static void WalkChmTocOrIndex(EbookTocVisitor *visitor, HtmlElement *list, UINT cp, bool isIndex, int level=1)
{
CrashIf(Tag_Ul != list->tag);
// some broken ToCs wrap every <li> into its own <ul>
for (; list && Tag_Ul == list->tag; list = list->next) {
for (HtmlElement *el = list->down; el; el = el->next) {
if (Tag_Li != el->tag)
continue; // ignore unexpected elements
bool valid;
HtmlElement *elObj = el->GetChildByTag(Tag_Object);
if (!elObj)
valid = false;
else if (isIndex)
valid = VisitChmIndexItem(visitor, elObj, cp, level);
else
valid = VisitChmTocItem(visitor, elObj, cp, level);
if (!valid)
continue; // skip incomplete elements and all their children
HtmlElement *nested = el->GetChildByTag(Tag_Ul);
// some broken ToCs have the <ul> follow right *after* a <li>
if (!nested && el->next && Tag_Ul == el->next->tag)
nested = el->next;
if (nested)
WalkChmTocOrIndex(visitor, nested, cp, isIndex, level + 1);
}
}
}
bool ChmDoc::ParseTocOrIndex(EbookTocVisitor *visitor, const char *path, bool isIndex)
{
if (!path)
return false;
// TODO: is path already UTF-8 encoded - or do we need str::conv::ToUtf8(ToStr(path)) ?
ScopedMem<unsigned char> htmlData(GetData(path, NULL));
const char *html = (char *)htmlData.Get();
if (!html)
return false;
HtmlParser p;
UINT cp = codepage;
// detect UTF-8 content by BOM
if (str::StartsWith(html, UTF8_BOM)) {
html += 3;
cp = CP_UTF8;
}
// enforce the default codepage, so that pre-encoded text and
// entities are in the same codepage and VisitChmTocItem yields
// consistent results
HtmlElement *el = p.Parse(html, CP_CHM_DEFAULT);
if (!el)
return false;
el = p.FindElementByName("body");
// since <body> is optional, also continue without one
el = p.FindElementByName("ul", el);
if (!el)
return WalkBrokenChmTocOrIndex(visitor, p, cp, isIndex);
WalkChmTocOrIndex(visitor, el, cp, isIndex);
return true;
}
static void WalkChmTocOrIndex(EbookTocVisitor *visitor, HtmlElement *list, UINT cp, bool isIndex, int level=1)
{
CrashIf(!list->NameIs("ul"));
// some broken ToCs wrap every <li> into its own <ul>
for (; list && list->NameIs("ul"); list = list->next) {
for (HtmlElement *el = list->down; el; el = el->next) {
if (!el->NameIs("li"))
continue; // ignore unexpected elements
bool valid = (isIndex ? VisitChmIndexItem : VisitChmTocItem)(visitor, el, cp, level);
if (!valid)
continue; // skip incomplete elements and all their children
HtmlElement *nested = el->GetChildByName("ul");
// some broken ToCs have the <ul> follow right *after* a <li>
if (!nested && el->next && el->next->NameIs("ul"))
nested = el->next;
if (nested)
WalkChmTocOrIndex(visitor, nested, cp, isIndex, level + 1);
}
}
}
