XML is important, I guess...

1 files changed, 165 insertions, 0 deletions

diff --git a/src/xmlreader.cpp b/src/xmlreader.cpp
new file mode 100644
index 0000000..9d299e6
--- /dev/null
+++ b/src/xmlreader.cpp
@@ -0,0 +1,165 @@
+#include "bu/xmlreader.h"
+#include "bu/stream.h"
+#include "bu/exceptions.h"
+
+Bu::XmlReader::XmlReader( Stream &rInput ) :
+        rInput( rInput ),
+        iCurToken( 0 ),
+        iNextToken( 0 ),
+        bIgnoreWS( true )
+{
+        nextToken();
+        stDocument();
+}
+
+Bu::XmlReader::~XmlReader()
+{
+}
+
+void Bu::XmlReader::fillBuffer()
+{
+        if( rInput.isEOS() )
+                return;
+        char buf[1024];
+        int iSize = rInput.read( buf, 1024 );
+        sBuf.append( buf, iSize );
+}
+
+void Bu::XmlReader::cleanupBuffer( int iUsed )
+{
+        for( int j = 0; j < iUsed; j++ )
+        {
+                if( sBuf[j] == '\n' )
+                {
+                        spNextToken.iLine++;
+                        spNextToken.iChar = 1;
+                }
+                else
+                {
+                        spNextToken.iChar++;
+                }
+        }
+
+        printf("--Deleting %d bytes from front of buffer.\n", iUsed );
+        sBuf.trimFront( iUsed );
+}
+
+int Bu::XmlReader::nextToken()
+{
+        fillBuffer();
+
+        int iUsed = 1;
+
+        iCurToken = iNextToken;
+        spCurToken = spNextToken;
+
+        switch( sBuf[0] )
+        {
+                case '<':
+                        if( !strncmp( sBuf.getStr(), "<?xml", 5 ) )
+                        {
+                                iNextToken = tokXmlDeclHead;
+                                iUsed = 5;
+                        }
+                        else
+                        {
+                                iNextToken = '<';
+                        }
+                        break;
+
+                case '?':
+                        if( sBuf[1] == '>' )
+                        {
+                                iNextToken = tokXmlDeclEnd;
+                                iUsed = 2;
+                        }
+                        else
+                        {
+                                iNextToken = '?';
+                        }
+                        break;
+
+                case ' ':
+                case '\t':
+                case '\n':
+                case '\r':
+                        for( int j = 1;; j++ )
+                        {
+                                if( j == sBuf.getSize() )
+                                {
+                                        if( rInput.isEOS() )
+                                                error("Reached end of input while waiting for whitespace to end.");
+
+                                        fillBuffer();
+                                }
+                                if( sBuf[j] == ' ' || sBuf[j] == '\t' ||
+                                        sBuf[j] == '\n' || sBuf[j] == '\r' )
+                                        iUsed++;
+                                else
+                                        break;
+                        }
+                        sStr.clear();
+                        sStr.append( sBuf, iUsed );
+                        iNextToken = tokWS;
+                        break;
+
+                case '=':
+                        iNextToken = sBuf[0];
+                        break;
+
+                default:
+                        if( (sBuf[0] >= 'a' && sBuf[0] <= 'z') ||
+                                (sBuf[0] >= 'A' && sBuf[0] <= 'Z') )
+                        {
+                                for( int j = 1;; j++ )
+                                {
+                                        if( j == sBuf.getSize() )
+                                        {
+                                                if( rInput.isEOS() )
+                                                        error("Reached end of input while waiting for a string to end.");
+
+                                                fillBuffer();
+                                        }
+                                        if( (sBuf[j] >= 'a' && sBuf[j] <= 'z') ||
+                                                (sBuf[j] >= 'A' && sBuf[j] <= 'Z') )
+                                                iUsed++;
+                                        else
+                                                break;
+                                }
+                                sStr.clear();
+                                sStr.append( sBuf, iUsed );
+                                iNextToken = tokIdent;
+                        }
+        }
+
+        cleanupBuffer( iUsed );
+
+        return iCurToken;
+}
+
+void Bu::XmlReader::error( const char *sMessage )
+{
+        throw Bu::XmlException("%d:%d:  %s",
+                spCurToken.iLine, spCurToken.iChar, sMessage );
+}
+
+void Bu::XmlReader::stDocument()
+{
+        stProlog();
+}
+
+void Bu::XmlReader::stProlog()
+{
+        stXmlDecl();
+}
+
+void Bu::XmlReader::stXmlDecl()
+{
+        if( nextToken() != tokXmlDeclHead )
+                error("You must begin your xml file with a declaration: <?xml ... ?>");
+        if( nextToken() != tokIdent )
+                error("A version comes first!");
+        if( sStr != "version" )
+                error("No, a version!");
+}
+