vodorod/parcer
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: to markdown

Browse Source
This commit is contained in:
amhsirak
2025-11-20 00:00:56 +05:30
parent da48d46f2a
commit 713d37465d
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
server/src/markdownify/html-to-markdown/html-to-markdown.go
View File

@@ -11,16 +11,16 @@ import (
"unicode/utf8" "unicode/utf8"
"github.com/PuerkitoBio/goquery" "github.com/PuerkitoBio/goquery"
md "github.com/getmaxun/html-to-markdown/v2" md "github.com/getmaxun/html-to-markdown"
"github.com/getmaxun/html-to-markdown/v2/plugin" "github.com/getmaxun/html-to-markdown/plugin"
converter "github.com/getmaxun/html-to-markdown/v2/converter"
"golang.org/x/net/html" "golang.org/x/net/html"
) )
// ConvertHTMLToMarkdown receives HTML and returns a markdown string allocated for C. // ConvertHTMLToMarkdown receives HTML and returns a markdown string allocated for C.
// Function name changed, comment rewritten.
func ConvertHTMLToMarkdown(input *C.char) *C.char { func ConvertHTMLToMarkdown(input *C.char) *C.char {
engine := converter.NewConverter("", true, nil) engine := md.NewConverter("", true, nil)
// engine.Use(plugin.GitHubFlavored()) engine.Use(plugin.GitHubFlavored())
registerPreHandler(engine) registerPreHandler(engine)
@@ -44,7 +44,7 @@ func main() {
// registerPreHandler configures a specialized PRE/code block rule // registerPreHandler configures a specialized PRE/code block rule
// to properly extract nested content and detect languages. // to properly extract nested content and detect languages.
func registerPreHandler(conv *converter.Converter) { func registerPreHandler(conv *md.Converter) {
isNoiseNode := func(class string) bool { isNoiseNode := func(class string) bool {
l := strings.ToLower(class) l := strings.ToLower(class)
return strings.Contains(l, "gutter") || strings.Contains(l, "line-numbers") return strings.Contains(l, "gutter") || strings.Contains(l, "line-numbers")