xref: /AOO41X/main/basic/source/comp/scanner.cxx (revision e1f63238eb022c8a12b30d46a012444ff20e0951)

/**************************************************************
 *
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing,
 * software distributed under the License is distributed on an
 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
 * KIND, either express or implied.  See the License for the
 * specific language governing permissions and limitations
 * under the License.
 *
 *************************************************************/



// MARKER(update_precomp.py): autogen include statement, do not remove
#include "precompiled_basic.hxx"

#include "sbcomp.hxx"
#include <stdio.h>
#include <string.h>
#include <ctype.h>
#if defined UNX
#include <stdlib.h>
#else
#include <math.h>   // atof()
#endif
#include <rtl/math.hxx>
#include <vcl/svapp.hxx>
#include <unotools/charclass.hxx>

#include <runtime.hxx>

SbiScanner::SbiScanner( const ::rtl::OUString& rBuf, StarBASIC* p ) : aBuf( rBuf )
{
    pBasic   = p;
    pLine    = NULL;
    nVal     = 0;
    eScanType = SbxVARIANT;
    nErrors  = 0;
    nBufPos  = 0;
    nCurCol1 = 0;
    nSavedCol1 = 0;
    nColLock = 0;
    nLine    = 0;
    nCol1    = 0;
    nCol2    = 0;
    nCol     = 0;
    bError   =
    bAbort   =
    bSpaces  =
    bNumber  =
    bSymbol  =
    bUsedForHilite =
    bCompatible =
    bVBASupportOn =
    bPrevLineExtentsComment = sal_False;
    bHash    =
    bErrors  = sal_True;
}

SbiScanner::~SbiScanner()
{}

void SbiScanner::LockColumn()
{
    if( !nColLock++ )
        nSavedCol1 = nCol1;
}

void SbiScanner::UnlockColumn()
{
    if( nColLock )
        nColLock--;
}

void SbiScanner::GenError( SbError code )
{
    if( GetSbData()->bBlockCompilerError )
    {
        bAbort = sal_True;
        return;
    }
    if( !bError && bErrors )
    {
        sal_Bool bRes = sal_True;
        // Nur einen Fehler pro Statement reporten
        bError = sal_True;
        if( pBasic )
        {
            // Falls EXPECTED oder UNEXPECTED kommen sollte, bezieht es sich
            // immer auf das letzte Token, also die Col1 uebernehmen
            sal_uInt16 nc = nColLock ? nSavedCol1 : nCol1;
            switch( code )
            {
                case SbERR_EXPECTED:
                case SbERR_UNEXPECTED:
                case SbERR_SYMBOL_EXPECTED:
                case SbERR_LABEL_EXPECTED:
                    nc = nCol1;
                    if( nc > nCol2 ) nCol2 = nc;
                    break;
            }
            bRes = pBasic->CError( code, aError, nLine, nc, nCol2 );
        }
        bAbort |= !bRes |
             ( code == SbERR_NO_MEMORY || code == SbERR_PROG_TOO_LARGE );
    }
    if( bErrors )
        nErrors++;
}

// Falls sofort ein Doppelpunkt folgt, wird sal_True zurueckgeliefert.
// Wird von SbiTokenizer::MayBeLabel() verwendet, um einen Label zu erkennen

sal_Bool SbiScanner::DoesColonFollow()
{
    if( pLine && *pLine == ':' )
    {
        pLine++; nCol++; return sal_True;
    }
    else return sal_False;
}

// Testen auf ein legales Suffix

static SbxDataType GetSuffixType( sal_Unicode c )
{
    static String aSuffixesStr = String::CreateFromAscii( "%&!#@ $" );
    if( c )
    {
        sal_uInt32 n = aSuffixesStr.Search( c );
        if( STRING_NOTFOUND != n && c != ' ' )
            return SbxDataType( (sal_uInt16) n + SbxINTEGER );
    }
    return SbxVARIANT;
}

// Einlesen des naechsten Symbols in die Variablen aSym, nVal und eType
// Returnwert ist sal_False bei EOF oder Fehlern
#define BUF_SIZE 80

namespace {

/** Returns true, if the passed character is a white space character. */
inline bool lclIsWhitespace( sal_Unicode cChar )
{
    return (cChar == ' ') || (cChar == '\t') || (cChar == '\f');
}

} // namespace

sal_Bool SbiScanner::NextSym()
{
    // Fuer den EOLN-Fall merken
    sal_uInt16 nOldLine = nLine;
    sal_uInt16 nOldCol1 = nCol1;
    sal_uInt16 nOldCol2 = nCol2;
    sal_Unicode buf[ BUF_SIZE ], *p = buf;
    bHash = sal_False;

    eScanType = SbxVARIANT;
    aSym.Erase();
    bSymbol =
    bNumber = bSpaces = sal_False;

    // Zeile einlesen?
    if( !pLine )
    {
        sal_Int32 n = nBufPos;
        sal_Int32 nLen = aBuf.getLength();
        if( nBufPos >= nLen )
            return sal_False;
        const sal_Unicode* p2 = aBuf.getStr();
        p2 += n;
        while( ( n < nLen ) && ( *p2 != '\n' ) && ( *p2 != '\r' ) )
            p2++, n++;
        // #163944# ignore trailing whitespace
        sal_Int32 nCopyEndPos = n;
        while( (nBufPos < nCopyEndPos) && lclIsWhitespace( aBuf[ nCopyEndPos - 1 ] ) )
            --nCopyEndPos;
        aLine = aBuf.copy( nBufPos, nCopyEndPos - nBufPos );
        if( n < nLen )
        {
            if( *p2 == '\r' && *( p2+1 ) == '\n' )
                n += 2;
            else
                n++;
        }
        nBufPos = n;
        pLine = aLine.getStr();
        nOldLine = ++nLine;
        nCol = nCol1 = nCol2 = nOldCol1 = nOldCol2 = 0;
        nColLock = 0;
    }

    // Leerstellen weg:
    while( lclIsWhitespace( *pLine ) )
        pLine++, nCol++, bSpaces = sal_True;

    nCol1 = nCol;

    // nur Leerzeile?
    if( !*pLine )
        goto eoln;

    if( bPrevLineExtentsComment )
        goto PrevLineCommentLbl;

    if( *pLine == '#' )
    {
        pLine++;
        nCol++;
        bHash = sal_True;
    }

    // Symbol? Dann Zeichen kopieren.
    if( BasicSimpleCharClass::isAlpha( *pLine, bCompatible ) || *pLine == '_' )
    {
        // Wenn nach '_' nichts kommt, ist es ein Zeilenabschluss!
        if( *pLine == '_' && !*(pLine+1) )
        {   pLine++;
            goto eoln;  }
        bSymbol = sal_True;
        short n = nCol;
        for ( ; (BasicSimpleCharClass::isAlphaNumeric( *pLine, bCompatible ) || ( *pLine == '_' ) ); pLine++ )
            nCol++;
        aSym = aLine.copy( n, nCol - n );

        // Special handling for "go to"
        if( bCompatible && *pLine && aSym.EqualsIgnoreCaseAscii( "go" ) )
        {
            const sal_Unicode* pTestLine = pLine;
            short nTestCol = nCol;
            while( lclIsWhitespace( *pTestLine ) )
            {
                pTestLine++;
                nTestCol++;
            }

            if( *pTestLine && *(pTestLine + 1) )
            {
                String aTestSym = aLine.copy( nTestCol, 2 );
                if( aTestSym.EqualsIgnoreCaseAscii( "to" ) )
                {
                    aSym = String::CreateFromAscii( "goto" );
                    pLine = pTestLine + 2;
                    nCol = nTestCol + 2;
                }
            }
        }

        // Abschliessendes '_' durch Space ersetzen, wenn Zeilenende folgt
        // (sonst falsche Zeilenfortsetzung)
        if( !bUsedForHilite && !*pLine && *(pLine-1) == '_' )
        {
            aSym.GetBufferAccess();     // #109693 force copy if necessary
            *((sal_Unicode*)(pLine-1)) = ' ';       // cast wegen const
        }
        // Typkennung?
        // Das Ausrufezeichen bitte nicht testen, wenn
        // danach noch ein Symbol anschliesst
        else if( *pLine != '!' || !BasicSimpleCharClass::isAlpha( pLine[ 1 ], bCompatible ) )
        {
            SbxDataType t = GetSuffixType( *pLine );
            if( t != SbxVARIANT )
            {
                eScanType = t;
                pLine++;
                nCol++;
            }
        }
    }

    // Zahl? Dann einlesen und konvertieren.
    else if( BasicSimpleCharClass::isDigit( *pLine & 0xFF )
        || ( *pLine == '.' && BasicSimpleCharClass::isDigit( *(pLine+1) & 0xFF ) ) )
    {
        short exp = 0;
        short comma = 0;
        short ndig = 0;
        short ncdig = 0;
        eScanType = SbxDOUBLE;
        sal_Bool bBufOverflow = sal_False;
        while( strchr( "0123456789.DEde", *pLine ) && *pLine )
        {
            // AB 4.1.1996: Buffer voll? -> leer weiter scannen
            if( (p-buf) == (BUF_SIZE-1) )
            {
                bBufOverflow = sal_True;
                pLine++, nCol++;
                continue;
            }
            // Komma oder Exponent?
            if( *pLine == '.' )
            {
                if( ++comma > 1 )
                {
                    pLine++; nCol++; continue;
                }
                else *p++ = *pLine++, nCol++;
            }
            else if( strchr( "DdEe", *pLine ) )
            {
                if (++exp > 1)
                {
                    pLine++; nCol++; continue;
                }
//              if( toupper( *pLine ) == 'D' )
//                  eScanType = SbxDOUBLE;
                *p++ = 'E'; pLine++; nCol++;
                // Vorzeichen hinter Exponent?
                if( *pLine == '+' )
                    pLine++, nCol++;
                else
                if( *pLine == '-' )
                    *p++ = *pLine++, nCol++;
            }
            else
            {
                *p++ = *pLine++, nCol++;
                if( comma && !exp ) ncdig++;
            }
            if (!exp) ndig++;
        }
        *p = 0;
        aSym = p; bNumber = sal_True;
        // Komma, Exponent mehrfach vorhanden?
        if( comma > 1 || exp > 1 )
        {   aError = '.';
            GenError( SbERR_BAD_CHAR_IN_NUMBER );   }

        // #57844 Lokalisierte Funktion benutzen
        nVal = rtl_math_uStringToDouble( buf, buf+(p-buf), '.', ',', NULL, NULL );
        // ALT: nVal = atof( buf );

        ndig = ndig - comma;
        if( !comma && !exp )
        {
            if( nVal >= SbxMININT && nVal <= SbxMAXINT )
                eScanType = SbxINTEGER;
            else
            if( nVal >= SbxMINLNG && nVal <= SbxMAXLNG )
                eScanType = SbxLONG;
        }
        if( bBufOverflow )
            GenError( SbERR_MATH_OVERFLOW );
        // zu viele Zahlen fuer SINGLE?
//      if (ndig > 15 || ncdig > 6)
//          eScanType = SbxDOUBLE;
//      else
//      if( nVal > SbxMAXSNG || nVal < SbxMINSNG )
//          eScanType = SbxDOUBLE;

        // Typkennung?
        SbxDataType t = GetSuffixType( *pLine );
        if( t != SbxVARIANT )
        {
            eScanType = t;
            pLine++;
            nCol++;
        }
    }

    // Hex/Oktalzahl? Einlesen und konvertieren:
    else if( *pLine == '&' )
    {
        pLine++; nCol++;
        sal_Unicode cmp1[] = { '0','1','2','3','4','5','6','7','8','9','A','B','C','D','E','F', 0 };
        sal_Unicode cmp2[] = { '0', '1', '2', '3', '4', '5', '6', '7', 0 };
        sal_Unicode *cmp = cmp1;
        //char *cmp = "0123456789ABCDEF";
        sal_Unicode base = 16;
        sal_Unicode ndig = 8;
        sal_Unicode xch  = *pLine++ & 0xFF; nCol++;
        switch( toupper( xch ) )
        {
            case 'O':
                cmp = cmp2; base = 8; ndig = 11; break;
                //cmp = "01234567"; base = 8; ndig = 11; break;
            case 'H':
                break;
            default :
                // Wird als Operator angesehen
                pLine--; nCol--; nCol1 = nCol-1; aSym = '&'; return SYMBOL;
        }
        bNumber = sal_True;
        long l = 0;
        int i;
        sal_Bool bBufOverflow = sal_False;
        while( BasicSimpleCharClass::isAlphaNumeric( *pLine & 0xFF, bCompatible ) )
        {
            sal_Unicode ch = sal::static_int_cast< sal_Unicode >(
                toupper( *pLine & 0xFF ) );
            pLine++; nCol++;
            // AB 4.1.1996: Buffer voll, leer weiter scannen
            if( (p-buf) == (BUF_SIZE-1) )
                bBufOverflow = sal_True;
            else if( String( cmp ).Search( ch ) != STRING_NOTFOUND )
            //else if( strchr( cmp, ch ) )
                *p++ = ch;
            else
            {
                aError = ch;
                GenError( SbERR_BAD_CHAR_IN_NUMBER );
            }
        }
        *p = 0;
        for( p = buf; *p; p++ )
        {
            i = (*p & 0xFF) - '0';
            if( i > 9 ) i -= 7;
            l = ( l * base ) + i;
            if( !ndig-- )
            {
                GenError( SbERR_MATH_OVERFLOW ); break;
            }
        }
        if( *pLine == '&' ) pLine++, nCol++;
        nVal = (double) l;
        eScanType = ( l >= SbxMININT && l <= SbxMAXINT ) ? SbxINTEGER : SbxLONG;
        if( bBufOverflow )
            GenError( SbERR_MATH_OVERFLOW );
    }

    // Strings:
    else if( *pLine == '"' || *pLine == '[' )
    {
        sal_Unicode cSep = *pLine;
        if( cSep == '[' )
            bSymbol = sal_True, cSep = ']';
        short n = nCol+1;
        while( *pLine )
        {
            do pLine++, nCol++;
            while( *pLine && ( *pLine != cSep ) );
            if( *pLine == cSep )
            {
                pLine++; nCol++;
                if( *pLine != cSep || cSep == ']' ) break;
            } else aError = cSep, GenError( SbERR_EXPECTED );
        }
        // If VBA Interop then doen't eat the [] chars
        if ( cSep == ']' && bVBASupportOn )
            aSym = aLine.copy( n - 1, nCol - n  + 1);
        else
            aSym = aLine.copy( n, nCol - n - 1 );
        // Doppelte Stringbegrenzer raus
        String s( cSep );
        s += cSep;
        sal_uInt16 nIdx = 0;
        do
        {
            nIdx = aSym.Search( s, nIdx );
            if( nIdx == STRING_NOTFOUND )
                break;
            aSym.Erase( nIdx, 1 );
            nIdx++;
        }
        while( true );
        if( cSep != ']' )
            eScanType = ( cSep == '#' ) ? SbxDATE : SbxSTRING;
    }
    // ungueltige Zeichen:
    else if( ( *pLine & 0xFF ) >= 0x7F )
    {
        GenError( SbERR_SYNTAX ); pLine++; nCol++;
    }
    // andere Gruppen:
    else
    {
        short n = 1;
        switch( *pLine++ )
        {
            case '<': if( *pLine == '>' || *pLine == '=' ) n = 2; break;
            case '>': if( *pLine == '=' ) n = 2; break;
            case ':': if( *pLine == '=' ) n = 2; break;
        }
        aSym = aLine.copy( nCol, n );
        pLine += n-1; nCol = nCol + n;
    }

    nCol2 = nCol-1;

PrevLineCommentLbl:
    // Kommentar?
    if( bPrevLineExtentsComment || (eScanType != SbxSTRING &&
        ( aSym.GetBuffer()[0] == '\'' || aSym.EqualsIgnoreCaseAscii( "REM" ) ) ) )
    {
        bPrevLineExtentsComment = sal_False;
        aSym = String::CreateFromAscii( "REM" );
        sal_uInt16 nLen = String( pLine ).Len();
        if( bCompatible && pLine[ nLen - 1 ] == '_' && pLine[ nLen - 2 ] == ' ' )
            bPrevLineExtentsComment = sal_True;
        nCol2 = nCol2 + nLen;
        pLine = NULL;
    }
    return sal_True;

    // Sonst Zeilen-Ende: aber bitte auf '_' testen, ob die
    // Zeile nicht weitergeht!
eoln:
    if( nCol && *--pLine == '_' )
    {
        pLine = NULL;
        bool bRes = NextSym();
        if( bVBASupportOn && aSym.GetBuffer()[0] == '.' )
        {
            // object _
            //    .Method
            // ^^^  <- spaces is legal in MSO VBA
            OSL_TRACE("*** resetting bSpaces***");
            bSpaces = sal_False;
        }
        return bRes;
    }
    else
    {
        pLine = NULL;
        nLine = nOldLine;
        nCol1 = nOldCol1;
        nCol2 = nOldCol2;
        aSym = '\n';
        nColLock = 0;
        return sal_True;
    }
}

LetterTable BasicSimpleCharClass::aLetterTable;

LetterTable::LetterTable( void )
{
    for( int i = 0 ; i < 256 ; ++i )
        IsLetterTab[i] = false;

