preprocessor.cpp

qt-x11-opensource-src-4.1.4.tar.gz源码

/****************************************************************************
**
** Copyright (C) 1992-2006 Trolltech ASA. All rights reserved.
**
** This file is part of the tools applications of the Qt Toolkit.
**
** This file may be used under the terms of the GNU General Public
** License version 2.0 as published by the Free Software Foundation
** and appearing in the file LICENSE.GPL included in the packaging of
** this file.  Please review the following information to ensure GNU
** General Public Licensing requirements will be met:
** http://www.trolltech.com/products/qt/opensource.html
**
** If you are unsure which license is appropriate for your use, please
** review the following information:
** http://www.trolltech.com/products/qt/licensing.html or contact the
** sales department at [e-mail].
**
** This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
** WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
**
****************************************************************************/

#include "preprocessor.h"
#include "utils.h"
#include "ppkeywords.cpp"
#include <QStringList>
#include <QFile>
#include <QDir>
#include <QFileInfo>

QList<QByteArray> Preprocessor::includes;
Macros Preprocessor::macros;
bool Preprocessor::onlyPreprocess = false;
QByteArray Preprocessor::protocol;
QSet<QByteArray> Preprocessor::preprocessedIncludes;


static inline bool hasNext(const Symbols &symbols, int i)
{ return (i < symbols.size()); }

static inline const Symbol &next(const Symbols &symbols, int &i)
{ return symbols.at(i++); }

static void skipUntilEndif(const Symbols &symbols, int &i)
{
    while(i < symbols.size() - 1 && symbols.at(i).pp_token != PP_ENDIF){
        switch (symbols.at(i).pp_token) {
        case PP_IF:
        case PP_IFDEF:
        case PP_IFNDEF:
            ++i;
            skipUntilEndif(symbols, i);
            break;
        default:
            ;
        }
        ++i;
    }
}
static bool skipBranch(const Symbols &symbols, int &i)
{
    while(i < symbols.size() - 1
          && (symbols.at(i).pp_token != PP_ENDIF
               && symbols.at(i).pp_token != PP_ELIF
               && symbols.at(i).pp_token != PP_ELSE)
       ){
        switch (symbols.at(i).pp_token) {
        case PP_IF:
        case PP_IFDEF:
        case PP_IFNDEF:
            ++i;
            skipUntilEndif(symbols, i);
            break;
        default:
            ;
        }
        ++i;
    }
    return (i < symbols.size() - 1);
}

static QByteArray cleaned(const QByteArray &input)
{
    QByteArray result;
    result.reserve(input.size());
    const char *data = input;
    char *output = result.data();

    int newlines = 0;
    while (*data) {
        while (*data && is_space(*data))
            ++data;
        bool takeLine = (*data == '#');
        if (*data == '%' && *(data+1) == ':') {
            takeLine = true;
            ++data;
        }
        if (takeLine) {
            *output = '#';
            ++output;
            do ++data; while (*data && is_space(*data));
        }
        while (*data) {
            if (*data == '\\' && *(data+1) == '\n') {
                ++newlines;
                data += 2;
                continue;
            }
            *output = *data;
            ++output;
            if (*data == '\n') {
                while (newlines) {
                    *output = '\n';
                    ++output;
                    --newlines;
                }
                ++data;
                break;
            }
            ++data;
        }
    }
    result.resize(output - result.constData());
    return result;
}

enum TokenizeMode { TokenizeFile, TokenizeLine };
static Symbols tokenize(const QByteArray &input, int lineNum = 1, TokenizeMode mode = TokenizeFile)
{
    Symbols symbols;
    const char *begin = input;
    const char *data = begin;
    while (*data) {
        const char *lexem = data;
        int state = 0;
        PP_Token token = PP_NOTOKEN;
        for (;;) {
            if (static_cast<signed char>(*data) < 0) {
                ++data;
                continue;
            }
            int nextindex = pp_keywords[state].next;
            int next = 0;
            if (*data == pp_keywords[state].defchar)
                next = pp_keywords[state].defnext;
            else if (!state || nextindex)
                    next = pp_keyword_trans[nextindex][(int)*data];
            if (!next)
                break;
            state = next;
            token = pp_keywords[state].token;
            ++data;
        }
        // suboptimal, is_ident_char  should use a table
        if (pp_keywords[state].ident && is_ident_char(*data))
            token = pp_keywords[state].ident;

