package htmlquery import ( "compress/gzip" "fmt" "io/ioutil" "net/http" "net/http/httptest" "os" "strings" "sync" "testing" "github.com/antchfx/xpath" "golang.org/x/net/html" ) const htmlSample = ` Hello,World!

City Gallery

London

Mountain View

London is the capital city of England. It is the most populous city in the United Kingdom, with a metropolitan area of over 13 million inhabitants.

Standing on the River Thames, London has been a major settlement for two millennia, its history going back to its founding by the Romans, who named it Londinium.

` var testDoc = loadHTML(htmlSample) func BenchmarkSelectorCache(b *testing.B) { DisableSelectorCache = false for i := 0; i < b.N; i++ { getQuery("/AAA/BBB/DDD/CCC/EEE/ancestor::*") } } func BenchmarkDisableSelectorCache(b *testing.B) { DisableSelectorCache = true for i := 0; i < b.N; i++ { getQuery("/AAA/BBB/DDD/CCC/EEE/ancestor::*") } } func TestSelectorCache(t *testing.T) { SelectorCacheMaxEntries = 2 for i := 1; i <= 3; i++ { getQuery(fmt.Sprintf("//a[position()=%d]", i)) } getQuery("//a[position()=3]") } func TestLoadURL(t *testing.T) { ts := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { fmt.Fprint(w, htmlSample) })) defer ts.Close() _, err := LoadURL(ts.URL) if err != nil { t.Fatal(err) } } func TestLoadURLWithGzipResponse(t *testing.T) { ts := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { w.Header().Add("Content-Encoding", "gzip") gz := gzip.NewWriter(w) defer gz.Close() fmt.Fprint(gz, htmlSample) })) defer ts.Close() _, err := LoadURL(ts.URL) if err != nil { t.Fatal(err) } } func TestLoadDoc(t *testing.T) { tempHTMLdoc, err := ioutil.TempFile("", "sample_*.html") if err != nil { t.Fatal(err) } tempHTMLFilename := tempHTMLdoc.Name() defer func(tempHTMLdoc *os.File, filename string) { tempHTMLdoc.Close() os.Remove(filename) }(tempHTMLdoc, tempHTMLFilename) tempHTMLdoc.Write([]byte(htmlSample)) if _, err := LoadDoc(tempHTMLFilename); err != nil { t.Fatal(err) } } func TestNavigator(t *testing.T) { top := FindOne(testDoc, "//html") nav := &NodeNavigator{curr: top, root: top, attr: -1} nav.MoveToChild() // HEAD nav.MoveToNext() if nav.NodeType() != xpath.TextNode { t.Fatalf("expectd node type is TextNode,but got %vs", nav.NodeType()) } nav.MoveToNext() // if nav.Value() != InnerText(FindOne(testDoc, "//body")) { t.Fatal("body not equal") } nav.MoveToPrevious() // nav.MoveToParent() // if nav.curr != top { t.Fatal("current node is not html node") } nav.MoveToNextAttribute() if nav.LocalName() != "lang" { t.Fatal("node not move to lang attribute") } nav.MoveToParent() nav.MoveToFirst() // if nav.curr.Type != html.DoctypeNode { t.Fatalf("expected node type is DoctypeNode,but got %d", nav.curr.Type) } } func TestXPath(t *testing.T) { node := FindOne(testDoc, "//html") if SelectAttr(node, "lang") != "en-US" { t.Fatal("//html[@lang] != en-Us") } node = FindOne(testDoc, "//header") if strings.Index(InnerText(node), "Logo") > 0 { t.Fatal("InnerText() have comment node text") } if !strings.Contains(OutputHTML(node, true), "Logo") { t.Fatal("OutputHTML() shoud have comment node text") } link := FindOne(testDoc, "//a[1]/@href") if link == nil { t.Fatal("link is nil") } if v := InnerText(link); v != "/London" { t.Fatalf("expect value is /London, but got %s", v) } } func TestXPathCdUp(t *testing.T) { doc := loadHTML(``) node := FindOne(doc, "//b/@attr/..") t.Logf("node = %#v", node) if node == nil || node.Data != "b" { t.Fatal("//b/@id/.. != ") } } func TestConcurrentQuery(t *testing.T) { var wg sync.WaitGroup for i := 0; i < 10; i++ { wg.Add(1) go func(i int) { defer wg.Done() s := `
a
` doc := loadHTML(s) if n := FindOne(doc, `//div`); n == nil { t.Fatalf("should find one but got nil [%d]", i) } }(i) } wg.Wait() } func TestLoadURL_ErrorCases(t *testing.T) { tests := []struct { name string contentEncoding string responseBody []byte expectedError string }{ { name: "ZlibReaderError", contentEncoding: "deflate", responseBody: []byte("invalid zlib data"), expectedError: "zlib", }, { name: "EmptyContentEncodingNilBody", contentEncoding: "", responseBody: nil, expectedError: "EOF", }, { name: "GzipReaderError", contentEncoding: "gzip", responseBody: []byte("invalid gzip data"), expectedError: "gzip", }, } for _, tt := range tests { t.Run(tt.name, func(t *testing.T) { // Create a local HTTP server server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { w.Header().Set("Content-Encoding", tt.contentEncoding) w.WriteHeader(http.StatusOK) w.Write(tt.responseBody) })) defer server.Close() node, err := LoadURL(server.URL) // Assert that the returned node is nil if node != nil { t.Errorf("Expected node to be nil, but got %v", node) } // Assert that an error is returned if err == nil { t.Error("Expected an error, but got nil") } else if !strings.Contains(err.Error(), tt.expectedError) { t.Errorf("Expected error to contain '%s', but got: %v", tt.expectedError, err) } }) } } func loadHTML(str string) *html.Node { node, err := Parse(strings.NewReader(str)) if err != nil { panic(err) } return node }