static void HtmlParser02()
{
HtmlParser p;
HtmlElement *root = p.Parse("<a><b/><c></c ><d at1=\"<quo&ted>\" at2='also quoted' att3=notquoted att4=end/></a>");
utassert(4 == p.ElementsCount());
utassert(4 == p.TotalAttrCount());
utassert(root->NameIs("a"));
utassert(NULL == root->next);
HtmlElement *el = root->down;
utassert(el->NameIs("b"));
utassert(root == el->up);
el = el->next;
utassert(el->NameIs("c"));
utassert(root == el->up);
el = el->next;
utassert(el->NameIs("d"));
utassert(NULL == el->next);
utassert(root == el->up);
ScopedMem<WCHAR> val(el->GetAttribute("at1"));
utassert(str::Eq(val, L"<quo&ted>"));
val.Set(el->GetAttribute("at2"));
utassert(str::Eq(val, L"also quoted"));
val.Set(el->GetAttribute("att3"));
utassert(str::Eq(val, L"notquoted"));
val.Set(el->GetAttribute("att4"));
utassert(str::Eq(val, L"end"));
}
static void HtmlParser05()
{
HtmlParser p;
HtmlElement *root = p.Parse("<!doctype><html><HEAD><meta name=foo></head><body><object t=la><param name=foo val=bar></object><ul><li></ul></object></body></Html>");
utassert(8 == p.ElementsCount());
utassert(4 == p.TotalAttrCount());
utassert(root->NameIs("html"));
utassert(NULL == root->up);
utassert(NULL == root->next);
HtmlElement *el = root->down;
utassert(el->NameIs("head"));
HtmlElement *el2 = el->down;
utassert(el2->NameIs("meta"));
utassert(NULL == el2->next);
utassert(NULL == el2->down);
el2 = el->next;
utassert(el2->NameIs("body"));
utassert(NULL == el2->next);
el2 = el2->down;
utassert(el2->NameIs("object"));
el = p.FindElementByName("html");
utassert(el);
el = p.FindElementByName("head", el);
utassert(el);
utassert(el->NameIs("head"));
el = p.FindElementByName("ul", el);
utassert(el);
}
HtmlElement *HtmlParser::FindParent(char *tagName)
{
if (str::Eq(tagName, "li")) {
// make a list item the child of the closest list
for (HtmlElement *el = currElement; el; el = el->up) {
if (el->NameIs("ul") || el->NameIs("ol"))
return el;
}
}
return currElement;
}
static void HtmlParser09()
{
HtmlParser p;
HtmlElement *root = p.Parse("<?xml version='1.0'?><!-- <html><body></html> --><root attr='<!-- comment -->' />");
utassert(1 == p.ElementsCount());
utassert(1 == p.TotalAttrCount());
utassert(root->NameIs("root"));
ScopedMem<WCHAR> val(root->GetAttribute("attr"));
utassert(str::Eq(val, L"<!-- comment -->"));
root = p.Parse("<!-- comment with \" and \' --><main />");
utassert(1 == p.ElementsCount());
utassert(0 == p.TotalAttrCount());
utassert(root->NameIs("main"));
}
static void HtmlParser00()
{
HtmlParser p;
HtmlElement *root = p.Parse("<a></A>");
utassert(p.ElementsCount() == 1);
utassert(root);
utassert(Tag_A == root->tag && !root->name);
utassert(root->NameIs("a"));
root = p.Parse("<b></B>");
utassert(p.ElementsCount() == 1);
utassert(root);
utassert(Tag_B == root->tag && !root->name);
utassert(root->NameIs("b"));
}
HtmlElement *HtmlParser::FindElementByNameNS(const char *name, const char *ns, HtmlElement *from)
{
HtmlElement *el = from ? from : rootElement;
if (from)
goto FindNext;
if (!el)
return NULL;
CheckNext:
if (el->NameIs(name) || ns && el->NameIsNS(name, ns))
return el;
FindNext:
if (el->down) {
el = el->down;
goto CheckNext;
}
if (el->next) {
el = el->next;
goto CheckNext;
}
// backup in the tree
HtmlElement *parent = el->up;
while (parent) {
if (parent->next) {
el = parent->next;
goto CheckNext;
}
parent = parent->up;
}
return NULL;
}
static void HtmlParser01()
{
HtmlParser p;
HtmlElement *root = p.Parse("<A><bAh></a>");
utassert(p.ElementsCount() == 2);
utassert(Tag_A == root->tag && !root->name);
utassert(NULL == root->up);
utassert(NULL == root->next);
HtmlElement *el = root->down;
utassert(NULL == el->firstAttr);
utassert(el->NameIs("bah") && el->NameIs("BAH"));
utassert(Tag_NotFound == el->tag && str::Eq("bAh", el->name));
utassert(el->up == root);
utassert(NULL == el->down);
utassert(NULL == el->next);
}
static void HtmlParser10()
{
HtmlParser p;
HtmlElement *root = p.Parse("<!xml version='1.0'?><x:a xmlns:x='http://example.org/ns/x'><x:b attr='val'/></x:a>");
utassert(2 == p.ElementsCount());
utassert(2 == p.TotalAttrCount());
utassert(root->NameIs("x:a") && root->NameIsNS("a", "http://example.org/ns/x"));
