zishuo/goquery
1
0
Fork 0
mirror of https://github.com/PuerkitoBio/goquery synced 2025-10-25 01:00:25 +08:00
Code Issues Packages Projects Releases Wiki Activity

Implement Single and SingleMatcher

Browse Source
This commit is contained in:
Martin Angers
2021-06-13 14:34:32 -04:00
parent e1f2d60505
commit 0126a1fd88
2 changed files with 76 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
example_test.go
View File

@@ -80,3 +80,31 @@ func ExampleNewDocumentFromReader_string() {
// Output: Header // Output: Header
} }
func ExampleSingle() {
html := `
<html>
<body>
<div>1</div>
<div>2</div>
<div>3</div>
</body>
</html>
`
doc, err := goquery.NewDocumentFromReader(strings.NewReader(html))
if err != nil {
log.Fatal(err)
}
// By default, the selector string selects all matching nodes
multiSel := doc.Find("div")
fmt.Println(multiSel.Text())
// Using goquery.Single, only the first match is selected
singleSel := doc.FindMatcher(goquery.Single("div"))
fmt.Println(singleSel.Text())
// Output:
// 123
// 1
}

49
type.go
View File

@@ -7,7 +7,6 @@ import (
"net/url" "net/url"
"github.com/andybalholm/cascadia" "github.com/andybalholm/cascadia"
"golang.org/x/net/html" "golang.org/x/net/html"
) )
@@ -122,6 +121,30 @@ type Matcher interface {
Filter([]*html.Node) []*html.Node Filter([]*html.Node) []*html.Node
} }
// Single compiles a selector string to a Matcher that stops after the first
// match is found.
//
// By default, Selection.Find and other functions that accept a selector string
// will use all matches corresponding to that selector. By using the Matcher
// returned by Single, at most the first match will be used.
func Single(selector string) Matcher {
return singleMatcher{compileMatcher(selector)}
}
// SingleMatcher returns a Matcher matches the same nodes as m, but that stops
// after the first match is found.
//
// By default, Selection.FindMatcher and other functions that accept a Matcher
// will use all corresponding matches. By using the Matcher returned by
// SingleMatcher, at most the first match will be used.
func SingleMatcher(m Matcher) Matcher {
if _, ok := m.(singleMatcher); ok {
// m is already a singleMatcher
return m
}
return singleMatcher{m}
}
// compileMatcher compiles the selector string s and returns // compileMatcher compiles the selector string s and returns
// the corresponding Matcher. If s is an invalid selector string, // the corresponding Matcher. If s is an invalid selector string,
// it returns a Matcher that fails all matches. // it returns a Matcher that fails all matches.
@@ -133,6 +156,30 @@ func compileMatcher(s string) Matcher {
return cs return cs
} }
type singleMatcher struct {
Matcher
}
func (m singleMatcher) MatchAll(n *html.Node) []*html.Node {
// Optimized version - stops finding at the first match (cascadia-compiled
// matchers all use this code path).
if mm, ok := m.Matcher.(interface{ MatchFirst(*html.Node) *html.Node }); ok {
node := mm.MatchFirst(n)
if node == nil {
return nil
}
return []*html.Node{node}
}
// Fallback version, for e.g. test mocks that don't provide the MatchFirst
// method.
nodes := m.Matcher.MatchAll(n)
if len(nodes) > 0 {
return nodes[:1:1]
}
return nil
}
// invalidMatcher is a Matcher that always fails to match. // invalidMatcher is a Matcher that always fails to match.
type invalidMatcher struct{} type invalidMatcher struct{}