Html.cpp

Go to the documentation of this file.

#include <dcl/Object.h>
 
#if __DCL_HAVE_ALLOC_DEBUG
#undef __DCL_ALLOC_LEVEL
#define __DCL_ALLOC_LEVEL   __DCL_ALLOC_INTERNAL
#endif
 
#include <dcl/String.h>
#include <dcl/Array.h>
#include <dcl/Regex.h>
#include <dcl/Html.h>
 
#if __DCL_HAVE_THIS_FILE__
#undef __THIS_FILE__
static const char_t* __THIS_FILE__ = __T("dcl/Html.cpp");
#endif
 
__DCL_BEGIN_NAMESPACE
 
// http://www.w3.org/TR/xhtml1/dtds.html#a_dtd_Special_characters
// http://www.w3.org/TR/xhtml1/DTD/xhtml-lat1.ent 
// http://www.w3.org/TR/xhtml1/DTD/xhtml-special.ent
// http://www.w3.org/TR/xhtml1/DTD/xhtml-symbol.ent
//
// notes!
//  User-Agent가 DTD를 읽어들이지 못하면 다음의 니모닉들이 제대로 표시하지 못할 수 있다.
//  apos의 경우 IE에서 &apos;를 문자열 그대로 표시한다.
 
struct EntitySpecial
{
    wchar_t         ch;
    const wchar_t*  psz;
    size_t          n;
};
 
static EntitySpecial __spacialChars[] =
{
    { L'<',     L"&lt;",    4 },
    { L'>',     L"&gt;",    4 },
    { L'\"',    L"&quot;",  6 },
    { L'\'',    L"&#39;",   5 },
    { L'\0',    NULL,       0 }
};
 
static const EntitySpecial* __special(wchar_t c)
{
    const EntitySpecial* pEntity = __spacialChars;
    for( ; pEntity->ch != L'\0'; pEntity++)
    {
        if (c == pEntity->ch)
            return pEntity;
    }
    return NULL;
}
 
static bool __exists(const wchar_t* psz, wchar_t c)
{
    if (psz == NULL)
        return true;
 
    for( ; *psz; psz++)
    {
        if (*psz == c)
            return true;
    }
    return false;
}
 
String Html::escape(
    const String& _str,
    const wchar_t* _chars       // " <>"'&
)
{
    StringBuilder r;
 
    const EntitySpecial* pEntity = NULL;
 
    const wchar_t* pStart = _str.data();
    const wchar_t* pCurrent = pStart;
 
    while(*pCurrent)
    {
        if (__exists(_chars, *pCurrent) 
            && (pEntity = __special(*pCurrent)))
        {
            if (pStart < pCurrent)
                r.append(pStart, pCurrent - pStart);
            r.append(pEntity->psz, pEntity->n);
            pStart = ++pCurrent;
        }
        else
            ++pCurrent;
    }
 
    if (pStart < pCurrent)
        r.append(pStart, pCurrent - pStart);
 
/*
    __DCL_TRACE2(
        "HtmlEntity::escape length src: %d, res: %d\n",
        nSourceLength,
        r.length()
    );
*/
 
    return r;
 
}
 
static String __GetSpace(int nCount)
{
    StringBuilder r;
    for(int i = 0; i < nCount; i++) {
        if (i % 2)
            r.append(L"&nbsp;", 6);
        else
            r.append(' ', 1);
    }
    return r;
}
 
String Html::format(
    const String&   _str,
    int             _tab2Space,
    const String&   _beginOfLine,
    const String&   _endOfLine
)
{   
    StringBuilder r;
    r = _beginOfLine;
 
    const EntitySpecial* pEntity = NULL;
    const wchar_t* pStart = _str.data();
    const wchar_t* pCurrent = pStart;
 
    while(*pCurrent) {
        switch(*pCurrent) {
        case L'\t' :
            if (pStart < pCurrent)
                r.append(pStart, pCurrent - pStart);
 
            if (_tab2Space > 0) {
                // TAB을 ' '으로 변환한다.
                int nSpace = _tab2Space;
                while(*(++pCurrent)) {
                    if (*pCurrent == L' ')
                        nSpace++;
                    else if (*pCurrent == L'\t')
                        nSpace += _tab2Space;
                    else
                        break;
                }
                r += __GetSpace(nSpace);
                pStart = pCurrent;
            }
            else {
                // TAB을 변환하지않는다.
                r.append(L'\t', 1);
                pStart = ++pCurrent;
            }
            break;
        case ' ' :
            if (pStart < pCurrent)
                r.append(pStart, pCurrent - pStart);
 
            if (_tab2Space > 0) {
                int nSpace = 1;
                while(*(++pCurrent)) {
                    if (*pCurrent == L' ')
                        nSpace++;
                    else if (*pCurrent == L'\t')
                        nSpace += _tab2Space;
                    else
                        break;
                }
                r += __GetSpace(nSpace);
                pStart = pCurrent;
            }
            else {
                int nSpace = 1;
                while(*(++pCurrent) == L' ')
                    nSpace++;
                r += __GetSpace(nSpace);
                pStart = pCurrent;
            }
            break;
        case '\r' :
            if (pStart < pCurrent)
                r.append(pStart, pCurrent - pStart);
 
            pStart = ++pCurrent;
            break;
        case '\n' :
            if (pStart < pCurrent)
                r.append(pStart, pCurrent - pStart);
            r += _endOfLine;
            r.append(L"\r\n", 2);
            r += _beginOfLine;
            pStart = ++pCurrent;
            break;
        default :
            if (__exists(L"<>&", *pCurrent)
                && (pEntity = __special(*pCurrent))) {
                // <, >, & 문자를 escape 한다.
                if (pStart < pCurrent)
                    r.append(pStart, pCurrent - pStart);
                r.append(pEntity->psz, pEntity->n);
                pStart = ++pCurrent;
            }
            else
                ++pCurrent;
        }
    }
 
    if (pStart < pCurrent)
        r.append(pStart, pCurrent - pStart);   // 나머지
 
    r += _endOfLine;
 
/*
    __DCL_TRACE2(
        "HtmlEntity::format length src: %d, res: %d\n",
        nSourceLength,
        r.length()
        );
*/  
    return r;
}
 
String Html::strip(
    const String& _str,
    const wchar_t* _elementNames    // ',' delimiter ex: "html,head,meta,!"
)
{
    String pattern;
    StringBuilder r;
 
    if (_elementNames) {
        StringBuilder sb;
        StringArray a;
        String::split(
            _elementNames, _elementNames + String::length(_elementNames),
            L',', a);
 
        for( StringArray::Iterator it = a.begin();
             it != a.end(); it++) {
            (*it).trim();
            if (!(*it).isEmpty())
            {
                if (!pattern.isEmpty())
                    sb += L"|";
                sb += L"(<+[ \t\r\n]*/*[ \t\r\n]*";
                sb += *it;
                sb += L"[ \t\r\n]*/*[^>]*>+)";
            }
        }
        pattern = sb.toString();
    }
    else {
        // NULL 이면 모든 html
        pattern = L"(<+[^>]*>+)|(&[a-z]+;)";
    }
 
    Regex re(pattern, Regex::ICASE);
    Regex::MatchResults match;
    const wchar_t* begin = _str.data();
    const wchar_t* end = begin + _str.length();
    while(begin < end && re.search(begin, end, match)) {
        __DCL_ASSERT(match.size() > 0);
        r.append(begin, match[0].first);
        begin = match[0].second;
    }
 
    if (begin < end) {
        r.append(begin, end);
    }
 
    return r;
}
 
__DCL_END_NAMESPACE