                for (int i = 0; i < frameCount; ++i)
+                {
+                    const Block::Frame& theFrame = pBlock->GetFrame(i);
+                    const long size = theFrame.len;
+                    const long long offset = theFrame.pos;
+                    printf("\t\t\t %15ld,%15llx\n", size, offset);
+                }
             }
 
             status = pCluster->GetNext(pBlockEntry, pBlockEntry);
@@ -296,15 +303,13 @@ int main(int argc, char* argv[])
             {
                 printf("\t\t\tError parsing next block of cluster\n");
                 fflush(stdout);
-                goto done;
+                return -1;
             }
         }
 
         pCluster = pSegment->GetNext(pCluster);
     }
 
-done:
+    fflush(stdout);
-    delete pSegment;
     return 0;
-
 }

webvttparser.cpp

@@ -1,23 +0,0 @@
-#include "webvttparser.hpp"
-
-namespace WebvttParser
-{
-
-IReader::IReader()
-{
-}
-
-IReader::~IReader()
-{
-}
-
-Parser::Parser(IReader* p) :
-    m_pReader(p)
-{
-}
-
-Parser::~Parser()
-{
-}
-
-}  //end namespace WebvttParser

webvttparser.hpp

@@ -1,60 +0,0 @@
-#ifndef WEBVTTPARSER_HPP
-#define WEBVTTPARSER_HPP
-
-namespace WebvttParser
-{
-
-class IReader
-{
-protected:
-    IReader();
-    virtual ~IReader();
-
-public:
-    virtual int Read(char& c) = 0;
-
-};
-
-class Parser
-{
-private:
-    Parser(const Parser&);
-    Parser& operator=(const Parser&);
-
-public:
-    Parser(IReader*);
-    ~Parser();
-
-    int Parse();
-
-    //need to know kind:
-    // caption, description, metadata, etc
-    // is this a webvtt stream?
-    //
-    //cue identifier
-    //timestamp (or defaults?) + cue settings
-    //payload
-    //  preserve each line
-
-    //machine states:
-    //  parsing stream id ("webvtt")
-    //  parsing file-wide metadata
-    //  parsing cue
-    //
-    //within cue:
-    //  parsing line
-    //    is this a cue identifier or timestamp line?
-    //
-    //  parsing cue id
-    //  parsing timestamp
-    //  parsing lines, looking for end-of-cue
-
-private:
-    IReader* const m_pReader;
-
-};
-
-
-}  //end namespace WebvttParser
-
-#endif