ooxml mathml import - first very basic implementation

still needs a number of cleanups (and handling more of course)

7 files changed, 337 insertions, 1 deletions

diff --git a/starmath/Library_sm.mk b/starmath/Library_sm.mk
index 6b9c7babdca4..9c4a6904450e 100644
--- a/starmath/Library_sm.mk
+++ b/starmath/Library_sm.mk
@@ -80,6 +80,7 @@ $(eval $(call gb_Library_add_exception_objects,sm,\
         starmath/source/mathtype \
         starmath/source/node \
         starmath/source/ooxmlexport \
+        starmath/source/ooxmlimport \
         starmath/source/parse \
         starmath/source/rect \
 		starmath/source/register \
diff --git a/starmath/inc/document.hxx b/starmath/inc/document.hxx
index 99f0d8b4cc79..c1b48629a454 100644
--- a/starmath/inc/document.hxx
+++ b/starmath/inc/document.hxx
@@ -53,6 +53,10 @@ class SfxMenuBarManager;
 class SfxPrinter;
 class Printer;
 class SmCursor;
+namespace ooxmlformulaimport
+{
+class XmlStream;
+}
 
 #define HINT_DATACHANGED    1004
 
@@ -175,6 +179,7 @@ class SmDocShell : public SfxObjectShell, public SfxListener
     void                InvalidateCursor();
 
     bool writeFormulaOoxml( ::sax_fastparser::FSHelperPtr m_pSerializer, oox::core::OoxmlVersion version );
+    bool readFormulaOoxml( ooxmlformulaimport::XmlStream& stream );
 
 public:
     TYPEINFO();
diff --git a/starmath/inc/unomodel.hxx b/starmath/inc/unomodel.hxx
index 66479f988ec5..1324c9f8bc88 100644
--- a/starmath/inc/unomodel.hxx
+++ b/starmath/inc/unomodel.hxx
@@ -38,6 +38,7 @@
 #include <comphelper/propertysethelper.hxx>
 #include <vcl/print.hxx>
 #include <oox/export/ooxmlexport.hxx>
+#include <oox/export/starmathimport.hxx>
 
 class SmFormat;
 
@@ -65,7 +66,8 @@ class SmModel : public SfxBaseModel,
                 public comphelper::PropertySetHelper,
                 public com::sun::star::lang::XServiceInfo,
                 public com::sun::star::view::XRenderable,
-                public OoxmlFormulaExportBase
+                public OoxmlFormulaExportBase,
+                public OoxmlFormulaImportBase
 {
     SmPrintUIOptions* m_pPrintUIOptions;
 protected:
@@ -104,6 +106,8 @@ public:
 
     // OoxmlFormulaExportBase
     virtual void writeFormulaOoxml( ::sax_fastparser::FSHelperPtr m_pSerializer, oox::core::OoxmlVersion version );
+    // OoxmlFormulaImportBase
+    virtual void readFormulaOoxml( ooxmlformulaimport::XmlStream& stream );
 
     static ::com::sun::star::uno::Sequence< rtl::OUString > getSupportedServiceNames_Static();
     static ::rtl::OUString getImplementationName_Static();
diff --git a/starmath/source/document.cxx b/starmath/source/document.cxx
index 7586e64d3089..36df9ae505dc 100644
--- a/starmath/source/document.cxx
+++ b/starmath/source/document.cxx
@@ -96,6 +96,7 @@
 #include <view.hxx>
 #include "mathtype.hxx"
 #include "ooxmlexport.hxx"
+#include "ooxmlimport.hxx"
 #include "mathmlimport.hxx"
 #include "mathmlexport.hxx"
 #include <sfx2/sfxsids.hrc>
@@ -998,6 +999,15 @@ bool SmDocShell::writeFormulaOoxml( ::sax_fastparser::FSHelperPtr m_pSerializer,
     return aEquation.ConvertFromStarMath( m_pSerializer );
 }
 
+bool SmDocShell::readFormulaOoxml( ooxmlformulaimport::XmlStream& stream )
+{
+    RTL_LOGFILE_CONTEXT( aLog, "starmath: SmDocShell::readFormulaOoxml" );
+
+    SmOoxmlImport aEquation( stream );
+    SetText( aEquation.ConvertToStarMath());
+    return true; // TODO just void?
+}
+
 sal_Bool SmDocShell::SaveCompleted( const ::com::sun::star::uno::Reference< ::com::sun::star::embed::XStorage >& xStorage )
 {
     RTL_LOGFILE_CONTEXT( aLog, "starmath: SmDocShell::SaveCompleted" );
diff --git a/starmath/source/ooxmlimport.cxx b/starmath/source/ooxmlimport.cxx
new file mode 100644
index 000000000000..cd7610fd215a
--- /dev/null
+++ b/starmath/source/ooxmlimport.cxx
@@ -0,0 +1,233 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * Version: MPL 1.1 / GPLv3+ / LGPLv3+
+ *
+ * The contents of this file are subject to the Mozilla Public License Version
+ * 1.1 (the "License"); you may not use this file except in compliance with
+ * the License or as specified alternatively below. You may obtain a copy of
+ * the License at http://www.mozilla.org/MPL/
+ *
+ * Software distributed under the License is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
+ * for the specific language governing rights and limitations under the
+ * License.
+ *
+ * Major Contributor(s):
+ * Copyright (C) 2011 Lubos Lunak <l.lunak@suse.cz> (initial developer)
+ *
+ * All Rights Reserved.
+ *
+ * For minor contributions see the git repository.
+ *
+ * Alternatively, the contents of this file may be used under the terms of
+ * either the GNU General Public License Version 3 or later (the "GPLv3+"), or
+ * the GNU Lesser General Public License Version 3 or later (the "LGPLv3+"),
+ * in which case the provisions of the GPLv3+ or the LGPLv3+ are applicable
+ * instead of those above.
+ */
+
+// MARKER(update_precomp.py): autogen include statement, do not remove
+#include "precompiled_starmath.hxx"
+
+#include "ooxmlimport.hxx"
+
+#include <oox/token/tokens.hxx>
+#include <oox/token/namespaces.hxx>
+
+using namespace oox;
+using namespace ooxmlformulaimport;
+using rtl::OUString;
+
+/*
+The primary internal data structure for the formula is the text representation
+(the SmNode tree is built from it), so read data must be converted into this format.
+*/
+
+#define M_TOKEN( token ) OOX_TOKEN( officeMath, token )
+#define OPENING_TAG( token ) OPENING( token )
+#define CLOSING_TAG( token ) CLOSING( token )
+
+// *sigh*
+#define STR( str ) OUString( RTL_CONSTASCII_USTRINGPARAM( str ))
+
+// TODO create IS_OPENING(), IS_CLOSING() instead of doing 'next == OPENING( next )' ?
+
+SmOoxmlImport::SmOoxmlImport( ooxmlformulaimport::XmlStream& s )
+: stream( s )
+{
+}
+
+OUString SmOoxmlImport::ConvertToStarMath()
+{
+    return handleStream();
+}
+
+// "toplevel" of reading, there will be oMath (if there was oMathPara, that was
+// up to the parent component to handle)
+
+// NOT complete
+OUString SmOoxmlImport::handleStream()
+{
+    checkOpeningTag( M_TOKEN( oMath ));
+    OUString ret;
+    bool out = false;
+    while( !out && !stream.nextIsEnd())
+    {
+        switch( stream.peekNextToken())
+        {
+            case OPENING( M_TOKEN( f )):
+                ret += STR( " " ) + handleF();
+                break;
+            case CLOSING( M_TOKEN( oMath )):
+                out = true;
+                break;
+            default:
+                handleUnexpectedTag();
+                break;
+        }
+    }
+    checkClosingTag( M_TOKEN( oMath ));
+    return ret;
+}
+
+// NOT complete
+OUString SmOoxmlImport::handleF()
+{
+    checkOpeningTag( M_TOKEN( f ));
+    if( stream.peekNextToken() == OPENING_TAG( M_TOKEN( fPr )))
+    {
+        // TODO
+    }
