encoder+decoder: escape/unescape HTML characters in bookmark descript…

…ions Signed-off-by: VirtualTam <[email protected]>
virtualtam · Oct 26, 2023 · 5fe95e9 · 5fe95e9
1 parent 3623544
commit 5fe95e9
Show file tree

Hide file tree

Showing 4 changed files with 73 additions and 2 deletions.
diff --git a/decoder.go b/decoder.go
@@ -1,6 +1,7 @@
 package netscape
 
 import (
+	"html"
 	"sort"
 	"strconv"
 	"strings"
@@ -105,7 +106,7 @@ func (d *Decoder) decodeFolder(f FolderNode) (Folder, error) {
 
 func (d Decoder) decodeBookmark(b BookmarkNode) (Bookmark, error) {
 	bookmark := Bookmark{
-		Description: b.Description,
+		Description: html.UnescapeString(b.Description),
 		URL:         b.Href,
 		Title:       b.Title,
 		Attributes:  map[string]string{},

diff --git a/decoder_test.go b/decoder_test.go
@@ -283,6 +283,50 @@ func TestDecodeBookmark(t *testing.T) {
 				Description: "Nested lists:\n- list1\n  - item1.1\n  - item1.2\n  - item1.3\n- list2\n  - item2.1",
 			},
 		},
+		{
+			tname: "bookmark with description containing escaped HTML characters",
+			input: BookmarkNode{
+				Description: "&#34;Fran &amp; Freddie&#39;s Diner&#34; &lt;[email protected]&gt;",
+				Href:        "https://domain.tld",
+				Title:       "Test Domain",
+			},
+			want: Bookmark{
+				Title:       "Test Domain",
+				URL:         "https://domain.tld",
+				Description: `"Fran & Freddie's Diner" <[email protected]>`,
+			},
+		},
+		{
+			tname: "bookmark with multi-line description containing escaped HTML characters",
+			input: BookmarkNode{
+				Description: `
+&gt; The format of here-documents is:
+
+` + "```shell" + `
+[n]&lt;&lt;[-]word
+        here-document
+delimiter
+` + "```" + `
+
+&gt; If any part of word is quoted, the delimiter is the result of quote removal on word, and the lines in the here-document are not expanded.`,
+				Href:  "https://domain.tld",
+				Title: "Test Domain",
+			},
+			want: Bookmark{
+				Title: "Test Domain",
+				URL:   "https://domain.tld",
+				Description: `
+> The format of here-documents is:
+
+` + "```shell" + `
+[n]<<[-]word
+        here-document
+delimiter
+` + "```" + `
+
+> If any part of word is quoted, the delimiter is the result of quote removal on word, and the lines in the here-document are not expanded.`,
+			},
+		},
 		{
 			tname: "bookmark with creation date",
 			input: BookmarkNode{

diff --git a/encoder.go b/encoder.go
@@ -4,6 +4,7 @@ import (
 	"bufio"
 	"encoding/xml"
 	"fmt"
+	"html"
 	"io"
 	"sort"
 	"strings"
@@ -230,7 +231,7 @@ func (p *printer) marshalBookmark(b *Bookmark) error {
 	}
 
 	if b.Description != "" {
-		_, err = p.writeString(fmt.Sprintf("<DD>%s\n", b.Description))
+		_, err = p.writeString(fmt.Sprintf("<DD>%s\n", html.EscapeString(b.Description)))
 		if err != nil {
 			return err
 		}

diff --git a/marshal_test.go b/marshal_test.go
@@ -45,6 +45,20 @@ func TestMarshal(t *testing.T) {
 							URL:         "https://test.domain.tld",
 							Title:       "Test Domain II",
 						},
+						{
+							Description: `
+> The format of here-documents is:
+
+` + "```shell" + `
+[n]<<[-]word
+        here-document
+delimiter
+` + "```" + `
+
+> If any part of word is quoted, the delimiter is the result of quote removal on word, and the lines in the here-document are not expanded.`,
+							URL:   "https://markdown.xml",
+							Title: "Markdown Description",
+						},
 					},
 				},
 			},
@@ -58,6 +72,17 @@ func TestMarshal(t *testing.T) {
     <DT><A HREF="https://domain.tld" PRIVATE="0">Test Domain</A>
     <DT><A HREF="https://test.domain.tld" PRIVATE="0">Test Domain II</A>
     <DD>Second test
+    <DT><A HREF="https://markdown.xml" PRIVATE="0">Markdown Description</A>
+    <DD>
+&gt; The format of here-documents is:
+
+` + "```shell" + `
+[n]&lt;&lt;[-]word
+        here-document
+delimiter
+` + "```" + `
+
+&gt; If any part of word is quoted, the delimiter is the result of quote removal on word, and the lines in the here-document are not expanded.
 </DL><p>
 `,
 		},