Release 1.2.8 (split out libCStreamedXML)

sent to the wrong buffer, and ended up in content rather than entity
name.

README

@@ -1,4 +1,4 @@
-libStreamedXML
+libStreamedXML        http://www.lwithers.me.uk/projects/libStreamedXML/
 ========================================================================
 (c)2006, Laurence Withers, <l@lwithers.me.uk>.
 Released under the GNU GPLv2. See file COPYING or 
@@ -11,4 +11,8 @@ To build: ./make.sh
 To install: ./make.sh install
     (you might want to set PREFIX, by default it's /usr/local)
 
-@TODO@
+Dependencies
+------------
+
+libutf8++
+  http://www.lwithers.me.uk/projects/libutf8++/

make.sh

@@ -64,6 +64,7 @@ OUTPUT_DIRS="obj html"
 #  none                   no special processing happens before each file
 #  C                      #line directives are inserted before each file
 #                         and VERSION, VERMAJOR etc. are #defined
+#  Ch                     Like C, but for header files (no VERSION #defines)
 #
 make_monolithic() {
     if [ $# -ne 2 ]
@@ -82,6 +83,9 @@ make_monolithic() {
     then
         HASHLINE=1
         VERDEFINE=1
+    elif [ "$2" == "Ch" ]
+    then
+        HASHLINE=1
     elif [ "$2" == "none" ]
     then
             HASHLINE=0 # dummy command
@@ -261,7 +265,7 @@ else
     targets="$@"
 fi
 
-for func in "${targets}"
+for func in ${targets}
 do
     case ${func} in
     clean)

scripts/.gitignore

@@ -1,5 +1,6 @@
 build.c++.app
 build.c++.lib
+build.c++.qtapp
 build.c++.tests
 build.c.app
 build.c.lib

src/libStreamedXML/BottomHeader.h

@@ -5,3 +5,7 @@
 */
 
 #endif
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/Callback.h

@@ -177,8 +177,9 @@ public:
     /*! \brief Entity reference.
 
     \param name The name of the entity being referred to.
-    \returns The expanded text.
-    \throws UnknownEntity if the entity is unknown.
+    \param[out] value The expanded text.
+    \retval true if the entity is valid.
+    \retval false if the entity is not valid.
 
     This function is called whenever an entity reference is encountered,
     and \a expandEntities is \a true in your parser. The five standard
@@ -186,12 +187,18 @@ public:
     but anything else will be passed to this function.
 
     */
-    virtual std::wstring entityRef(const std::wstring& name)
+    virtual bool entityRef(const std::wstring& name, std::wstring& value)
     {
-        throw UnknownEntity(name);
+        (void)name;
+        (void)value;
+        return false;
     }
 };
 
 
 
 };
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/Decoder.cpp

@@ -62,3 +62,7 @@ void Decoder::restart()
 
 
 }
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/Decoder.h

@@ -70,3 +70,7 @@ private:
 
 
 }
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/Exceptions.cpp

@@ -8,14 +8,14 @@ namespace lsx {
 
 
 
-Exception::Exception(const std::wstring& reason)
-    : reason(reason)
+Exception::Exception(const std::wstring& reason, int line, int col)
+    : reason(reason), line(line), col(col)
 {
 }
 
 
 
-const char* Exception::what()
+const char* Exception::what() const throw()
 {
     if(utf8Reason.empty()) utf8Reason = utf8::encode(reason, true);
     return utf8Reason.c_str();
@@ -23,39 +23,46 @@ const char* Exception::what()
 
 
 
-NotWellFormed::NotWellFormed(const std::wstring& error)
-    : Exception(format(error)), error(error)
+NotWellFormed::NotWellFormed(const std::wstring& error, int line, int col)
+    : Exception(format(error, line, col), line, col), error(error)
 {
 }
 
 
 
-std::wstring NotWellFormed::format(const std::wstring& error)
+std::wstring NotWellFormed::format(const std::wstring& error, int line, int col)
 {
     std::wostringstream ost;
-    ost << L"Streamed XML is not well formed:\n"
-        << error;
+    ost << L"Streamed XML is not well formed:"
+        << L"\n  line  : " << line + 1
+        << L"\n  column: " << col + 1
+        << L"\n  reason: " << error;
     return ost.str();
 }
 
 
 
-UnknownEntity::UnknownEntity(const std::wstring& name)
-    : Exception(format(name)), name(name)
+UnknownEntity::UnknownEntity(const std::wstring& name, int line, int col)
+    : Exception(format(name, line, col), line, col), name(name)
 {
 }
 