+    checkOpeningTag( M_TOKEN( num ));
+    OUString num = readR();
+    checkClosingTag( M_TOKEN( num ));
+    checkOpeningTag( M_TOKEN( den ));
+    OUString den = readR();
+    checkClosingTag( M_TOKEN( den ));
+    checkClosingTag( M_TOKEN( f ));
+    return STR( "{" ) + num + STR( "} over {" ) + den + STR( "}" );
+}
+
+// NOT complete
+OUString SmOoxmlImport::readR()
+{
+    checkOpeningTag( M_TOKEN( r ));
+
+//    checkOpeningTag( OOX_TOKEN( doc, rPr ));
+//    checkOpeningTag( OOX_TOKEN( doc, rFonts ));
+//    checkClosingTag( OOX_TOKEN( doc, rFonts ));
+//    checkClosingTag( OOX_TOKEN( doc, rPr ));
+
+    // TODO can there be more t's ?
+    checkOpeningTag( M_TOKEN( t ));
+    OUString text = stream.getCharacters();
+    if( !stream.getAttributes().getBool( OOX_TOKEN( xml, space ), false ))
+        text = text.trim();
+    checkClosingTag( M_TOKEN( t ));
+    checkClosingTag( M_TOKEN( r ));
+    return text;
+}
+
+void SmOoxmlImport::checkOpeningTag( int token )
+{
+    checkTag( OPENING( token ), "opening" );
+}
+
+void SmOoxmlImport::checkClosingTag( int token )
+{
+    checkTag( CLOSING( token ), "closing" );
+}
+
+void SmOoxmlImport::checkTag( int token, const char* txt )
+{
+    if( stream.peekNextToken() == token )
+    {
+        stream.getNextToken(); // read it
+        return; // ok
+    }
+    if( recoverAndFindTag( token ))
+    {
+        stream.getNextToken(); // read it
+        return; // ok, skipped some tokens
+    }
+    fprintf( stderr, "Expected %s tag %d not found.\n", txt, token );
+}
+
+bool SmOoxmlImport::recoverAndFindTag( int token )
+{
+    int depth = 0;
+    for(;;)
+    {
+        if( depth > 0 ) // we're inside a nested element, skip those
+        {
+            int next = stream.getNextToken();
+            if( next == OPENING( next ))
+            {
+                fprintf( stderr, "Skipping opening tag %d\n", next );
+                ++depth;
+            }
+            else if( next == CLOSING( next ))
+            { // TODO debug output without the OPENING/CLOSING bits set
+                fprintf( stderr, "Skipping closing tag %d\n", next );
+                --depth;
+            }
+            else if( next == XML_TOKEN_INVALID ) // end of stream
+            {
+                fprintf( stderr, "Unexpected end of stream reached.\n" );
+                return false;
+            }
+            else
+            {
+                fprintf( stderr, "Malformed token %d\n", next );
+                abort();
+            }
+            continue;
+        }
+        int next = stream.peekNextToken();
+        if( next == CLOSING( next ))
+            return false; // that would be leaving current element, so not found
+        if( next == token )
+            return true; // ok, found
+        if( next == OPENING( next ))
+        {
+            fprintf( stderr, "Skipping opening tag %d\n", next );
+            stream.getNextToken();
+            ++depth;
+        }
+        else if( next == XML_TOKEN_INVALID )
+        {
+            fprintf( stderr, "Unexpected end of stream reached.\n" );
+            return false;
+        }
+        else
+            abort();
+    }
+}
+
+void SmOoxmlImport::skipElement( int token )
+{
+    int closing = ( token & ~TAG_OPENING ) | TAG_CLOSING; // make it a closing tag
+    assert( stream.peekNextToken() == OPENING( token ));
+    // just find the matching closing tag
+    if( recoverAndFindTag( closing ))
+    {
+        stream.getNextToken(); // read it
+        return;
+    }
+    fprintf( stderr, "Expected end of element %d not found.\n", token );
+}
+
+void SmOoxmlImport::handleUnexpectedTag()
+{
+    int next = stream.peekNextToken();
+    if( next == XML_TOKEN_INVALID )
