void EpubFormatter::HandleHtmlTag(HtmlToken *t)
{
CrashIf(!t->IsTag());
if (hiddenDepth && t->IsEndTag() && tagNesting.Count() == hiddenDepth &&
t->tag == tagNesting.Last()) {
hiddenDepth = 0;
UpdateTagNesting(t);
return;
}
if (0 == hiddenDepth && t->IsStartTag() && t->GetAttrByName("hidden"))
hiddenDepth = tagNesting.Count() + 1;
if (hiddenDepth > 0)
UpdateTagNesting(t);
else if (Tag_Image == t->tag || Tag_Svg_Image == t->tag)
HandleTagSvgImage(t);
else
HtmlFormatter::HandleHtmlTag(t);
}
void HtmlFormatter::HandleTagStyle(HtmlToken* t) {
if (!t->IsStartTag())
return;
AttrInfo* attr = t->GetAttrByName("type");
if (attr && !attr->ValIs("text/css"))
return;
const char* start = t->s + t->sLen + 1;
while (t && !t->IsError() && (!t->IsEndTag() || t->tag != Tag_Style)) {
t = htmlParser->Next();
}
if (!t || !t->IsEndTag() || Tag_Style != t->tag)
return;
const char* end = t->s - 2;
CrashIf(start > end);
ParseStyleSheet(start, end - start);
UpdateTagNesting(t);
}
// Returns next part of html or NULL if finished
HtmlToken *HtmlPullParser::Next()
{
if (currPos >= end)
return NULL;
Next:
const char *start = currPos;
if (*currPos != '<') {
// this must text between tags
if (!SkipUntil(currPos, end, '<') && IsSpaceOnly(start, currPos)) {
// ignore whitespace after the last tag
return NULL;
}
currToken.SetText(start, currPos);
return &currToken;
}
// '<' - tag begins
++start;
// skip <? and <! (processing instructions and comments)
if (('?' == *start) || ('!' == *start)) {
if ('!' == *start && start + 2 < end && str::StartsWith(start, "!--")) {
currPos = start + 2;
if (!SkipUntil(currPos, end, "-->")) {
currToken.SetError(HtmlToken::UnclosedTag, start);
return &currToken;
}
currPos += 2;
}
else if (!SkipUntil(currPos, end, '>')) {
currToken.SetError(HtmlToken::UnclosedTag, start);
return &currToken;
}
++currPos;
goto Next;
}
if (!SkipUntilTagEnd(currPos, end)) {
currToken.SetError(HtmlToken::UnclosedTag, start);
return &currToken;
}
CrashIf('>' != *currPos);
if (currPos == start || currPos == start + 1 && *start == '/') {
// skip empty tags (<> and </>), because we're lenient
++currPos;
goto Next;
}
if (('/' == *start) && ('/' == currPos[-1])) { // </foo/>
currToken.SetError(HtmlToken::InvalidTag, start);
} else if ('/' == *start) { // </foo>
currToken.SetTag(HtmlToken::EndTag, start + 1, currPos);
} else if ('/' == currPos[-1]) { // <foo/>
currToken.SetTag(HtmlToken::EmptyElementTag, start, currPos - 1);
} else {
currToken.SetTag(HtmlToken::StartTag, start, currPos);
}
++currPos;
UpdateTagNesting(&tagNesting, &currToken);
return &currToken;
}
void HtmlFormatter::HandleHtmlTag(HtmlToken* t) {
CrashIf(!t->IsTag());
UpdateTagNesting(t);
HtmlTag tag = t->tag;
if (Tag_P == tag) {
HandleTagP(t);
} else if (Tag_Hr == tag) {
EmitHr();
} else if ((Tag_B == tag) || (Tag_Strong == tag)) {
ChangeFontStyle(FontStyleBold, t->IsStartTag());
} else if ((Tag_I == tag) || (Tag_Em == tag)) {
ChangeFontStyle(FontStyleItalic, t->IsStartTag());
} else if (Tag_U == tag) {
if (!currLinkIdx)
ChangeFontStyle(FontStyleUnderline, t->IsStartTag());
} else if (Tag_Strike == tag) {
ChangeFontStyle(FontStyleStrikeout, t->IsStartTag());
} else if (Tag_Br == tag) {
HandleTagBr();
} else if (Tag_Font == tag) {
HandleTagFont(t);
} else if (Tag_A == tag) {
HandleTagA(t);
} else if (Tag_Blockquote == tag) {
// TODO: implement me
HandleTagList(t);
} else if (Tag_Div == tag) {
// TODO: implement me
HandleTagP(t, true);
} else if (IsTagH(tag)) {
HandleTagHx(t);
} else if (Tag_Sup == tag) {
// TODO: implement me
} else if (Tag_Sub == tag) {
// TODO: implement me
} else if (Tag_Span == tag) {
// TODO: implement me
} else if (Tag_Center == tag) {
HandleTagP(t, true);
if (!t->IsEndTag())
CurrStyle()->align = Align_Center;
} else if ((Tag_Ul == tag) || (Tag_Ol == tag)) {
HandleTagList(t);
} else if (Tag_Li == tag) {
// TODO: display bullet/number
FlushCurrLine(true);
} else if (Tag_Dt == tag) {
FlushCurrLine(true);
ChangeFontStyle(FontStyleBold, t->IsStartTag());
if (t->IsStartTag())
CurrStyle()->align = Align_Left;
} else if (Tag_Dd == tag) {
// TODO: separate indentation from list depth
HandleTagList(t);
} else if (Tag_Table == tag) {
// TODO: implement me
HandleTagList(t);
} else if (Tag_Tr == tag) {
// display tables row-by-row for now
FlushCurrLine(true);
if (t->IsStartTag())
SetAlignment(Align_Left);
else if (t->IsEndTag())
RevertStyleChange();
} else if (Tag_Code == tag || Tag_Tt == tag) {
if (t->IsStartTag())
SetFont(L"Courier New", (FontStyle)CurrFont()->GetStyle());
else if (t->IsEndTag())
RevertStyleChange();
} else if (Tag_Pre == tag) {
HandleTagPre(t);
} else if (Tag_Img == tag) {
HandleTagImg(t);
} else if (Tag_Pagebreak == tag) {
// not really a HTML tag, but many ebook
// formats use it
HandleTagPagebreak(t);
} else if (Tag_Link == tag) {
HandleTagLink(t);
} else if (Tag_Style == tag) {
HandleTagStyle(t);
} else {
// TODO: temporary debugging
// lf("unhandled tag: %d", tag);
}
// any tag could contain anchor information
HandleAnchorAttr(t);
// any tag could contain a reading direction change
HandleDirAttr(t);
}
