func buildHTML(tokenizer *html.Tokenizer) (s string, err error) {
buf := new(bytes.Buffer)
bp := 0
if tag, _ := tokenizer.TagName(); string(tag) == "div" {
div := tokenizer.Raw()
buf.Write(div)
bp = len(div)
err = nextToken(tokenizer)
}
ep := bp
for err != io.EOF {
if err != nil && err != io.EOF {
return
}
ep = buf.Len()
b := tokenizer.Raw()
if _, err := buf.Write(b); err != nil {
return "", err
}
err = nextToken(tokenizer)
}
b := buf.Bytes()
if bp > 0 {
b = b[bp:ep]
}
return string(b), nil
}
func skipSubtreeIfUicRemove(z *html.Tokenizer, tt html.TokenType, tagName string, attrs []html.Attribute) bool {
_, foundRemoveTag := getAttr(attrs, UicRemove)
if !foundRemoveTag {
return false
}
if isSelfClosingTag(tagName, tt) {
return true
}
depth := 0
for {
tt := z.Next()
tag, _ := z.TagName()
switch {
case tt == html.ErrorToken:
return true
case tt == html.StartTagToken && !isSelfClosingTag(string(tag), tt):
depth++
case tt == html.EndTagToken:
depth--
if depth < 0 {
return true
}
}
}
}
func flushTagToken(htmlBuf *[]byte, tz *html.Tokenizer, url string) string {
*htmlBuf = append(*htmlBuf, '<')
tagName, hasAttr := tz.TagName()
*htmlBuf = append(*htmlBuf, tagName...)
if hasAttr {
for {
attrKey, attrValue, hasMore := tz.TagAttr()
*htmlBuf = append(*htmlBuf, ' ')
*htmlBuf = append(*htmlBuf, attrKey...)
*htmlBuf = append(*htmlBuf, '=', '"')
if tagAttrToProxy[string(tagName)][string(attrKey)] {
urlInAttr := string(attrValue)
*htmlBuf = append(*htmlBuf, []byte(GetProxiedUrl(urlInAttr, url))...)
} else {
*htmlBuf = append(*htmlBuf, attrValue...)
}
*htmlBuf = append(*htmlBuf, '"')
if !hasMore {
break
}
}
}
*htmlBuf = append(*htmlBuf, '>')
if string(tagName) == "head" {
*htmlBuf = append(*htmlBuf, []byte(getJsHookTag())...)
}
return string(tagName)
}
func (item *AnimeConventionItem) readResgiterNowurl(t *html.Tokenizer) {
t.Next()
if _, hasmore := t.TagName(); hasmore {
if key, val, _ := t.TagAttr(); strings.EqualFold(string(key), "href") {
item.registerNowURL = string(val)
}
}
}
func (item *AnimeConventionItem) readNameAndLink(t *html.Tokenizer) {
if label := t.Next(); label == html.StartTagToken {
_, hasmore := t.TagName()
if hasmore {
if key, val, _ := t.TagAttr(); strings.EqualFold(string(key), "href") {
item.siteURL = string(val)
}
}
}
if label := t.Next(); label == html.TextToken {
item.name = string(t.Text())
}
}
func parseTableX86(z *html.Tokenizer, in *Intrinsic) *Intrinsic {
in.Performance = make(map[string]Timing)
for {
tt := z.Next()
switch tt {
case html.ErrorToken:
return in
case html.StartTagToken, html.EndTagToken:
tn, _ := z.TagName()
tns := strings.ToLower(string(tn))
switch tns {
case "tr":
if tt == html.StartTagToken {
n := 0
p := Timing{}
for {
tt = z.Next()
tn, _ = z.TagName()
tns = strings.ToLower(string(tn))
if tt == html.EndTagToken && tns == "tr" {
break
}
if tt == html.StartTagToken && tns == "td" {
switch n {
case 0:
p.Arch = getText(z)
case 1:
p.Latency, _ = strconv.ParseFloat(getText(z), 64)
case 2:
p.Throughput, _ = strconv.ParseFloat(getText(z), 64)
in.Performance[p.Arch] = p
}
n++
}
}
} else {
panic("tr ended")
}
case "table":
if tt == html.EndTagToken {
return in
} else {
panic("table started")
}
}
}
}
}
func (item *AnimeConventionItem) Parse(t *html.Tokenizer) {
for {
label := t.Next()
switch label {
case html.ErrorToken:
fmt.Errorf("%v\n", t.Err())
return
case html.TextToken:
switch string(t.Text()) {
case "Advance Rates:":
//fmt.Println("rate")
item.readadvanceRate(t)
case "At-Door Rates:":
item.readatDoorRate(t)
}
case html.StartTagToken, html.EndTagToken, html.SelfClosingTagToken:
tag, hasmore := t.TagName()
if strings.EqualFold(string(tag), "big") {
item.readResgiterNowurl(t)
} else if hasmore {
key, val, hasmore := t.TagAttr()
if strings.EqualFold(string(key), "itemprop") {
//fmt.Println(string(val))
switch string(val) {
case "description":
item.readDescription(t)
case "latitude":
item.readLatitude(t)
case "longitude":
item.readLongitude(t)
case "startDate":
item.readStartDate(t)
case "endDate":
item.readEndDate(t)
case "location":
item.readLocation(t)
case "addressLocality":
item.readCity(t)
case "addressRegion":
item.readState(t)
case "addressCountry":
item.readCountry(t, hasmore)
case "name":
item.readNameAndLink(t)
}
}
}
}
}
}
func (item *AnimeConventionItem) readLocation(t *html.Tokenizer) {
for {
if label := t.Next(); label == html.StartTagToken {
_, hasmore := t.TagName()
if hasmore {
if _, val, _ := t.TagAttr(); strings.EqualFold(string(val), "name") {
break
}
}
}
}
if label := t.Next(); label == html.TextToken {
item.location = string(t.Text())
}
}
func (item *AnimeConventionItem) readRates(t *html.Tokenizer) string {
rates := ""
for {
label := t.Next()
if label == html.EndTagToken {
val, _ := t.TagName()
if strings.EqualFold(string(val), "p") {
break
}
}
if label == html.TextToken {
rates = strings.Join([]string{rates, string(t.Text())}, "\n")
}
}
return strings.TrimSpace(rates)
}
func parseFragment(z *html.Tokenizer) (f Fragment, dependencies []*FetchDefinition, err error) {
attrs := make([]html.Attribute, 0, 10)
dependencies = make([]*FetchDefinition, 0, 0)
buff := bytes.NewBuffer(nil)
forloop:
for {
tt := z.Next()
tag, _ := z.TagName()
raw := byteCopy(z.Raw()) // create a copy here, because readAttributes modifies z.Raw, if attributes contain an &
attrs = readAttributes(z, attrs)
switch {
case tt == html.ErrorToken:
if z.Err() != io.EOF {
return nil, nil, z.Err()
}
break forloop
case tt == html.StartTagToken || tt == html.SelfClosingTagToken:
if string(tag) == UicInclude {
if replaceTextStart, replaceTextEnd, err := getInclude(z, attrs); err != nil {
return nil, nil, err
} else {
fmt.Fprintf(buff, replaceTextStart)
// Enhancement: WriteOut sub tree, to allow alternative content
// for optional includes.
fmt.Fprintf(buff, replaceTextEnd)
continue
}
}
if skipSubtreeIfUicRemove(z, tt, string(tag), attrs) {
continue
}
case tt == html.EndTagToken:
if string(tag) == UicFragment || string(tag) == UicTail {
break forloop
}
}
buff.Write(raw)
}
return StringFragment(buff.String()), dependencies, nil
}
func (parser *HtmlContentParser) parseHead(z *html.Tokenizer, c *MemoryContent) error {
attrs := make([]html.Attribute, 0, 10)
headBuff := bytes.NewBuffer(nil)
forloop:
for {
tt := z.Next()
tag, _ := z.TagName()
raw := byteCopy(z.Raw()) // create a copy here, because readAttributes modifies z.Raw, if attributes contain an &
attrs = readAttributes(z, attrs)
switch {
case tt == html.ErrorToken:
if z.Err() != io.EOF {
return z.Err()
}
break forloop
case tt == html.StartTagToken || tt == html.SelfClosingTagToken:
if skipSubtreeIfUicRemove(z, tt, string(tag), attrs) {
continue
}
if string(tag) == "script" && attrHasValue(attrs, "type", ScriptTypeMeta) {
if err := parseMetaJson(z, c); err != nil {
return err
}
continue
}
case tt == html.EndTagToken:
if string(tag) == "head" {
break forloop
}
}
headBuff.Write(raw)
}
s := headBuff.String()
st := strings.Trim(s, " \n")
if len(st) > 0 {
c.head = StringFragment(st)
}
return nil
}
func parseMetaJson(z *html.Tokenizer, c *MemoryContent) error {
tt := z.Next()
if tt != html.TextToken {
return fmt.Errorf("expected text node for meta json, but found %v, (%s)", tt.String(), z.Raw())
}
bytes := z.Text()
err := json.Unmarshal(bytes, &c.meta)
if err != nil {
return fmt.Errorf("error while parsing json from meta json element: %v", err.Error())
}
tt = z.Next()
tag, _ := z.TagName()
if tt != html.EndTagToken || string(tag) != "script" {
return fmt.Errorf("Tag not properly ended. Expected </script>, but found %s", z.Raw())
}
return nil
}
func skipCompleteTag(z *html.Tokenizer, tagName string) error {
forloop:
for {
tt := z.Next()
tag, _ := z.TagName()
switch {
case tt == html.ErrorToken:
if z.Err() != io.EOF {
return z.Err()
}
break forloop
case tt == html.EndTagToken:
tagAsString := string(tag)
if tagAsString == tagName {
break forloop
}
}
}
return nil
}
func (list *AnimeConventionList) Parse(t *html.Tokenizer) {
for {
next := t.Next()
switch next {
case html.ErrorToken:
return
case html.TextToken:
//fmt.Println(string(t.Text()))
case html.StartTagToken, html.EndTagToken, html.SelfClosingTagToken:
_, hasmore := t.TagName()
if hasmore {
key, val, _ := t.TagAttr()
if strings.EqualFold(string(key), "href") && strings.HasPrefix(string(val), "/events/info.shtml") {
var item = &AnimeConventionItem{}
item.url = strings.Join([]string{"http://animecons.com", string(val)}, "")
list.taskNum++
go item.crawlInformation(&list.ConventionList)
//time.Sleep(100 * time.Millisecond)
}
}
}
}
}
func getTextR(z *html.Tokenizer) string {
r := ""
depth := 1
for {
tt := z.Next()
switch tt {
case html.ErrorToken:
panic(z.Err())
case html.TextToken:
r += string(z.Text())
case html.StartTagToken:
tn, _ := z.TagName()
tns := strings.ToLower(string(tn))
switch tns {
case "div":
r += "\r"
depth++
case "span":
r += "'"
depth++
}
case html.EndTagToken:
tn, _ := z.TagName()
tns := strings.ToLower(string(tn))
switch tns {
case "div":
depth--
case "span":
r += "'"
depth--
}
}
if depth == 0 {
return r
}
}
}
func extractTitleFromTree(z *html.Tokenizer) string {
depth := 0
for {
tt := z.Next()
switch tt {
case html.ErrorToken:
return ""
case html.TextToken:
if depth > 0 {
title := strings.TrimSpace(string(z.Text()))
lower := strings.ToLower(title)
if strings.HasPrefix(lower, "imgur") {
return ""
}
return title
}
case html.StartTagToken:
tn, _ := z.TagName()
if string(tn) == "title" {
depth++
}
}
}
}
func buildTokenArray(tokenizer *xhtml.Tokenizer) []tagInfo {
tokens := []tagInfo{}
for tt := tokenizer.Next(); tt != xhtml.ErrorToken; tt = tokenizer.Next() {
switch tt {
case xhtml.TextToken:
txt := string(tokenizer.Text())
if len(tokens) == 0 {
info := tagInfo{
raw: txt,
}
tokens = append(tokens, info)
}
tn, _ := tokenizer.TagName()
key, val, _ := tokenizer.TagAttr()
info := tagInfo{
tag: string(tn),
key: string(key),
val: string(val),
txt: txt,
}
tokens = append(tokens, info)
case xhtml.StartTagToken:
tn, _ := tokenizer.TagName()
key, val, _ := tokenizer.TagAttr()
info := tagInfo{
tag: string(tn),
key: string(key),
val: string(val),
}
tokens = append(tokens, info)
case xhtml.SelfClosingTagToken, xhtml.EndTagToken:
tn, _ := tokenizer.TagName()
key, val, _ := tokenizer.TagAttr()
info := tagInfo{
tag: string(tn),
key: string(key),
val: string(val),
closingTag: true,
}
tokens = append(tokens, info)
}
}
return tokens
}
// getTagName gets a tagName from tokenizer.
func getTagName(tokenizer *html.Tokenizer) string {
tagName, _ := tokenizer.TagName()
return string(tagName)
}
func (parser *HtmlContentParser) parseBody(z *html.Tokenizer, c *MemoryContent) error {
attrs := make([]html.Attribute, 0, 10)
bodyBuff := bytes.NewBuffer(nil)
attrs = readAttributes(z, attrs)
if len(attrs) > 0 {
c.bodyAttributes = StringFragment(joinAttrs(attrs))
}
forloop:
for {
tt := z.Next()
tag, _ := z.TagName()
raw := byteCopy(z.Raw()) // create a copy here, because readAttributes modifies z.Raw, if attributes contain an &
attrs = readAttributes(z, attrs)
switch {
case tt == html.ErrorToken:
if z.Err() != io.EOF {
return z.Err()
}
break forloop
case tt == html.StartTagToken || tt == html.SelfClosingTagToken:
if skipSubtreeIfUicRemove(z, tt, string(tag), attrs) {
continue
}
if string(tag) == UicFragment {
if f, deps, err := parseFragment(z); err != nil {
return err
} else {
c.body[getFragmentName(attrs)] = f
for _, dep := range deps {
c.requiredContent[dep.URL] = dep
}
}
continue
}
if string(tag) == UicTail {
if f, deps, err := parseFragment(z); err != nil {
return err
} else {
c.tail = f
for _, dep := range deps {
c.requiredContent[dep.URL] = dep
}
}
continue
}
if string(tag) == UicFetch {
if fd, err := getFetch(z, attrs); err != nil {
return err
} else {
c.requiredContent[fd.URL] = fd
continue
}
}
if string(tag) == UicInclude {
if replaceTextStart, replaceTextEnd, err := getInclude(z, attrs); err != nil {
return err
} else {
bodyBuff.WriteString(replaceTextStart)
// Enhancement: WriteOut sub tree, to allow alternative content
// for optional includes.
bodyBuff.WriteString(replaceTextEnd)
continue
}
}
case tt == html.EndTagToken:
if string(tag) == "body" {
break forloop
}
}
bodyBuff.Write(raw)
}
s := bodyBuff.String()
if _, defaultFragmentExists := c.body[""]; !defaultFragmentExists {
if st := strings.Trim(s, " \n"); len(st) > 0 {
c.body[""] = StringFragment(st)
}
}
return nil
}
func pullNode(tokens *html.Tokenizer, root *Markup) {
var node *Markup
for {
token := tokens.Next()
switch token {
case html.ErrorToken:
return
case html.TextToken, html.CommentToken, html.DoctypeToken:
text := strings.TrimSpace(string(tokens.Text()))
if text == "" {
continue
}
if token == html.CommentToken {
text = "<!--" + text + "-->"
}
if node != nil {
NewText(text).Apply(node)
continue
}
NewText(text).Apply(root)
continue
case html.StartTagToken, html.EndTagToken, html.SelfClosingTagToken:
if token == html.EndTagToken {
node = nil
return
}
tagName, hasAttr := tokens.TagName()
node = NewMarkup(string(tagName), token == html.SelfClosingTagToken)
node.Apply(root)
if hasAttr {
attrLoop:
for {
key, val, more := tokens.TagAttr()
NewAttr(string(key), string(val)).Apply(node)
if !more {
break attrLoop
}
}
}
if token == html.SelfClosingTagToken {
continue
}
pullNode(tokens, node)
}
}
}