    IsLetterTab[0xC0] = true;   // � , CAPITAL LETTER A WITH GRAVE ACCENT
    IsLetterTab[0xC1] = true;   // � , CAPITAL LETTER A WITH ACUTE ACCENT
    IsLetterTab[0xC2] = true;   // � , CAPITAL LETTER A WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xC3] = true;   // � , CAPITAL LETTER A WITH TILDE
    IsLetterTab[0xC4] = true;   // � , CAPITAL LETTER A WITH DIAERESIS
    IsLetterTab[0xC5] = true;   // � , CAPITAL LETTER A WITH RING ABOVE
    IsLetterTab[0xC6] = true;   // � , CAPITAL LIGATURE AE
    IsLetterTab[0xC7] = true;   // � , CAPITAL LETTER C WITH CEDILLA
    IsLetterTab[0xC8] = true;   // � , CAPITAL LETTER E WITH GRAVE ACCENT
    IsLetterTab[0xC9] = true;   // � , CAPITAL LETTER E WITH ACUTE ACCENT
    IsLetterTab[0xCA] = true;   // � , CAPITAL LETTER E WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xCB] = true;   // � , CAPITAL LETTER E WITH DIAERESIS
    IsLetterTab[0xCC] = true;   // � , CAPITAL LETTER I WITH GRAVE ACCENT
    IsLetterTab[0xCD] = true;   // � , CAPITAL LETTER I WITH ACUTE ACCENT
    IsLetterTab[0xCE] = true;   // � , CAPITAL LETTER I WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xCF] = true;   // � , CAPITAL LETTER I WITH DIAERESIS
    IsLetterTab[0xD0] = true;   // � , CAPITAL LETTER ETH
    IsLetterTab[0xD1] = true;   // � , CAPITAL LETTER N WITH TILDE
    IsLetterTab[0xD2] = true;   // � , CAPITAL LETTER O WITH GRAVE ACCENT
    IsLetterTab[0xD3] = true;   // � , CAPITAL LETTER O WITH ACUTE ACCENT
    IsLetterTab[0xD4] = true;   // � , CAPITAL LETTER O WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xD5] = true;   // � , CAPITAL LETTER O WITH TILDE
    IsLetterTab[0xD6] = true;   // � , CAPITAL LETTER O WITH DIAERESIS
    IsLetterTab[0xD8] = true;   // � , CAPITAL LETTER O WITH STROKE
    IsLetterTab[0xD9] = true;   // � , CAPITAL LETTER U WITH GRAVE ACCENT
    IsLetterTab[0xDA] = true;   // � , CAPITAL LETTER U WITH ACUTE ACCENT
    IsLetterTab[0xDB] = true;   // � , CAPITAL LETTER U WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xDC] = true;   // � , CAPITAL LETTER U WITH DIAERESIS
    IsLetterTab[0xDD] = true;   // � , CAPITAL LETTER Y WITH ACUTE ACCENT
    IsLetterTab[0xDE] = true;   // � , CAPITAL LETTER THORN
    IsLetterTab[0xDF] = true;   // � , SMALL LETTER SHARP S
    IsLetterTab[0xE0] = true;   // � , SMALL LETTER A WITH GRAVE ACCENT
    IsLetterTab[0xE1] = true;   // � , SMALL LETTER A WITH ACUTE ACCENT
    IsLetterTab[0xE2] = true;   // � , SMALL LETTER A WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xE3] = true;   // � , SMALL LETTER A WITH TILDE
    IsLetterTab[0xE4] = true;   // � , SMALL LETTER A WITH DIAERESIS
    IsLetterTab[0xE5] = true;   // � , SMALL LETTER A WITH RING ABOVE
    IsLetterTab[0xE6] = true;   // � , SMALL LIGATURE AE
    IsLetterTab[0xE7] = true;   // � , SMALL LETTER C WITH CEDILLA
    IsLetterTab[0xE8] = true;   // � , SMALL LETTER E WITH GRAVE ACCENT
    IsLetterTab[0xE9] = true;   // � , SMALL LETTER E WITH ACUTE ACCENT
    IsLetterTab[0xEA] = true;   // � , SMALL LETTER E WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xEB] = true;   // � , SMALL LETTER E WITH DIAERESIS
    IsLetterTab[0xEC] = true;   // � , SMALL LETTER I WITH GRAVE ACCENT
    IsLetterTab[0xED] = true;   // � , SMALL LETTER I WITH ACUTE ACCENT
    IsLetterTab[0xEE] = true;   // � , SMALL LETTER I WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xEF] = true;   // � , SMALL LETTER I WITH DIAERESIS
    IsLetterTab[0xF0] = true;   // � , SMALL LETTER ETH
    IsLetterTab[0xF1] = true;   // � , SMALL LETTER N WITH TILDE
    IsLetterTab[0xF2] = true;   // � , SMALL LETTER O WITH GRAVE ACCENT
    IsLetterTab[0xF3] = true;   // � , SMALL LETTER O WITH ACUTE ACCENT
    IsLetterTab[0xF4] = true;   // � , SMALL LETTER O WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xF5] = true;   // � , SMALL LETTER O WITH TILDE
    IsLetterTab[0xF6] = true;   // � , SMALL LETTER O WITH DIAERESIS
    IsLetterTab[0xF8] = true;   // � , SMALL LETTER O WITH OBLIQUE BAR
    IsLetterTab[0xF9] = true;   // � , SMALL LETTER U WITH GRAVE ACCENT
    IsLetterTab[0xFA] = true;   // � , SMALL LETTER U WITH ACUTE ACCENT
    IsLetterTab[0xFB] = true;   // � , SMALL LETTER U WITH CIRCUMFLEX ACCENT
    IsLetterTab[0xFC] = true;   // � , SMALL LETTER U WITH DIAERESIS
    IsLetterTab[0xFD] = true;   // � , SMALL LETTER Y WITH ACUTE ACCENT
    IsLetterTab[0xFE] = true;   // � , SMALL LETTER THORN
    IsLetterTab[0xFF] = true;   // � , SMALL LETTER Y WITH DIAERESIS
}

bool LetterTable::isLetterUnicode( sal_Unicode c )
{
    static CharClass* pCharClass = NULL;
    if( pCharClass == NULL )
        pCharClass = new CharClass( Application::GetSettings().GetLocale() );
    String aStr( c );
    bool bRet = pCharClass->isLetter( aStr, 0 );
    return bRet;
}