        switch (token) {
        case NOTOKEN:
            ++data;
            break;
        case PP_IFDEF:
            symbols += Symbol(lineNum, PP_IF, QByteArray());
            symbols += Symbol(lineNum, PP_DEFINED, QByteArray());
            continue;
        case PP_IFNDEF:
            symbols += Symbol(lineNum, PP_IF, QByteArray());
            symbols += Symbol(lineNum, PP_NOT, QByteArray());
            symbols += Symbol(lineNum, PP_DEFINED, QByteArray());
            continue;
        case PP_QUOTE:
            data = skipQuote(data);
            token = PP_STRING_LITERAL;
            break;
        case PP_SINGLEQUOTE:
            while (*data && (*data != '\''
                             || (*(data-1)=='\\'
                                  && *(data-2)!='\\')))
                ++data;
            if (*data)
                ++data;
            token = PP_CHARACTER_LITERAL;
            break;
        case PP_DIGIT:
            while (is_digit_char(*data))
                ++data;
            if (!*data || *data != '.') {
                token = PP_INTEGER_LITERAL;
                if (data - lexem == 1 &&
                     (*data == 'x' || *data == 'X')
                     && *lexem == '0') {
                    ++data;
                    while (is_hex_char(*data))
                        ++data;
                }
                break;
            }
            token = PP_FLOATING_LITERAL;
            ++data;
            // fall through
        case PP_FLOATING_LITERAL:
            while (is_digit_char(*data))
                ++data;
            if (*data == '+' || *data == '-')
                ++data;
            if (*data == 'e' || *data == 'E') {
                ++data;
                while (is_digit_char(*data))
                    ++data;
            }
            if (*data == 'f' || *data == 'F'
                || *data == 'l' || *data == 'L')
                ++data;
            break;
        case PP_CHARACTER:
            while (is_ident_char(*data))
                ++data;
            token = PP_IDENTIFIER;
            break;
        case PP_C_COMMENT:
            while (*data && (*(data-1) != '/' || *(data-2) != '*')) {
                if (*data == '\n')
                    ++lineNum;
                ++data;
            }
            token = PP_WHITESPACE; // one comment, one whitespace
            // fall through;
        case PP_WHITESPACE:
            while (*data && (*data == ' ' || *data == '\t'))
                ++data;
            break;
        case PP_CPP_COMMENT:
            while (*data && *data != '\n')
                ++data;
            continue; // ignore safly, the newline is a seperator
        case PP_NEWLINE:
            if (mode == TokenizeLine)
                goto exit;
            ++lineNum;
            break;
        default:
            break;
        }
        symbols += Symbol(lineNum, token, input, lexem-begin, data-lexem);
    }
exit:
    symbols += Symbol(); // eof symbol
    return symbols;
}

//static Symbols tokenize(const Symbol &symbol)
//{ return tokenize(symbol.lexem(), symbol.lineNum, TokenizeLine); }

static Symbols substitute(const Macros &macros, const Symbols& symbols, int &i,
                          bool discardWhitespace = false, QList<QByteArray> safeset = QList<QByteArray>())
{
    QByteArray lexem = symbols.at(i-1).lexem();

    if (!macros.contains(lexem) || safeset.contains(lexem))
        return Symbols(1, symbols.at(i-1));

    QByteArray macro = macros.value(lexem);

                       // ### cannot do parameters yet, TODO
    if (macro.size() && macro.at(0) == '(')
        return Symbols(1, symbols.at(i-1));

    safeset += lexem;
    Symbols syms = tokenize(macro, symbols.at(i-1).lineNum, TokenizeLine);

