func parse2(z *html.Tokenizer) (*Schedule, error) {
schedule := &Schedule{}
currentDate := ""
for {
tt := z.Next()
switch tt {
case html.ErrorToken:
return schedule, nil
case html.StartTagToken:
t := z.Token()
if isTokenTagWithAttr("font", "class", "PageHeading", &t, z) {
z.Next()
currentDate = z.Token().Data
} else if isTokenTagWithAttr("tr", "bgcolor", "#ffffff", &t, z) || isTokenTagWithAttr("tr", "bgcolor", "#f5f5f5", &t, z) {
game, err := parseGame(currentDate, z)
if err != nil {
return nil, err
}
schedule.Games = append(schedule.Games, game)
}
}
}
}
func getMatchInfoTitle(z *html.Tokenizer) string {
eof := false
for !eof {
tt := z.Next()
switch {
case tt == html.ErrorToken:
eof = true
case tt == html.StartTagToken:
t := z.Token()
// Check if the token is a <title> tag
isTitle := t.Data == "title"
if isTitle {
z.Next()
// This is the title
return z.Token().Data
}
}
}
// If we reached here something went wrong :^(
Error.Printf("Could not get title...")
return ""
}
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 getQ(tknzer html.Tokenizer, ch chan string) {
tknzer.Next()
tknzer.Next()
tknzer.Next()
tknzer.Next()
ch <- string(tknzer.Text())
}
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 getMatchInfoBets(z *html.Tokenizer) (bets []*Bet) {
var bettor string
var item string
var statTrak bool
eof := false
for !eof {
tt := z.Next()
switch {
case tt == html.ErrorToken:
eof = true
case tt == html.StartTagToken:
t := z.Token()
isDiv := t.Data == "div"
isSpan := t.Data == "span"
if isSpan {
for _, a := range t.Attr {
if a.Key == "class" && a.Val == "user" {
z.Next()
z.Next()
t := z.Token()
bettor = strings.TrimSpace(t.Data)
}
}
}
if isDiv {
for _, a := range t.Attr {
if a.Key == "class" && strings.Contains(a.Val, "item") {
z.Next()
z.Next()
t = z.Token()
// Get StatTrak status
statTrak = strings.Contains(t.Attr[0].Val, "clreff")
if statTrak {
z.Next()
z.Next()
z.Next()
z.Next()
t = z.Token()
}
item = t.Attr[2].Val
thisBet := &Bet{bettor, item, statTrak}
bets = append(bets, thisBet)
}
}
}
}
}
return
}
func getText(z *html.Tokenizer) string {
tt := z.Next()
switch tt {
case html.ErrorToken:
panic(z.Err())
case html.TextToken:
return string(z.Text())
}
return ""
}
func (item *AnimeConventionItem) readCountry(t *html.Tokenizer, hasmore bool) {
if hasmore {
_, val, _ := t.TagAttr()
item.country = string(val)
return
}
t.Next()
item.country = string(t.Text())
}
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 advanceToTextToken(z *html.Tokenizer) *html.Token {
for {
tt := z.Next()
switch tt {
case html.ErrorToken:
return nil
case html.TextToken:
t := z.Token()
return &t
}
}
}
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 parseGame(date string, z *html.Tokenizer) (Game, error) {
var game Game
td := advanceToStartTag("td", z)
if td == nil {
return game, errors.New("Unable to find Game Number")
}
z.Next()
gameNum := strings.TrimSpace(z.Token().Data)
td = advanceToStartTag("td", z)
if td == nil {
return game, errors.New("Unable to find Game Time")
}
td = advanceToStartTag("div", z)
if td == nil {
return game, errors.New("Unable to find Game Time")
}
z.Next()
gameTime := strings.TrimSpace(z.Token().Data)
if gameTime == "" {
t := advanceToTextToken(z)
gameTime = strings.TrimSpace(t.Data)
}
var homeTeam, homeScore, awayTeam, awayScore string
skipAwayScore := false
homeTeam = parseTeamName(z)
homeScore = parseScore(z)
if len(homeScore) > 3 {
awayTeam = homeScore
homeScore = ""
skipAwayScore = true
} else {
awayTeam = parseTeamName(z)
}
if !skipAwayScore {
awayScore = parseScore(z)
} else {
awayScore = ""
}
gameDate, err := time.Parse("1/2/2006 3:04 PM", date+" "+gameTime)
if err != nil {
return game, err
}
return Game{gameDate, gameNum, homeTeam, homeScore, awayTeam, awayScore}, nil
}
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 advanceToStartTag(tagName string, z *html.Tokenizer) *html.Token {
for {
tt := z.Next()
switch tt {
case html.ErrorToken:
return nil
case html.StartTagToken:
t := z.Token()
if t.Data == tagName {
return &t
}
}
}
}
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)
}
// ParseToken is to parse token
func ParseToken(z *html.Tokenizer, tag string) {
for {
tt := z.Next()
switch {
case tt == html.ErrorToken:
// End of the document, we're done
return
case tt == html.StartTagToken:
t := z.Token()
// check element
checkElement(t, tag)
}
}
}
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 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
}
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 parseToken(tokenizer *html.Tokenizer, htmlDoc *htmlDocument, parent *tagElement) (bool, bool, string) {
tokenType := tokenizer.Next()
switch tokenType {
case html.ErrorToken:
return true, false, ""
case html.TextToken:
text := string(tokenizer.Text())
if strings.TrimSpace(text) == "" {
break
}
textElement := &textElement{text: text}
appendElement(htmlDoc, parent, textElement)
case html.StartTagToken:
tagElement := &tagElement{tagName: getTagName(tokenizer), startTagRaw: string(tokenizer.Raw())}
appendElement(htmlDoc, parent, tagElement)
for {
errorToken, parentEnded, unsetEndTag := parseToken(tokenizer, htmlDoc, tagElement)
if errorToken {
return true, false, ""
}
if parentEnded {
if unsetEndTag != "" {
return false, false, unsetEndTag
}
break
}
if unsetEndTag != "" {
return false, false, setEndTagRaw(tokenizer, tagElement, unsetEndTag)
}
}
case html.EndTagToken:
return false, true, setEndTagRaw(tokenizer, parent, getTagName(tokenizer))
case html.DoctypeToken, html.SelfClosingTagToken, html.CommentToken:
tagElement := &tagElement{tagName: getTagName(tokenizer), startTagRaw: string(tokenizer.Raw())}
appendElement(htmlDoc, parent, tagElement)
}
return false, false, ""
}
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 getMatchInfoDateTime(z *html.Tokenizer) (matchDate, matchTime string) {
eof := false
for !eof {
tt := z.Next()
switch {
case tt == html.ErrorToken:
eof = true
case tt == html.StartTagToken:
t := z.Token()
isDiv := t.Data == "div"
if isDiv {
possibleDate := false
for _, a := range t.Attr {
if a.Key == "class" && a.Val == "half" {
possibleDate = true
}
if possibleDate && a.Key == "title" {
// Definitely a date now, grab both date and time
matchDate = a.Val
z.Next()
matchTime = z.Token().Data
// Trim the whitespace around time
matchTime = strings.TrimSpace(matchTime)
return
}
}
}
}
}
Error.Printf("Could not get date and time...")
return "", ""
}
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 (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 (article *NYTArticle) DoParse(parser *html.Tokenizer) error {
articleOpeningTagLoop:
for {
token := parser.Next()
switch {
case token == html.ErrorToken:
return fmt.Errorf("problem moving article %s to open tag", article.GetTitle())
case token == html.StartTagToken:
tmp := parser.Token()
isStartArticle := tmp.Data == "p"
if isStartArticle {
for _, attr := range tmp.Attr {
if attr.Key == "class" && attr.Val == "story-body-text story-content" {
break articleOpeningTagLoop
}
}
}
}
}
isInParagraph := true
articleClosingTagLoop:
for {
token := parser.Next()
switch {
case token == html.ErrorToken:
return fmt.Errorf("problem scraping article %s", article.GetTitle())
case token == html.StartTagToken:
tmp := parser.Token()
isEndArticle := tmp.Data == "footer"
if isEndArticle {
for _, attr := range tmp.Attr {
if attr.Key == "class" && attr.Val == "story-footer story-content" {
break articleClosingTagLoop
}
}
}
if tmp.Data == "p" {
for _, attr := range tmp.Attr {
if attr.Key == "class" && strings.Contains(attr.Val, "story-body-text") {
isInParagraph = true
}
}
if isInParagraph {
continue
}
}
// is a link
if tmp.Data == "a" {
shouldSkip := false
for _, attr := range tmp.Attr {
if attr.Key == "class" && strings.Contains(attr.Val, "visually-hidden") {
shouldSkip = true
}
}
if shouldSkip {
continue
}
parser.Next()
tmp = parser.Token()
newBody := strings.TrimSpace(article.GetData()) + " " + strings.TrimSpace(tmp.Data) + " "
article.SetData(newBody)
isInParagraph = true
}
case token == html.EndTagToken:
tmp := parser.Token()
if tmp.Data == "p" {
isInParagraph = false
}
default:
if !isInParagraph {
continue
}
tmp := parser.Token()
newBody := article.GetData()
// add a space on the left just in case there is a comment or something
if unicode.IsPunct(rune(tmp.Data[0])) {
newBody = strings.TrimSpace(newBody)
}
newBody = newBody + strings.TrimSpace(tmp.Data)
article.SetData(newBody)
isInParagraph = false
}
}
fmt.Println(article.GetData())
return nil
}
func nextToken(tokenizer *html.Tokenizer) error {
if t := tokenizer.Next(); t == html.ErrorToken {
return tokenizer.Err()
}
return nil
}
