Strip HTML tags (but keep any text content) when rendering text

go.mod

@@ -7,6 +7,7 @@ require (
 	github.com/Masterminds/sprig/v3 v3.2.2
 	github.com/gomarkdown/markdown v0.0.0-20210915032930-fe0e174ee09a
 	github.com/google/uuid v1.3.0 // indirect
+	github.com/grokify/html-strip-tags-go v0.0.1
 	github.com/huandu/xstrings v1.3.2 // indirect
 	github.com/imdario/mergo v0.3.12 // indirect
 	github.com/mattn/go-runewidth v0.0.13 // indirect

go.sum

@@ -21,6 +21,8 @@ github.com/gomarkdown/markdown v0.0.0-20210915032930-fe0e174ee09a/go.mod h1:JDGc
 github.com/google/uuid v1.1.1/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/google/uuid v1.3.0 h1:t6JiXgmwXMjEs8VusXIJk2BXHsn+wx8BZdTaoZ5fu7I=
 github.com/google/uuid v1.3.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
+github.com/grokify/html-strip-tags-go v0.0.1 h1:0fThFwLbW7P/kOiTBs03FsJSV9RM2M/Q/MOnCQxKMo0=
+github.com/grokify/html-strip-tags-go v0.0.1/go.mod h1:2Su6romC5/1VXOQMaWL2yb618ARB8iVo6/DR99A6d78=
 github.com/huandu/xstrings v1.3.1/go.mod h1:y5/lhBue+AyNmUVz9RLU9xbLR0o4KIIExikq4ovT0aE=
 github.com/huandu/xstrings v1.3.2 h1:L18LIDzqlW6xN2rEkpdV8+oL/IXWJ1APd+vsdYy4Wdw=
 github.com/huandu/xstrings v1.3.2/go.mod h1:y5/lhBue+AyNmUVz9RLU9xbLR0o4KIIExikq4ovT0aE=

internal/renderer/renderer.go

@@ -22,9 +22,9 @@ import (
 	"fmt"
 	"io"
 	"regexp"
-	"strings"
 
 	"github.com/gomarkdown/markdown/ast"
+	"github.com/grokify/html-strip-tags-go"
 	"github.com/olekukonko/tablewriter"
 )
 
@@ -51,6 +51,21 @@ var (
 // matches a FULL string that contains no non-whitespace characters
 var emptyLineRegex = regexp.MustCompile(`\A[\s]*\z`)
 
+// fairly tolerant to handle weird HTML
+var tagPairRegexString = `<[\n\f ]*%s([\n\f ]+[^\n\f \/>"'=]+[\n\f ]*(=[\n\f ]*([a-zA-Z1-9\-]+|"[^\n\f"]+"|'[^\n\f']+'))?)*[\n\f ]*>.*?<[\n\f ]*/[\n\f ]*%s[\n\f ]*>`
+
+// HTML block tags whose contents should not be rendered
+var htmlNoRenderRegex = []*regexp.Regexp{
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "fieldset", "fieldset")),
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "form", "form")),
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "iframe", "iframe")),
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "script", "script")),
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "style", "style")),
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "canvas", "canvas")),
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "dialog", "dialog")),
+	regexp.MustCompile(fmt.Sprintf(tagPairRegexString, "progress", "progress")),
+}
+
 // Renderer implements markdown.Renderer.
 type Renderer struct{}
 
@@ -125,7 +140,7 @@ func (r Renderer) subscript(w io.Writer, node *ast.Subscript, entering bool) {
 	if entering {
 		if node := node.AsLeaf(); node != nil {
 			w.Write(subOpen)
-			w.Write([]byte(strings.ReplaceAll(string(node.Literal), "\n", " ")))
+			w.Write(bytes.ReplaceAll(node.Literal, lineBreak, space))
 			w.Write(subClose)
 		}
 	}
@@ -134,7 +149,7 @@ func (r Renderer) superscript(w io.Writer, node *ast.Superscript, entering bool)
 	if entering {
 		if node := node.AsLeaf(); node != nil {
 			w.Write(supOpen)
-			w.Write([]byte(strings.ReplaceAll(string(node.Literal), "\n", " ")))
+			w.Write(bytes.ReplaceAll(node.Literal, lineBreak, space))
 			w.Write(supClose)
 		}
 	}
@@ -338,6 +353,7 @@ func (r Renderer) list(w io.Writer, node *ast.List, level int) {
 }
 
 var lineBreakCharacters = regexp.MustCompile(`[\n\r]+`)
+var hardBreakTag = regexp.MustCompile(`< *br */? *>`)
 
 func textWithNewlineReplacement(node ast.Node, replacement []byte) []byte {
 	buf := bytes.Buffer{}
@@ -346,13 +362,20 @@ func textWithNewlineReplacement(node ast.Node, replacement []byte) []byte {
 	if node, ok := node.(*ast.Link); ok && node.Footnote != nil {
 		fmt.Fprintf(&buf, "[^%d]", node.NoteID)
 	}
-	if node := node.AsLeaf(); node != nil {
+	if leaf := node.AsLeaf(); leaf != nil {
 		// replace all newlines in text with preferred symbols; this may
 		// be spaces for general text, allowing for soft wrapping, which
 		// is recommended as per Gemini spec p. 5.4.1, or line breaks
 		// with a blockquote symbols for blockquotes, or just nothing
 		buf.Write(delimiter)
-		buf.Write(lineBreakCharacters.ReplaceAll(node.Literal, replacement))
+		switch node.(type) {
+		case *ast.Hardbreak:
+			buf.Write(lineBreak)
+		case *ast.HTMLSpan:
+			buf.Write(leaf.Content)
+		default:
+			buf.Write(lineBreakCharacters.ReplaceAll(leaf.Literal, replacement))
+		}
 		buf.Write(delimiter)
 	}
 	if node := node.AsContainer(); node != nil {
@@ -440,6 +463,22 @@ func (r Renderer) table(w io.Writer, node *ast.Table, entering bool) {
 	}
 }
 
+func (r Renderer) htmlBlock(w io.Writer, node *ast.HTMLBlock, entering bool) {
+	if entering {
+		// Only render contents of allowed tags
+		literal := node.Literal
+		for _, re := range htmlNoRenderRegex {
+			literal = re.ReplaceAllLiteral(literal, []byte(""))
+		}
+		if len(literal) > 0 {
+			literalWithBreaks := hardBreakTag.ReplaceAll(lineBreakCharacters.ReplaceAll(literal, space), lineBreak)
+			w.Write([]byte(strip.StripTags(string(literalWithBreaks))))
+			w.Write(lineBreak)
+			w.Write(lineBreak)
+		}
+	}
+}
+
 // RenderNode implements Renderer.RenderNode().
 func (r Renderer) RenderNode(w io.Writer, node ast.Node, entering bool) ast.WalkStatus {
 	// entering in gomarkdown was made to have elements of type switch
@@ -487,6 +526,8 @@ func (r Renderer) RenderNode(w io.Writer, node ast.Node, entering bool) ast.Walk
 		r.table(w, node, entering)
 		noNewLine = false
 		fetchLinks = true
+	case *ast.HTMLBlock:
+		r.htmlBlock(w, node, entering)
 	}
 	if !noNewLine && !entering {
 		w.Write(lineBreak)