ensure itemscope always starts a new item

master
David M. Nesting 2020-10-07 16:30:44 -04:00
parent a5d3d8ae37
commit ca93c08d53
2 changed files with 76 additions and 95 deletions

View File

@ -121,8 +121,18 @@ func (p *Parser) Parse() (*Microdata, error) {
}) })
for _, node := range topLevelItemNodes { for _, node := range topLevelItemNodes {
item := NewItem() p.data.Items = append(p.data.Items, p.readItem(nil, node))
p.data.Items = append(p.data.Items, item) }
return p.data, nil
}
func (p *Parser) readItem(item *Item, node *html.Node) *Item {
var parent *Item
if _, exists := getAttr("itemscope", node); exists {
parent, item = item, NewItem()
if itemtypes, exists := getAttr("itemtype", node); exists { if itemtypes, exists := getAttr("itemtype", node); exists {
for _, itemtype := range strings.Split(strings.TrimSpace(itemtypes), " ") { for _, itemtype := range strings.Split(strings.TrimSpace(itemtypes), " ") {
itemtype = strings.TrimSpace(itemtype) itemtype = strings.TrimSpace(itemtype)
@ -143,53 +153,24 @@ func (p *Parser) Parse() (*Microdata, error) {
itemref = strings.TrimSpace(itemref) itemref = strings.TrimSpace(itemref)
if refnode, exists := p.identifiedNodes[itemref]; exists { if refnode, exists := p.identifiedNodes[itemref]; exists {
if refnode != node {
p.readItem(item, refnode) p.readItem(item, refnode)
} }
} }
} }
for child := node.FirstChild; child != nil; {
p.readItem(item, child)
child = child.NextSibling
} }
} }
return p.data, nil
}
func (p *Parser) readItem(item *Item, node *html.Node) {
if itemprop, exists := getAttr("itemprop", node); exists { if itemprop, exists := getAttr("itemprop", node); exists {
if _, exists := getAttr("itemscope", node); exists { if parent != nil {
subitem := NewItem() // an itemprop on an itemscope has value of the item created by the itemscope
if itemrefs, exists := getAttr("itemref", node); exists {
for _, itemref := range strings.Split(strings.TrimSpace(itemrefs), " ") {
itemref = strings.TrimSpace(itemref)
if refnode, exists := p.identifiedNodes[itemref]; exists {
if refnode != node {
p.readItem(subitem, refnode)
}
}
}
}
for child := node.FirstChild; child != nil; {
p.readItem(subitem, child)
child = child.NextSibling
}
for _, propertyName := range strings.Split(strings.TrimSpace(itemprop), " ") { for _, propertyName := range strings.Split(strings.TrimSpace(itemprop), " ") {
propertyName = strings.TrimSpace(propertyName) propertyName = strings.TrimSpace(propertyName)
if propertyName != "" { if propertyName != "" {
item.AddItem(propertyName, subitem) parent.AddItem(propertyName, item)
} }
} }
} else {
return
}
var propertyValue string var propertyValue string
switch node.DataAtom { switch node.DataAtom {
@ -241,7 +222,7 @@ func (p *Parser) readItem(item *Item, node *html.Node) {
} }
} }
} }
}
} }
for child := node.FirstChild; child != nil; { for child := node.FirstChild; child != nil; {
@ -249,6 +230,7 @@ func (p *Parser) readItem(item *Item, node *html.Node) {
child = child.NextSibling child = child.NextSibling
} }
return item
} }
func getAttr(name string, node *html.Node) (string, bool) { func getAttr(name string, node *html.Node) (string, bool) {

View File

@ -583,8 +583,9 @@ func TestSkipSelfReferencingItemref(t *testing.T) {
actual := ParseData(html, t) actual := ParseData(html, t)
child := NewItem() child := NewItem()
child.AddString("title", "Foo") child.AddType("http://data-vocabulary.org/Breadcrumb")
child.AddString("url", "http://example.com/foo/bar") child.AddString("url", "http://example.com/foo/bar")
child.AddString("title", "Foo")
item := NewItem() item := NewItem()
item.AddType("http://schema.org/WebPage") item.AddType("http://schema.org/WebPage")
@ -609,11 +610,9 @@ func TestPropertiesInContainedItem(t *testing.T) {
<div itemscope itemtype="http://schema.org/Person"> <div itemscope itemtype="http://schema.org/Person">
<meta itemprop="bar" content="bar value"> <meta itemprop="bar" content="bar value">
</div> </div>
<div itemscope itemtype="http://schema.org/Person" itemprop="author"> <div itemscope itemtype="http://schema.org/Person" itemprop="author">
<meta itemprop="baz" content="baz value"> <meta itemprop="baz" content="baz value">
</div> </div>
</body>` </body>`
actual := ParseData(html, t) actual := ParseData(html, t)