    Symbols result;
    int j = 0;
    bool skip = false;
    while (hasNext(syms, j+1)) {
        const Symbol &sym = next(syms, j);
        if (discardWhitespace && sym.pp_token == PP_WHITESPACE)
            continue;
        if (sym.pp_token == PP_IDENTIFIER && !skip)
            result += substitute(macros, syms, j, discardWhitespace, safeset);
        else
            result += sym;
        skip = (sym.pp_token == PP_DEFINED || skip && sym.pp_token == PP_LPAREN);
    }
    return result;
}

struct PP_Expression
{
    PP_Expression():i(0){}
    Macros macros;
    Symbols symbols;
    int i;

    int value() { i = 0; return unary_expression_lookup() ?  conditional_expression() : 0; }

    inline bool hasNext() const { return (i <= symbols.size()); }
    inline PP_Token next() { return symbols.at(i++).pp_token; }
    bool test(PP_Token);
    inline void prev() {--i;}
    PP_Token lookup(int k = 1);
    inline PP_Token token() { return symbols.at(i-1).pp_token;}
    inline QByteArray lexem() { return symbols.at(i-1).lexem();}

    int conditional_expression();
    int logical_OR_expression();
    int logical_AND_expression();
    int inclusive_OR_expression();
    int exclusive_OR_expression();
    int AND_expression();
    int equality_expression();
    int relational_expression();
    int shift_expression();
    int additive_expression();
    int multiplicative_expression();
    int unary_expression();
    bool unary_expression_lookup();
    int primary_expression();
    bool primary_expression_lookup();
};

inline bool PP_Expression::test(PP_Token token)
{
    if (i < symbols.size() && symbols.at(i).pp_token == token) {
        ++i;
        return true;
    }
    return false;
}

inline PP_Token PP_Expression::lookup(int k)
{
    const int l = i - 1 + k;
    return l < symbols.size() ? symbols.at(l).pp_token : PP_NOTOKEN;
}

int PP_Expression::conditional_expression()
{
    int value = logical_OR_expression();
    if (test(PP_QUESTION)) {
        int alt1 = conditional_expression();
        int alt2 = test(PP_COLON) ? conditional_expression() : 0;
        return value ? alt1 : alt2;
    }
    return value;
}

int PP_Expression::logical_OR_expression()
{
    int value = logical_AND_expression();
    if (test(PP_OROR))
        return logical_OR_expression() || value;
    return value;
}

int PP_Expression::logical_AND_expression()
{
    int value = inclusive_OR_expression();
    if (test(PP_ANDAND))
        return logical_AND_expression() && value;
    return value;
}

int PP_Expression::inclusive_OR_expression()
{
    int value = exclusive_OR_expression();
    if (test(PP_OR))
        return value | inclusive_OR_expression();
    return value;
}

int PP_Expression::exclusive_OR_expression()
{
    int value = AND_expression();
    if (test(PP_HAT))
        return value ^ exclusive_OR_expression();
    return value;
}

int PP_Expression::AND_expression()
{
    int value = equality_expression();
    if (test(PP_AND))
        return value & AND_expression();
    return value;
}

int PP_Expression::equality_expression()
{
    int value = relational_expression();
    switch (next()) {
    case PP_EQEQ:
        return value == equality_expression();
    case PP_NE:
        return value != equality_expression();
    default:
        prev();
        return value;
    }
}

int PP_Expression::relational_expression()
{
    int value = shift_expression();
    switch (next()) {
    case PP_LANGLE:
        return value < relational_expression();
    case PP_RANGLE:
        return value > relational_expression();
    case PP_LE:
        return value <= relational_expression();
    case PP_GE:
        return value >= relational_expression();
    default:
        prev();
        return value;
    }
}

int PP_Expression::shift_expression()
{
    int value = additive_expression();
    switch (next()) {
    case PP_LTLT:
        return value << shift_expression();
    case PP_GTGT: