From d62c9724b7f4cb728cd5b5496fdcc0eba8330772 Mon Sep 17 00:00:00 2001
From: MohamedBassem <me@mbassem.com>
Date: Sat, 21 Sep 2024 16:56:42 +0000
Subject: feature(web): Preserve title, tags and createdAt when importing a
 netscape html. Fixes #401

---
 apps/web/lib/netscapeBookmarkParser.ts | 35 ++++++++++++++++++++++------------
 1 file changed, 23 insertions(+), 12 deletions(-)

(limited to 'apps/web/lib/netscapeBookmarkParser.ts')

diff --git a/apps/web/lib/netscapeBookmarkParser.ts b/apps/web/lib/netscapeBookmarkParser.ts
index ac5f3ec2..196c0525 100644
--- a/apps/web/lib/netscapeBookmarkParser.ts
+++ b/apps/web/lib/netscapeBookmarkParser.ts
@@ -1,20 +1,31 @@
-function extractUrls(html: string): string[] {
-  const regex = /<a\s+(?:[^>]*?\s+)?href="(http[^"]*)"/gi;
-  let match;
-  const urls = [];
-
-  while ((match = regex.exec(html)) !== null) {
-    urls.push(match[1]);
-  }
-
-  return urls;
-}
+// Copied from https://gist.github.com/devster31/4e8c6548fd16ffb75c02e6f24e27f9b9
+import * as cheerio from "cheerio";
 
 export async function parseNetscapeBookmarkFile(file: File) {
   const textContent = await file.text();
+
   if (!textContent.startsWith("<!DOCTYPE NETSCAPE-Bookmark-file-1>")) {
     throw Error("The uploaded html file does not seem to be a bookmark file");
   }
 
-  return extractUrls(textContent).map((url) => new URL(url));
+  const $ = cheerio.load(textContent);
+
+  return $("a")
+    .map(function (_index, a) {
+      const $a = $(a);
+      const addDate = $a.attr("add_date");
+      let tags: string[] = [];
+      try {
+        tags = $a.attr("tags")?.split(",") ?? [];
+      } catch (e) {
+        /* empty */
+      }
+      return {
+        title: $a.text(),
+        url: $a.attr("href"),
+        tags: tags,
+        addDate: typeof addDate === "undefined" ? undefined : parseInt(addDate),
+      };
+    })
+    .get();
 }
-- 
cgit v1.2.3-70-g09d2