/////////////////////////////////////////////////////////////////////////////
-// Name: winpars.cpp
+// Name: src/html/winpars.cpp
// Purpose: wxHtmlParser class (generic parser)
// Author: Vaclav Slavik
// RCS-ID: $Id$
#include "wx/wxprec.h"
-#include "wx/defs.h"
-#if wxUSE_HTML && wxUSE_STREAMS
-
#ifdef __BORLANDC__
-#pragma hdrstop
+ #pragma hdrstop
#endif
-#ifndef WXPRECOMP
+#if wxUSE_HTML && wxUSE_STREAMS
+
+#ifndef WX_PRECOMP
#include "wx/intl.h"
#include "wx/dc.h"
+ #include "wx/log.h"
+ #include "wx/settings.h"
#endif
#include "wx/html/htmldefs.h"
#include "wx/html/winpars.h"
#include "wx/html/htmlwin.h"
#include "wx/fontmap.h"
-#include "wx/log.h"
-#include "wx/settings.h"
#include "wx/uri.h"
m_CharHeight = m_CharWidth = 0;
m_UseLink = false;
#if !wxUSE_UNICODE
+ m_nbsp = 0;
m_EncConv = NULL;
m_InputEnc = wxFONTENCODING_ISO8859_1;
m_OutputEnc = wxFONTENCODING_DEFAULT;
#endif
+ m_whitespaceMode = Whitespace_Normal;
m_lastWordCell = NULL;
+ m_posColumn = 0;
{
int i, j, k, l, m;
m_Modules.DeleteObject(module);
}
-void wxHtmlWinParser::SetFonts(const wxString& normal_face, const wxString& fixed_face,
+// build all HTML font sizes (1..7) from the given base size
+static void wxBuildFontSizes(int *sizes, int size)
+{
+ // using a fixed factor (1.2, from CSS2) is a bad idea as explained at
+ // http://www.w3.org/TR/CSS21/fonts.html#font-size-props but this is by far
+ // simplest thing to do so still do it like this for now
+ sizes[0] = int(size * 0.75); // exception to 1.2 rule, otherwise too small
+ sizes[1] = int(size * 0.83);
+ sizes[2] = size;
+ sizes[3] = int(size * 1.2);
+ sizes[4] = int(size * 1.44);
+ sizes[5] = int(size * 1.73);
+ sizes[6] = int(size * 2);
+}
+
+static int wxGetDefaultHTMLFontSize()
+{
+ // base the default font size on the size of the default system font but
+ // also ensure that we have a font of reasonable size, otherwise small HTML
+ // fonts are unreadable
+ int size = wxNORMAL_FONT->GetPointSize();
+ if ( size < 10 )
+ size = 10;
+ return size;
+}
+
+void wxHtmlWinParser::SetFonts(const wxString& normal_face,
+ const wxString& fixed_face,
const int *sizes)
{
- static int default_sizes[7] =
- {
- wxHTML_FONT_SIZE_1,
- wxHTML_FONT_SIZE_2,
- wxHTML_FONT_SIZE_3,
- wxHTML_FONT_SIZE_4,
- wxHTML_FONT_SIZE_5,
- wxHTML_FONT_SIZE_6,
- wxHTML_FONT_SIZE_7
- };
+ static int default_sizes[7] = { 0 };
+ if ( !sizes )
+ {
+ if ( !default_sizes[0] )
+ wxBuildFontSizes(default_sizes, wxGetDefaultHTMLFontSize());
- if (sizes == NULL) sizes = default_sizes;
+ sizes = default_sizes;
+ }
int i, j, k, l, m;
- for (i = 0; i < 7; i++) m_FontsSizes[i] = sizes[i];
+ for (i = 0; i < 7; i++)
+ m_FontsSizes[i] = sizes[i];
+
m_FontFaceFixed = fixed_face;
m_FontFaceNormal = normal_face;
const wxString& normal_face,
const wxString& fixed_face)
{
- wxFont defaultFont = wxSystemSettings::GetFont(wxSYS_DEFAULT_GUI_FONT);
-
- int f_sizes[7];
if (size == -1)
- size = defaultFont.GetPointSize();
+ size = wxGetDefaultHTMLFontSize();
- f_sizes[0] = int(size * 0.6);
- f_sizes[1] = int(size * 0.8);
- f_sizes[2] = size;
- f_sizes[3] = int(size * 1.2);
- f_sizes[4] = int(size * 1.4);
- f_sizes[5] = int(size * 1.6);
- f_sizes[6] = int(size * 1.8);
+ int f_sizes[7];
+ wxBuildFontSizes(f_sizes, size);
- wxString normal = normal_face.empty() ?
- defaultFont.GetFaceName() : normal_face;
+ wxString normal = normal_face;
+ if ( normal.empty() )
+ normal = wxNORMAL_FONT->GetFaceName();
SetFonts(normal, fixed_face, f_sizes);
}
m_FontBold = m_FontItalic = m_FontUnderlined = m_FontFixed = FALSE;
m_FontSize = 3; //default one
CreateCurrentFont(); // we're selecting default font into
- m_DC->GetTextExtent( wxT("H"), &m_CharWidth, &m_CharHeight);
- /* NOTE : we're not using GetCharWidth/Height() because
- of differences under X and win
- */
+
+ // we're not using GetCharWidth/Height() because of
+ // differences under X and win
+ wxCoord w,h;
+ m_DC->GetTextExtent( wxT("H"), &w, &h);
+ m_CharWidth = w;
+ m_CharHeight = h;
m_UseLink = false;
m_Link = wxHtmlLinkInfo( wxEmptyString );
m_tmpLastWasSpace = false;
m_lastWordCell = NULL;
+ // open the toplevel container that contains everything else and that
+ // is never closed (this makes parser's life easier):
OpenContainer();
+
+ // then open the first container into which page's content will go:
OpenContainer();
#if !wxUSE_UNICODE
if ( status == wxHTML_BLOCK )
return NULL;
- return GetFS()->OpenFile(myurl);
+ int flags = wxFS_READ;
+ if (type == wxHTML_URL_IMAGE)
+ flags |= wxFS_SEEKABLE;
+
+ return GetFS()->OpenFile(myurl, flags);
}
-void wxHtmlWinParser::AddText(const wxChar* txt)
+#define NBSP_UNICODE_VALUE (wxChar(160))
+#if !wxUSE_UNICODE
+ #define CUR_NBSP_VALUE m_nbsp
+#else
+ #define CUR_NBSP_VALUE NBSP_UNICODE_VALUE
+#endif
+
+void wxHtmlWinParser::AddText(const wxString& txt)
{
- size_t i = 0,
- x,
- lng = wxStrlen(txt);
- register wxChar d;
- int templen = 0;
- wxChar nbsp = GetEntitiesParser()->GetCharForCode(160 /* nbsp */);
-
- if (lng+1 > m_tmpStrBufSize)
- {
- delete[] m_tmpStrBuf;
- m_tmpStrBuf = new wxChar[lng+1];
- m_tmpStrBufSize = lng+1;
- }
- wxChar *temp = m_tmpStrBuf;
+#if !wxUSE_UNICODE
+ if ( m_nbsp == 0 )
+ m_nbsp = GetEntitiesParser()->GetCharForCode(NBSP_UNICODE_VALUE);
+#endif
- if (m_tmpLastWasSpace)
+ if ( m_whitespaceMode == Whitespace_Normal )
{
- while ((i < lng) &&
- ((txt[i] == wxT('\n')) || (txt[i] == wxT('\r')) || (txt[i] == wxT(' ')) ||
- (txt[i] == wxT('\t')))) i++;
- }
+ int templen = 0;
- while (i < lng)
- {
- x = 0;
- d = temp[templen++] = txt[i];
- if ((d == wxT('\n')) || (d == wxT('\r')) || (d == wxT(' ')) || (d == wxT('\t')))
+ size_t lng = txt.length();
+ if (lng+1 > m_tmpStrBufSize)
{
- i++, x++;
- while ((i < lng) && ((txt[i] == wxT('\n')) || (txt[i] == wxT('\r')) ||
- (txt[i] == wxT(' ')) || (txt[i] == wxT('\t')))) i++, x++;
+ delete[] m_tmpStrBuf;
+ m_tmpStrBuf = new wxChar[lng+1];
+ m_tmpStrBufSize = lng+1;
}
- else i++;
+ wxChar *temp = m_tmpStrBuf;
- if (x)
+ wxString::const_iterator i = txt.begin();
+ const wxString::const_iterator end = txt.end();
+
+ if (m_tmpLastWasSpace)
{
- temp[templen-1] = wxT(' ');
- DoAddText(temp, templen, nbsp);
- m_tmpLastWasSpace = true;
+ while ( (i < end) &&
+ (*i == wxT('\n') || *i == wxT('\r') || *i == wxT(' ') ||
+ *i == wxT('\t')) )
+ {
+ ++i;
+ }
}
- }
- if (templen && (templen > 1 || temp[0] != wxT(' ')))
+ while (i < end)
+ {
+ size_t x = 0;
+ const wxChar d = temp[templen++] = *i;
+ if ((d == wxT('\n')) || (d == wxT('\r')) || (d == wxT(' ')) || (d == wxT('\t')))
+ {
+ ++i, ++x;
+ while ( (i < end) &&
+ (*i == wxT('\n') || *i == wxT('\r') ||
+ *i == wxT(' ') || *i == wxT('\t')) )
+ {
+ ++i;
+ ++x;
+ }
+ }
+ else
+ {
+ ++i;
+ }
+
+ if (x)
+ {
+ temp[templen-1] = wxT(' ');
+ FlushWordBuf(temp, templen);
+ m_tmpLastWasSpace = true;
+ }
+ }
+
+ if (templen && (templen > 1 || temp[0] != wxT(' ')))
+ {
+ FlushWordBuf(temp, templen);
+ m_tmpLastWasSpace = false;
+ }
+ }
+ else // m_whitespaceMode == Whitespace_Pre
{
- DoAddText(temp, templen, nbsp);
+ if ( txt.find(CUR_NBSP_VALUE) != wxString::npos )
+ {
+ // we need to substitute spaces for here just like we
+ // did in the Whitespace_Normal branch above
+ wxString txt2(txt);
+ txt2.Replace(CUR_NBSP_VALUE, ' ');
+ AddPreBlock(txt2);
+ }
+ else
+ {
+ AddPreBlock(txt);
+ }
+
+ // don't eat any whitespace in <pre> block
m_tmpLastWasSpace = false;
}
}
-void wxHtmlWinParser::DoAddText(wxChar *temp, int& templen, wxChar nbsp)
+void wxHtmlWinParser::FlushWordBuf(wxChar *buf, int& len)
{
- temp[templen] = 0;
- templen = 0;
+ buf[len] = 0;
+
+ for ( int i = 0; i < len; i++ )
+ {
+ if ( buf[i] == CUR_NBSP_VALUE )
+ buf[i] = ' ';
+ }
+
#if !wxUSE_UNICODE
if (m_EncConv)
- m_EncConv->Convert(temp);
+ m_EncConv->Convert(buf);
#endif
- size_t len = wxStrlen(temp);
- for (size_t j = 0; j < len; j++)
- {
- if (temp[j] == nbsp)
- temp[j] = wxT(' ');
- }
- wxHtmlCell *c = new wxHtmlWordCell(temp, *(GetDC()));
+ AddWord(wxString(buf, len));
+
+ len = 0;
+}
- ApplyStateToCell(c);
+void wxHtmlWinParser::AddWord(wxHtmlWordCell *word)
+{
+ ApplyStateToCell(word);
- m_Container->InsertCell(c);
- ((wxHtmlWordCell*)c)->SetPreviousWord(m_lastWordCell);
- m_lastWordCell = (wxHtmlWordCell*)c;
+ m_Container->InsertCell(word);
+ word->SetPreviousWord(m_lastWordCell);
+ m_lastWordCell = word;
}
+void wxHtmlWinParser::AddPreBlock(const wxString& text)
+{
+ if ( text.find('\t') != wxString::npos )
+ {
+ wxString text2;
+ text2.reserve(text.length());
+
+ const wxString::const_iterator end = text.end();
+ wxString::const_iterator copyFrom = text.begin();
+ size_t pos = 0;
+ int posColumn = m_posColumn;
+ for ( wxString::const_iterator i = copyFrom; i != end; ++i, ++pos )
+ {
+ if ( *i == '\t' )
+ {
+ if ( copyFrom != i )
+ text2.append(copyFrom, i);
+
+ const unsigned SPACES_PER_TAB = 8;
+ const size_t expandTo = SPACES_PER_TAB - posColumn % SPACES_PER_TAB;
+ text2.append(expandTo, ' ');
+
+ posColumn += expandTo;
+ copyFrom = i + 1;
+ }
+ else
+ {
+ ++posColumn;
+ }
+ }
+ if ( copyFrom != text.end() )
+ text2.append(copyFrom, text.end());
+
+ AddWord(new wxHtmlWordWithTabsCell(text2, text, m_posColumn, *(GetDC())));
+
+ m_posColumn = posColumn;
+ }
+ else
+ {
+ // no special formatting needed
+ AddWord(text);
+ m_posColumn += text.length();
+ }
+}
wxHtmlContainerCell* wxHtmlWinParser::OpenContainer()
{
m_Container = new wxHtmlContainerCell(m_Container);
m_Container->SetAlignHor(m_Align);
+ m_posColumn = 0;
m_tmpLastWasSpace = true;
/* to avoid space being first character in paragraph */
return m_Container;
void wxHtmlWinParser::SetFontSize(int s)
{
- if (s < 1) s = 1;
- else if (s > 7) s = 7;
+ if (s < 1)
+ s = 1;
+ else if (s > 7)
+ s = 7;
m_FontSize = s;
}
+void wxHtmlWinParser::SetDC(wxDC *dc, double pixel_scale, double font_scale)
+{
+ m_DC = dc;
+ m_PixelScale = pixel_scale;
+ m_FontScale = font_scale;
+}
+
wxFont* wxHtmlWinParser::CreateCurrentFont()
{
{
*faceptr = face;
*fontptr = new wxFont(
- (int) (m_FontsSizes[fs] * m_PixelScale),
+ (int) (m_FontsSizes[fs] * m_FontScale),
ff ? wxMODERN : wxSWISS,
fi ? wxITALIC : wxNORMAL,
fb ? wxBOLD : wxNORMAL,
void wxHtmlWinParser::SetFontFace(const wxString& face)
{
- if (GetFontFixed()) m_FontFaceFixed = face;
- else m_FontFaceNormal = face;
+ if (GetFontFixed())
+ m_FontFaceFixed = face;
+ else
+ m_FontFaceNormal = face;
#if !wxUSE_UNICODE
if (m_InputEnc != wxFONTENCODING_DEFAULT)
#if !wxUSE_UNICODE
void wxHtmlWinParser::SetInputEncoding(wxFontEncoding enc)
{
+ // the character used for non-breakable space may change:
+ m_nbsp = 0;
+
m_InputEnc = m_OutputEnc = wxFONTENCODING_DEFAULT;
if (m_EncConv)
{
m_EncConv = NULL;
}
- if (enc == wxFONTENCODING_DEFAULT) return;
+ if (enc == wxFONTENCODING_DEFAULT)
+ return;
wxFontEncoding altfix, altnorm;
bool availfix, availnorm;
- // exact match?
availnorm = wxFontMapper::Get()->IsEncodingAvailable(enc, m_FontFaceNormal);
availfix = wxFontMapper::Get()->IsEncodingAvailable(enc, m_FontFaceFixed);
+
if (availnorm && availfix)
+ {
+ // exact match?
m_OutputEnc = enc;
+ }
- // alternatives?
else if (wxFontMapper::Get()->GetAltForEncoding(enc, &altnorm, m_FontFaceNormal, false) &&
wxFontMapper::Get()->GetAltForEncoding(enc, &altfix, m_FontFaceFixed, false) &&
altnorm == altfix)
+ {
+ // alternatives?
m_OutputEnc = altnorm;
-
- // at least normal face?
+ }
else if (availnorm)
+ {
+ // at least normal face?
m_OutputEnc = enc;
+ }
else if (wxFontMapper::Get()->GetAltForEncoding(enc, &altnorm, m_FontFaceNormal, false))
+ {
m_OutputEnc = altnorm;
-
+ }
else
{
#ifndef __WXMAC__
- // okay, let convert to ISO_8859-1, available always
+ // okay, let's convert to ISO_8859-1, available always
m_OutputEnc = wxFONTENCODING_DEFAULT;
#else
m_OutputEnc = wxLocale::GetSystemEncoding() ;
m_InputEnc = enc;
if (m_OutputEnc == wxFONTENCODING_DEFAULT)
+ {
GetEntitiesParser()->SetEncoding(wxFONTENCODING_SYSTEM);
+ }
else
+ {
GetEntitiesParser()->SetEncoding(m_OutputEnc);
+ }
- if (m_InputEnc == m_OutputEnc) return;
+ if (m_InputEnc == m_OutputEnc)
+ return;
m_EncConv = new wxEncodingConverter();
if (!m_EncConv->Init(m_InputEnc,
(m_OutputEnc == wxFONTENCODING_DEFAULT) ?
wxFONTENCODING_ISO8859_1 : m_OutputEnc,
wxCONVERT_SUBSTITUTE))
- { // total failture :-(
+ { // total failure :-(
wxLogError(_("Failed to display HTML document in %s encoding"),
wxFontMapper::GetEncodingName(enc).c_str());
m_InputEnc = m_OutputEnc = wxFONTENCODING_DEFAULT;
}
#endif
-