 
 
-std::wstring UnknownEntity::format(const std::wstring& name)
+std::wstring UnknownEntity::format(const std::wstring& name, int line, int col)
 {
     std::wostringstream ost;
-    ost << L"Encountered an unknown entity named '"
-        << name
-        << L"' in the Streamed XML.";
+    ost << L"Encountered an unknown entity in the Streamed XML:"
+        << L"\n  line  : " << line + 1
+        << L"\n  column: " << col + 1
+        << L"\n  name  : " << name;
     return ost.str();
 }
 
 
 
 }
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/Exceptions.h

@@ -14,30 +14,38 @@ This class is the base class of all exceptions. It simply provides a mechanism f
 message; more specific details must be stored in base classes.
 
 */
-class Exception {
+class Exception : public std::exception {
 public:
     /*! \brief Constructor.
 
     \param reason Reason for the exception.
+    \param line Line number of input (starting from 0).
+    \param col Column of input (starting from 0).
 
     The constructor simply stores the reason for the exception, which may be later accessed for
     reporting to the user.
 
     */
-    Exception(const std::wstring& reason);
+    Exception(const std::wstring& reason, int line, int col);
                
     /// Destructor.
     virtual ~Exception() throw()
     { }
     
     /// Find what caused the error.
-    virtual const char* what();
+    virtual const char* what() const throw();
 
     /// Reason for the exception.
     const std::wstring reason;
+
+    /// Line on which the exception occurred (starting from 0).
+    const int line;
+
+    /// Column on which the exception occurred (starting from 0).
+    const int col;
     
 private:
-    std::string utf8Reason;
+    mutable std::string utf8Reason;
 };
 
 
@@ -45,18 +53,20 @@ private:
 /// Thrown when XML is not well formed.
 class NotWellFormed : public Exception {
 private:
-    static std::wstring format(const std::wstring& error);
+    static std::wstring format(const std::wstring& error, int line, int col);
 
 public:
     /*! \brief Constructor.
 
     \param error Reason for the error.
+    \param line Line on which the exception occurred.
+    \param col Column on which the exception occurred.
 
     The constructor will store the reason for the error. It will also prepare a suitable message
     for the Exception base class.
 
     */
-    NotWellFormed(const std::wstring& error);
+    NotWellFormed(const std::wstring& error, int line, int col);
     
     /// Destructor.
     virtual ~NotWellFormed() throw()
@@ -71,18 +81,20 @@ public:
 /// Thrown when an unknown entity is referred to.
 class UnknownEntity : public Exception {
 private:
-    static std::wstring format(const std::wstring& name);
+    static std::wstring format(const std::wstring& name, int line, int col);
 
 public:
     /*! \brief Constructor.
 
     \param name Name of the unknown entity.
-
+    \param line Line on which the exception occurred.
+    \param col Column on which the exception occurred.
+    
     The constructor will store the name of the unknown entity. It will also prepare a suitable
     message for the Exception base class.
 
     */
-    UnknownEntity(const std::wstring& name);
+    UnknownEntity(const std::wstring& name, int line, int col);
     
     /// Destructor.
     virtual ~UnknownEntity() throw()
@@ -95,3 +107,7 @@ public:
 
 
 }
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/ForwardDeclare.h

@@ -22,3 +22,7 @@ class UnknownEntity;
 
 
 }
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/Parser.cpp

@@ -37,6 +37,8 @@ void Parser::reset()
     skipNextNewline = false;
     state = StateNone;
     restartCount = 0;
+    line = 0;
+    col = 0;
 }
 
 
@@ -59,7 +61,7 @@ void Parser::feedChar(wchar_t ch)
 
 #define ERROR(_reason) do { \
     state = StateError; \
-    throw NotWellFormed(_reason); \
+    throw NotWellFormed(_reason, line, col); \
 }while(0)
     if(ch == xmlRestartMarker[restartCount]) {
         if(++restartCount == 11) {
@@ -84,11 +86,15 @@ void Parser::feedChar(wchar_t ch)
             skipNextNewline = true;
             ch = L'\n';
             c = ClassWhitespace;
+            ++line;
+            col = 0;
             goto doBuffer;
 
         case 0x2028:
             ch = L'\n';
             c = ClassWhitespace;
+            ++line;
+            col = 0;
             break;
 