HtmlElement *node = p.FindElementByName("b");
utassert(!node);
node = p.FindElementByNameNS("b", "http://example.org/ns/x");
utassert(node);
utassert(node->NameIs("x:b") && node->NameIsNS("b", "http://example.org/ns/x"));
ScopedMem<WCHAR> val(node->GetAttribute("attr"));
utassert(str::Eq(val, L"val"));
// TODO: XML tags are case sensitive (HTML tags aren't)
node = p.FindElementByName("X:B");
utassert(node && node->NameIs("X:B"));
}
HtmlElement *HtmlElement::GetChildByName(const char *name, int idx) const
{
for (HtmlElement *el = down; el; el = el->next) {
if (el->NameIs(name)) {
if (0 == idx)
return el;
idx--;
}
}
return NULL;
}
static void HtmlParser11()
{
HtmlParser p;
HtmlElement *root = p.Parse("<root/><!-- comment -->");
utassert(1 == p.ElementsCount());
utassert(0 == p.TotalAttrCount());
utassert(root && root->NameIs("root"));
root = p.Parse("<root><!---></root>");
utassert(!root);
}
static void HtmlParserFile()
{
WCHAR *fileName = L"HtmlParseTest00.html";
// We assume we're being run from obj-[dbg|rel], so the test
// files are in ..\src\utils directory relative to exe's dir
ScopedMem<WCHAR> exePath(GetExePath());
const WCHAR *exeDir = path::GetBaseName(exePath);
ScopedMem<WCHAR> p1(path::Join(exeDir, L"..\\src\\utils"));
ScopedMem<WCHAR> p2(path::Join(p1, fileName));
char *d = file::ReadAll(p2, NULL);
// it's ok if we fail - we assume we were not run from the
// right location
if (!d)
return;
HtmlParser p;
HtmlElement *root = p.ParseInPlace(d);
utassert(root);
utassert(709 == p.ElementsCount());
utassert(955 == p.TotalAttrCount());
utassert(root->NameIs("html"));
HtmlElement *el = root->down;
utassert(el->NameIs("head"));
el = el->next;
utassert(el->NameIs("body"));
el = el->down;
utassert(el->NameIs("object"));
el = el->next;
utassert(el->NameIs("ul"));
el = el->down;
utassert(el->NameIs("li"));
el = el->down;
utassert(el->NameIs("object"));
ScopedMem<WCHAR> val(el->GetAttribute("type"));
utassert(str::Eq(val, L"text/sitemap"));
el = el->down;
utassert(el->NameIs("param"));
utassert(!el->down);
utassert(el->next->NameIs("param"));
el = p.FindElementByName("body");
utassert(el);
el = p.FindElementByName("ul", el);
utassert(el);
int count = 0;
while (el) {
++count;
el = p.FindElementByName("ul", el);
}
utassert(18 == count);
free(d);
}
void HtmlParser::CloseTag(char *tagName)
{
str::ToLower(tagName);
// to allow for lack of closing tags, e.g. in case like
// <a><b><c></a>, we look for the first parent with matching name
for (HtmlElement *el = currElement; el; el = el->up) {
if (el->NameIs(tagName)) {
currElement = el->up;
return;
}
}
// ignore the unexpected closing tag
}
static void HtmlParser03()
{
HtmlParser p;
HtmlElement *root = p.Parse("<el att =v"al/>");
utassert(1 == p.ElementsCount());
utassert(1 == p.TotalAttrCount());
utassert(root->NameIs("el"));
utassert(NULL == root->next);
utassert(NULL == root->up);
utassert(NULL == root->down);
ScopedMem<WCHAR> val(root->GetAttribute("att"));
utassert(str::Eq(val, L"v\"al"));
utassert(!root->firstAttr->next);
}
void HtmlParser::CloseTag(HtmlToken *tok)
{
char *tagName = NULL;
if (Tag_NotFound == tok->tag) {
tagName = (char *)tok->s;
char *tagEnd = tagName + tok->nLen;
*tagEnd = '\0';
}
// to allow for lack of closing tags, e.g. in case like
// <a><b><c></a>, we look for the first parent with matching name
for (HtmlElement *el = currElement; el; el = el->up) {
if (tagName ? el->NameIs(tagName) : tok->tag == el->tag) {
currElement = el->up;
return;
}
}
// ignore the unexpected closing tag
}
static void HtmlParser06()
{
HtmlParser p;
HtmlElement *root = p.Parse("<ul><p>ignore<li><br><meta><li><ol><li></ul><dropme>");
utassert(9 == p.ElementsCount());
utassert(0 == p.TotalAttrCount());
utassert(root->NameIs("ul"));
utassert(!root->next);
HtmlElement *el = root->GetChildByTag(Tag_Li);
utassert(el);
utassert(el->down->NameIs("br"));
utassert(el->down->next->NameIs("meta"));
utassert(!el->down->next->next);
el = root->GetChildByTag(Tag_Li, 1);
utassert(el);
utassert(!el->next);
el = el->GetChildByTag(Tag_Ol);
utassert(!el->next);
utassert(el->down->NameIs("li"));
utassert(!el->down->down);
}
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);
}
}
}