void HtmlFormatter::HandleDirAttr(HtmlToken* t) {
// only apply reading direction changes to block elements (for now)
if (t->IsStartTag() && !IsInlineTag(t->tag)) {
AttrInfo* attr = t->GetAttrByName("dir");
if (attr)
dirRtl = CurrStyle()->dirRtl = attr->ValIs("RTL");
}
}
AttrInfo *HtmlToken::GetAttrByNameNS(const char *name, const char *attrNS)
{
nextAttr = NULL; // start from the beginning
for (AttrInfo *a = NextAttr(); a; a = NextAttr()) {
if (a->NameIsNS(name, attrNS))
return a;
}
return NULL;
}
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);
}
static void Test00(const char *s, HtmlToken::TokenType expectedType) {
HtmlPullParser parser(s, str::Len(s));
HtmlToken *t = parser.Next();
assert(t->type == expectedType);
assert(t->NameIs("p"));
assert(Tag_P == t->tag);
AttrInfo *a = t->GetAttrByName("a1");
assert(a->NameIs("a1"));
assert(a->ValIs(">"));
a = t->GetAttrByName("foo");
assert(a->NameIs("foo"));
assert(a->ValIs("bar"));
a = t->GetAttrByName("nope");
assert(!a);
t = parser.Next();
assert(!t);
}
void HtmlFileFormatter::HandleTagLink(HtmlToken *t)
{
CrashIf(!htmlDoc);
if (t->IsEndTag())
return;
AttrInfo *attr = t->GetAttrByName("rel");
if (!attr || !attr->ValIs("stylesheet"))
return;
attr = t->GetAttrByName("type");
if (attr && !attr->ValIs("text/css"))
return;
attr = t->GetAttrByName("href");
if (!attr)
return;
size_t len;
ScopedMem<char> src(str::DupN(attr->val, attr->valLen));
ScopedMem<char> data(htmlDoc->GetFileData(src, &len));
if (data)
ParseStyleSheet(data, len);
}
// cf. http://www.w3.org/TR/html4/charset.html#h-5.2.2
static UINT ExtractHttpCharset(const char *html, size_t htmlLen)
{
if (!strstr(html, "charset="))
return 0;
HtmlPullParser parser(html, min(htmlLen, 1024));
HtmlToken *tok;
while ((tok = parser.Next()) && !tok->IsError()) {
if (tok->tag != Tag_Meta)
continue;
AttrInfo *attr = tok->GetAttrByName("http-equiv");
if (!attr || !attr->ValIs("Content-Type"))
continue;
attr = tok->GetAttrByName("content");
ScopedMem<char> mimetype, charset;
if (!attr || !str::Parse(attr->val, attr->valLen, "%S;%_charset=%S", &mimetype, &charset))
continue;
static struct {
const char *name;
UINT codepage;
} codepages[] = {
{ "ISO-8859-1", 1252 }, { "Latin1", 1252 }, { "CP1252", 1252 }, { "Windows-1252", 1252 },
{ "ISO-8859-2", 28592 }, { "Latin2", 28592 },
{ "CP1251", 1251 }, { "Windows-1251", 1251 }, { "KOI8-R", 20866 },
{ "shift-jis", 932 }, { "x-euc", 932 }, { "euc-kr", 949 },
{ "Big5", 950 }, { "GB2312", 936 },
{ "UTF-8", CP_UTF8 },
};
for (int i = 0; i < dimof(codepages); i++) {
if (str::EqI(charset, codepages[i].name))
return codepages[i].codepage;
}
break;
}
return 0;
}