         case L' ':
@@ -101,6 +107,8 @@ void Parser::feedChar(wchar_t ch)
             if(skipNextNewline) return;
             ch = L'\n';
             c = ClassWhitespace;
+            ++line;
+            col = 0;
             break;
 
         case L':':
@@ -155,26 +163,26 @@ doBuffer:
                     break;
 
                 case ClassOpenTag:
-                    if(!buffer.empty()) callback->whiteSpace(buffer);
                     state = StateOpen;
+                    if(!buffer.empty()) callback->whiteSpace(buffer);
                     buffer.clear();
                     break;
 
                 case ClassEntity:
                     if(expandEntities) {
                         if(!elementDepth) ERROR(L"Entities cannot appear at stream level.");
+                        state = StateEntity;
                         if(!buffer.empty()) callback->whiteSpace(buffer);
                         buffer.clear();
                         parsingAttr = false;
-                        state = StateEntity;
                         break;
                     }
                     
                     // fall through
                 default:
                     if(!elementDepth) ERROR(L"Content cannot appear at stream level.");
-                    if(!buffer.empty()) callback->whiteSpace(buffer);
                     state = StateData;
+                    if(!buffer.empty()) callback->whiteSpace(buffer);
                     buffer = ch;
                     break;
             }
@@ -183,16 +191,16 @@ doBuffer:
         case StateData:
             switch(c) {
                 case ClassOpenTag:
+                    state = StateOpen;
                     callback->content(buffer);
                     buffer.clear();
-                    state = StateOpen;
                     break;
 
                 case ClassEntity:
+                    state = StateEntity;
                     callback->content(buffer);
                     buffer.clear();
                     parsingAttr = false;
-                    state = StateEntity;
                     break;
 
                 default:
@@ -217,9 +225,9 @@ doBuffer:
 
         case StateCDATA2:
             if(ch == L'>') {
+                state = StateNone;
                 callback->cdata(buffer);
                 buffer.clear();
-                state = StateNone;
             } else if(ch == L']') {
                 buffer += ch;
             } else {
@@ -290,9 +298,9 @@ doBuffer:
         case StatePI2:
             if(ch != L'>') ERROR(L"Invalid target for PI");
             else {
+                state = StateNone;
                 callback->PI(buffer2, L"");
                 buffer.clear();
-                state = StateNone;
             }
             break;
             
@@ -303,12 +311,15 @@ doBuffer:
             
         case StatePI3:
             if(ch == L'>') {
+                state = StateNone;
                 callback->PI(buffer2, buffer);
                 buffer.clear();
-                state = StateNone;
+            } else if(ch == '?') {
+                buffer += L'?';
             } else {
                 buffer += L'?';
                 buffer += ch;
+                state = StatePIData;
             }
             break;
 
@@ -343,14 +354,15 @@ doBuffer:
             else {
                 buffer += L'-';
                 buffer += ch;
+                state = StateComment;
             }
             break;
 
         case StateComment3:
             if(ch != L'>') ERROR(L"`--' not valid in comments");
+            state = StateNone;
             callback->comment(buffer);
             buffer.clear();
-            state = StateNone;
             break;
 
         case StateElemName:
@@ -370,10 +382,10 @@ doBuffer:
                     switch(ch) {
                         case L'>':
                             elemStack.push_back(buffer);
-                            callback->element(buffer, elemAttrs);
                             state = StateNone;
-                            ++elementDepth;
                             buffer.clear();
+                            callback->element(buffer, elemAttrs);
+                            ++elementDepth;
                             break;
                         
                         case L'/':
@@ -400,9 +412,9 @@ doBuffer:
                     switch(ch) {
                         case L'>':
                             elemStack.push_back(elemName);
-                            callback->element(elemName, elemAttrs);
-                            buffer.clear();
                             state = StateNone;
+                            buffer.clear();
+                            callback->element(elemName, elemAttrs);
                             ++elementDepth;
                             break;
                         
@@ -462,10 +474,10 @@ doBuffer:
                     if(ch == L'/') {
                         state = StateNeedClose;
                     } else if(ch == L'>') {
+                        state = StateNone;
+                        buffer.clear();
                         callback->element(elemName, elemAttrs);
                         elemStack.push_back(elemName);
-                        buffer.clear();
-                        state = StateNone;
                         ++elementDepth;
                     } else ERROR(L"Invalid character after attribute.");
                     break;
@@ -474,10 +486,10 @@ doBuffer:
 
         case StateNeedClose:
             if(ch != L'>') ERROR(L"Stray `/' in open tag.");
+            state = StateNone;
+            buffer.clear();
             callback->element(elemName, elemAttrs);
             callback->closeTag(elemName);
-            buffer.clear();
-            state = StateNone;
             break;
 
         case StateClose:
@@ -501,9 +513,9 @@ doBuffer:
                     if(ch != L'>') ERROR(L"Invalid character in close tag name.");
                     if(elemStack.back() != buffer) ERROR(L"Mismatched close tag.");
                     elemStack.pop_back();
-                    callback->closeTag(buffer);
-                    buffer.clear();
                     state = StateNone;
+                    buffer.clear();
+                    callback->closeTag(buffer);
                     --elementDepth;
             }
             break;
@@ -513,9 +525,9 @@ doBuffer:
             if(ch != L'>') ERROR(L"Invalid data in close tag.");
             if(elemStack.back() != elemName) ERROR(L"Mismatched close tag.");
             elemStack.pop_back();
-            callback->closeTag(elemName);
-            buffer.clear();
             state = StateNone;
+            buffer.clear();
+            callback->closeTag(elemName);
             --elementDepth;
             break;
 
@@ -566,6 +578,7 @@ doBuffer:
             break;
     }
 #undef ERROR
+    ++col;
 }
 
 
@@ -577,9 +590,16 @@ std::wstring Parser::entityRef(const std::wstring& ent)
     if(ent == L"apos") return L"'";
     if(ent == L"lt") return L"<";
     if(ent == L"gt") return L">";
-    return callback->entityRef(ent);
+
+    std::wstring result;
+    if(callback->entityRef(ent, result)) return result;
+    throw UnknownEntity(ent, line, col);
 }
 
 
 
 }
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/Parser.h

@@ -120,6 +120,9 @@ private:
     bool skipNextNewline;
     int elementDepth, xmlCount, restartCount;
 
+    // for reporting errors
+    int line, col;
+
 public:
     /*! \brief Constructor.
 
@@ -221,3 +224,7 @@ public:
 
 
 }
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/TopHeader.h

@@ -14,3 +14,7 @@
 #include <sstream>
 #include <utf8>
 #include <stdint.h>
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/TopSource.cpp

@@ -8,3 +8,7 @@
 
 // Below are all the includes used throughout the library.
 
+/* options for text editors
+kate: replace-trailing-space-save true; space-indent true; tab-width 4;
+vim: expandtab:ts=4:sw=4
+*/

src/libStreamedXML/build.lib

@@ -31,7 +31,7 @@ then
         echo " Compiling"
 
         SONAME="${libStreamedXML_BASE}.so.${SOMAJOR}.${SOMINOR}"
-        do_cmd ${CXX} ${CFLAGS} -shared -fpic -o "${libStreamedXML}" \
+        do_cmd ${CXX} ${CFLAGS} -Iobj -shared -fpic -o "${libStreamedXML}" \
             -Wl,-soname,${SONAME} \
             ${SRC} ${SO_EXTRA} || return 1
 

src/libStreamedXML/build.monolithic

@@ -9,7 +9,7 @@ MONOLITHIC_TESTS="src/libStreamedXML/build.lib src/libStreamedXML/build.monolith
 if [ -z "${libStreamedXML_MONOLITHIC}" ]
 then
     MONOLITHIC_SOURCE="$(echo src/libStreamedXML/{TopHeader,ForwardDeclare,Exceptions,Callback,Parser,Decoder,BottomHeader}.h)"
-    make_monolithic ${HDR} C || return 1
+    make_monolithic ${HDR} Ch || return 1
 
     MONOLITHIC_SOURCE="$(echo src/libStreamedXML/{TopSource,Exceptions,Parser,Decoder}.cpp)"
     make_monolithic ${SRC} C || return 1

src/libStreamedXML/soversion

@@ -11,7 +11,7 @@
 # be bumped on a binary-incompatible release. They are both single
 # integers.
 SOMAJOR=0
-SOMINOR=0
+SOMINOR=1
 
 # SOMICRO is bumped every time there is a binary-compatible release.
-SOMICRO=0
+SOMICRO=3

version

@@ -11,8 +11,8 @@
 # expected to be in 'major.minor.micro' format. It can optionally be 
 # suffixed with a string.
 VERMAJOR=1
-VERMINOR=1
-VERMICRO=0
+VERMINOR=2
+VERMICRO=8
 VEREXTRA=""
 
 # kate: replace-trailing-space-save true; space-indent true; tab-width 4;