I have been using a plugin called “URL Hyperlink” from the marketplace. It converts URLs into markdown URL with the page title retrieved automatically. The plugin has been great, except that sometimes it retrieves back page titles with encoded characters (e.g. "e) that I had to manually fix.
I deleted this plugin by mistake. Because all other similar plugins have far more serious issues: they tend to retrieve/refresh on page scrolling thus consuming huge amount of resources. They make page scrolling with many URLs not bareable sometimes.
However, I could not go back to “URL hyperlink” plugin. It’s still listed in the Marketplace but the github page is no longer there.
I wonder if anyone has the plugin installed and kindly shares the binaries. Or if you have really good alternatives, please let me know. I have tried pretty much everything I can found, none as good as “URL Hyperlink”.
Your plugin is the best. Only by trying other similar plugins, did I realize the nuances. For example, other plugins try to insert icons into the formatted URL. That triggers repeated download in real time and making scrolling pages with many links unbearable. Your plugin clearly is triggered only by URL paste or change. Great design.
Hope we can get the bug fixed. It will be the perfect plugin. LOL
‘URL Hyperlink’ plugin is not compatible with the Logseq db version, so I created a new plugin.
‘External-link’ plugin (unpublished)
I came up with a simpler way of handling URLs. Press the icon next to the URL to convert it. Icons are generated for what Logseq recognises as links.
Excellent feature. Is the little link icon a new feature of the DB version ?
I am not on the DB version yet. When I copy & paste a URL, it appears only as a link with no icon. With the link icon, clicking on it is more natural than the current trigger of URL change. I like it.
I am very happy with the current plugin design. The two nagging issues are not being able to retrieve page titles.
Sometimes it returns encoded characters. And pretty much all pages from wechat (qq.com) are not retrieved at all. Interestingly, if the wechat(qq.com) link is opened in Edge browser, Edge is always able to convert the URL to MD style. If I simply copy the URL field from Edge after the URL is open, and then paste to logseq, the URL is already converted to MD style with the correct title. It implies Edge is not affected . Example: h t t p s ://mp.weixin.qq.com/s/i8pWVpcToXNZy2GP_udjJw
wechat is unique. Normal web pages use a tag in the HTML source, but wechat is empty of them. It seems to use the less common tag and reflects it in JavaScript. It is probably possible to fetch the title. May or may not get the title due to the length of the source, don’t know.
If encoded characters are returned, pattern processing may be required for the site. I do not know which site it is. Normally, the charset is found from the tag, but if it is not specified, or if it is incorrect, it will not be processed correctly.