+        return; // end of stream
+    if( next == CLOSING( next ))
+    {
+        stream.getNextToken(); // just skip it
+        return;
+    }
+    skipElement( stream.peekNextToken());
+}
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/starmath/source/ooxmlimport.hxx b/starmath/source/ooxmlimport.hxx
new file mode 100644
index 000000000000..80be1d836b5c
--- /dev/null
+++ b/starmath/source/ooxmlimport.hxx
@@ -0,0 +1,78 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * Version: MPL 1.1 / GPLv3+ / LGPLv3+
+ *
+ * The contents of this file are subject to the Mozilla Public License Version
+ * 1.1 (the "License"); you may not use this file except in compliance with
+ * the License or as specified alternatively below. You may obtain a copy of
+ * the License at http://www.mozilla.org/MPL/
+ *
+ * Software distributed under the License is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
+ * for the specific language governing rights and limitations under the
+ * License.
+ *
+ * Major Contributor(s):
+ * Copyright (C) 2011 Lubos Lunak <l.lunak@suse.cz> (initial developer)
+ *
+ * All Rights Reserved.
+ *
+ * For minor contributions see the git repository.
+ *
+ * Alternatively, the contents of this file may be used under the terms of
+ * either the GNU General Public License Version 3 or later (the "GPLv3+"), or
+ * the GNU Lesser General Public License Version 3 or later (the "LGPLv3+"),
+ * in which case the provisions of the GPLv3+ or the LGPLv3+ are applicable
+ * instead of those above.
+ */
+
+#ifndef SM_OOXMLIMPORT_HXX
+#define SM_OOXMLIMPORT_HXX
+
+#include <oox/export/starmathimport.hxx>
+
+#include "node.hxx"
+
+/**
+ Class implementing reading of formulas from OOXML. The toplevel element is expected
+ to be oMath (handle oMathPara outside of this code).
+ */
+class SmOoxmlImport
+{
+public:
+    SmOoxmlImport( ooxmlformulaimport::XmlStream& stream );
+    rtl::OUString ConvertToStarMath();
+private:
+    rtl::OUString handleStream();
+    rtl::OUString handleF();
+    rtl::OUString readR();
+    /**
+     Checks that the next token is the given opening tag, if not, writes out a warning
+     and tries to recover (skips tags until found or until the current element would end).
+    */
+    void checkOpeningTag( int token );
+    /**
+     Checks that the next token is the given opening tag, if not, writes out a warning
+     and tries to recover (skips tags until found or until the current element would end).
+    */
+    void checkClosingTag( int token );
+    // helper for the two above
+    void checkTag( int token, const char* txt );
+    /**
+     Tries to find the given token, until either found (returns true) or end of current element.
+    */
+    bool recoverAndFindTag( int token );
+    /**
+     Skips the given element (i.e. reads up to and including the matching closing tag).
+    */
+    void skipElement( int token );
+    /**
+     Handle the current (unexpected) tag.
+    */
+    void handleUnexpectedTag();
+    ooxmlformulaimport::XmlStream& stream;
+};
+
+#endif
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/starmath/source/unomodel.cxx b/starmath/source/unomodel.cxx
index 2cb17987b2ac..8b7ea8169235 100644
--- a/starmath/source/unomodel.cxx
+++ b/starmath/source/unomodel.cxx
@@ -1137,4 +1137,9 @@ void SmModel::writeFormulaOoxml( ::sax_fastparser::FSHelperPtr m_pSerializer, oo
     static_cast< SmDocShell* >( GetObjectShell())->writeFormulaOoxml( m_pSerializer, version );
 }
 
+void SmModel::readFormulaOoxml( ooxmlformulaimport::XmlStream& stream )
+{
+    static_cast< SmDocShell* >( GetObjectShell())->readFormulaOoxml( stream );
+}
+
 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */