func (markdown Markdown) Extract(creativeWork *schema.CreativeWork, path string) error {
markdownContent, err := ioutil.ReadFile(path)
if nil != err {
return err
}
html := blackfriday.MarkdownCommon(markdownContent)
doc, err := goquery.NewDocumentFromReader(bytes.NewReader(html))
if nil != err {
return err
}
doc.Find("a[href]").Each(func(i int, s *goquery.Selection) {
link, _ := s.Attr("href")
url, _ := url.Parse(link)
if !url.IsAbs() && strings.HasSuffix(link, ".md") {
s.SetAttr("href", fmt.Sprint(link[:len(link)-3], ".jsonld"))
}
})
creativeWork.Name = doc.Find("h1").Text()
creativeWork.Text, err = doc.Find("body").Html()
if nil != err {
return err
}
return nil
}
func (markdown Markdown) Extract(creativeWork *schema.CreativeWork, path string) error {
markdownContent, err := ioutil.ReadFile(path)
if nil != err {
return err
}
unsafe := blackfriday.MarkdownCommon(markdownContent)
p := bluemonday.UGCPolicy()
p.RequireNoFollowOnLinks(false)
p.AllowAttrs("class").Matching(regexp.MustCompile("^language-[a-zA-Z0-9]+$")).OnElements("code")
html := p.SanitizeBytes(unsafe)
doc, err := goquery.NewDocumentFromReader(bytes.NewReader(html))
if nil != err {
return err
}
doc.Find("a[href]").Each(func(i int, s *goquery.Selection) {
link, _ := s.Attr("href")
url, _ := url.Parse(link)
if !url.IsAbs() {
s.SetAttr("href", strings.Replace(link, ".md", ".jsonld", 1))
}
})
creativeWork.Name = doc.Find("h1").Text()
creativeWork.Text, err = doc.Find("body").Html()
if nil != err {
return err
}
return nil
}
func (git Git) Extract(creativeWork *schema.CreativeWork, path string) error {
if "" == gitPath {
return nil
}
cmd := exec.Command(gitPath, "log", "--format=%an;%ae;%aI", path)
stdout, err := cmd.StdoutPipe()
if nil != err {
return err
}
if err := cmd.Start(); err != nil {
return err
}
scanner := bufio.NewScanner(stdout)
for scanner.Scan() {
parts := strings.Split(strings.TrimSpace(scanner.Text()), ";")
author := schema.NewPerson(parts[0], parts[1])
creativeWork.Author = append([]schema.Person{*author}, creativeWork.Author...)
creativeWork.DateCreated = parts[2]
if "" == creativeWork.DateModified {
creativeWork.DateModified = parts[2]
}
}
if err := scanner.Err(); err != nil {
return err
}
if err := cmd.Wait(); err != nil {
log.Fatalln("You are not in a git repository.")
return err
}
return nil
}