func testCodepage(t *testing.T, name string, inReader, outReader func(io.Reader) io.Reader) {
data := make([]byte, 256)
for i := range data {
data[i] = byte(i)
}
inr := inReader(bytes.NewBuffer(data))
r, err := charset.NewReader(name, inr)
if err != nil {
t.Fatalf("cannot make reader for charset %q: %v", name, err)
}
outr := outReader(r)
r = outr
var outbuf bytes.Buffer
w, err := charset.NewWriter(name, &outbuf)
if err != nil {
t.Fatalf("cannot make writer for charset %q: %v", name, err)
}
_, err = io.Copy(w, r)
if err != nil {
t.Fatalf("copy failed: %v", err)
}
err = w.Close()
if err != nil {
t.Fatalf("close failed: %v", err)
}
if len(outbuf.Bytes()) != len(data) {
t.Fatalf("short result of roundtrip, charset %q, readers %T, %T; expected 256, got %d", name, inr, outr, len(outbuf.Bytes()))
}
for i, x := range outbuf.Bytes() {
if data[i] != x {
t.Fatalf("charset %q, round trip expected %d, got %d", name, i, data[i])
}
}
}
// Will receive an input stream which would convert the response to utf-8
// The given function must close the reader r, in order to close the response body.
func HandleStringReader(f func(r io.Reader, ctx *goproxy.ProxyCtx) io.Reader) goproxy.RespHandler {
return goproxy.FuncRespHandler(func(resp *http.Response, ctx *goproxy.ProxyCtx) *http.Response {
if ctx.Error != nil {
return nil
}
charsetName := ctx.Charset()
if charsetName == "" {
charsetName = "utf-8"
}
if strings.ToLower(charsetName) != "utf-8" {
r, err := charset.NewReader(charsetName, resp.Body)
if err != nil {
ctx.Warnf("Cannot convert from %v to utf-8: %v", charsetName, err)
return resp
}
tr, err := charset.TranslatorTo(charsetName)
if err != nil {
ctx.Warnf("Can't translate to %v from utf-8: %v", charsetName, err)
return resp
}
if err != nil {
ctx.Warnf("Cannot translate to %v: %v", charsetName, err)
return resp
}
newr := charset.NewTranslatingReader(f(r, ctx), tr)
resp.Body = &readFirstCloseBoth{ioutil.NopCloser(newr), resp.Body}
} else {
//no translation is needed, already at utf-8
resp.Body = &readFirstCloseBoth{ioutil.NopCloser(f(resp.Body, ctx)), resp.Body}
}
return resp
})
}
func ExampleNewReader() {
r, err := charset.NewReader("latin1", strings.NewReader("\xa35 for Pepp\xe9"))
if err != nil {
log.Fatal(err)
}
result, err := ioutil.ReadAll(r)
if err != nil {
log.Fatal(err)
}
fmt.Printf("%s\n", result)
// Output: £5 for Peppé
}
func main() {
flag.Usage = func() {
fmt.Fprintf(os.Stderr, "usage: tcs [-l] [-v] [charset]\n")
fmt.Fprintf(os.Stderr, "\ttcs [-f charset] [-t charset] [file]\n")
}
flag.Parse()
if *listFlag {
cs := ""
switch flag.NArg() {
case 1:
cs = flag.Arg(0)
case 0:
default:
flag.Usage()
}
listCharsets(*verboseFlag, cs)
return
}
var f *os.File
switch flag.NArg() {
case 0:
f = os.Stdin
case 1:
var err error
f, err = os.Open(flag.Arg(0))
if err != nil {
fatalf("cannot open %q: %v", err)
}
}
r, err := charset.NewReader(*fromCharset, f)
if err != nil {
fatalf("cannot translate from %q: %v", *fromCharset, err)
}
w, err := charset.NewWriter(*toCharset, os.Stdout)
if err != nil {
fatalf("cannot translate to %q: ", err)
}
_, err = io.Copy(w, r)
if err != nil {
fatalf("%v", err)
}
}
// Crawl fetches the HTML body and returns an Article
func (c Crawler) Crawl() (*Article, error) {
article := new(Article)
c.assignParseCandidate()
err := c.assignHTML()
if err != nil {
return nil, err
}
// This was supposed to have been set by assignHTML, so something went wrong
if c.RawHTML == "" {
return article, nil
}
reader := strings.NewReader(c.RawHTML)
document, err := goquery.NewDocumentFromReader(reader)
if err != nil {
return nil, err
}
selection := document.Find("meta").EachWithBreak(func(i int, s *goquery.Selection) bool {
attr, exists := s.Attr("http-equiv")
if exists && attr == "Content-Type" {
return false
}
return true
})
if selection != nil {
attr, _ := selection.Attr("content")
attr = strings.Replace(attr, " ", "", -1)
if strings.HasPrefix(attr, "text/html;charset=") {
cs := strings.TrimPrefix(attr, "text/html;charset=")
cs = strings.ToLower(cs)
if cs != "utf-8" {
r, err1 := charset.NewReader(cs, strings.NewReader(c.RawHTML))
if err1 != nil {
// On error, skip the read
c.RawHTML = ""
} else {
utf8, _ := ioutil.ReadAll(r)
c.RawHTML = string(utf8)
}
reader = strings.NewReader(c.RawHTML)
document, err = goquery.NewDocumentFromReader(reader)
}
}
}
if err != nil {
return nil, err
}
extractor := NewExtractor(c.config)
html, err := document.Html()
if err != nil {
return nil, err
}
startTime := time.Now().UnixNano()
article.RawHTML = html
article.FinalURL = c.helper.url
article.LinkHash = c.helper.linkHash
article.Doc = document
article.Title = extractor.getTitle(article)
article.MetaLang = extractor.getMetaLanguage(article)
article.MetaFavicon = extractor.getFavicon(article)
article.MetaDescription = extractor.getMetaContentWithSelector(article, "meta[name#=(?i)^description$]")
article.MetaKeywords = extractor.getMetaContentWithSelector(article, "meta[name#=(?i)^keywords$]")
article.CanonicalLink = extractor.getCanonicalLink(article)
article.Domain = extractor.getDomain(article)
article.Tags = extractor.getTags(article)
cleaner := NewCleaner(c.config)
article.Doc = cleaner.clean(article)
article.TopImage = OpenGraphResolver(article)
if article.TopImage == "" {
article.TopImage = WebPageResolver(article)
}
article.TopNode = extractor.calculateBestNode(article)
if article.TopNode != nil {
article.TopNode = extractor.postCleanup(article.TopNode)
outputFormatter := new(outputFormatter)
article.CleanedText, article.Links = outputFormatter.getFormattedText(article)
videoExtractor := NewVideoExtractor()
article.Movies = videoExtractor.GetVideos(article)
}
article.Delta = time.Now().UnixNano() - startTime
return article, nil
}
func (this Crawler) Crawl() *Article {
article := new(Article)
this.assignParseCandidate()
this.assignHtml()
if this.rawHtml == "" {
return article
}
reader := strings.NewReader(this.rawHtml)
document, err := goquery.NewDocumentFromReader(reader)
if err != nil {
panic(err.Error())
}
attr := ""
selection := document.Find("meta").EachWithBreak(func(i int, s *goquery.Selection) bool {
attr, exists := s.Attr("http-equiv")
if exists && attr == "Content-Type" {
return false
}
return true
})
if selection != nil {
attr, _ = selection.Attr("content")
attr = strings.Replace(attr, " ", "", -1)
if strings.HasPrefix(attr, "text/html;charset=") {
cs := strings.TrimPrefix(attr, "text/html;charset=")
cs = strings.ToLower(cs)
if cs != "utf-8" {
r, err := charset.NewReader(cs, strings.NewReader(this.rawHtml))
if err != nil {
// On error, skip the read
this.rawHtml = ""
} else {
utf8, _ := ioutil.ReadAll(r)
this.rawHtml = string(utf8)
}
reader = strings.NewReader(this.rawHtml)
document, err = goquery.NewDocumentFromReader(reader)
}
}
}
if err == nil {
extractor := NewExtractor(this.config)
html, _ := document.Html()
start := TimeInNanoseconds()
article.RawHtml = html
article.FinalUrl = this.helper.url
article.LinkHash = this.helper.linkHash
article.Doc = document
article.Title = extractor.getTitle(article)
article.MetaLang = extractor.getMetaLanguage(article)
article.MetaFavicon = extractor.getFavicon(article)
article.MetaDescription = extractor.getMetaContentWithSelector(article, "meta[name#=(?i)^description$]")
article.MetaKeywords = extractor.getMetaContentWithSelector(article, "meta[name#=(?i)^keywords$]")
article.CanonicalLink = extractor.getCanonicalLink(article)
article.Domain = extractor.getDomain(article)
article.Tags = extractor.getTags(article)
cleaner := NewCleaner(this.config)
article.Doc = cleaner.clean(article)
article.TopImage = OpenGraphResolver(article)
if article.TopImage == "" {
article.TopImage = WebPageResolver(article)
}
article.TopNode = extractor.calculateBestNode(article)
if article.TopNode != nil {
article.TopNode = extractor.postCleanup(article.TopNode)
outputFormatter := new(outputFormatter)
article.CleanedText = outputFormatter.getFormattedText(article)
videoExtractor := NewVideoExtractor()
article.Movies = videoExtractor.GetVideos(article)
}
stop := TimeInNanoseconds()
delta := stop - start
article.Delta = delta
} else {
panic(err.Error())
}
return article
}
