Currently links to: https://chrome.google.com/webstore/detail/webchatgpt/lpfemeioodjbpieminkklglpmhlngfcn

Should link to: https://chrome.google.com/webstore/detail/gpt2markdown/mlfimpibamecbdnofjnbkjomeieclnjl

Hi ,

Here is a table example I got with the extension,

<div class="markdown prose w-full break-words dark:prose-invert dark"><table><thead><tr><th>Name</th><th>Popular Topics</th></tr></thead><tbody><tr><td>WatchMojo</td><td>Pop culture, entertainment</td></tr><tr><td>TopTenz</td><td>Various</td></tr><tr><td>TheRichest</td><td>Wealth, luxury</td></tr><tr><td>Topsify</td><td>Music</td></tr><tr><td>Tops and Rankings</td><td>Various</td></tr><tr><td>Top Trending</td><td>Trending topics</td></tr><tr><td>Top5s</td><td>Various</td></tr><tr><td>Top 10 Archive</td><td>History, science</td></tr><tr><td>Top 10 Daily</td><td>Current events, pop culture</td></tr><tr><td>Top 10 Nerd</td><td>Geek, nerd culture</td></tr></tbody></table></div>

expected output is markdown table format:

| aaa | bbb |

The table content:

Thanks for this great extension!

I'm having an issue where an ordered (numbered) list isn't being included in the generated Markdown.

Here is a piece of DOM copied as HTML using Chrome Inspector taken from a conversation with ChatGPT which exhibits the issue:

<div class="markdown prose w-full break-words dark:prose-invert light"><p>When using a large language model like GPT-3, it's important to choose prompts that are specific enough to elicit a coherent response, but not so restrictive that they limit the model's creativity and ability to generate diverse and interesting content.</p><p>Here are some tips for creating prompts that are suitable for use with a large language model:</p><ol><li><p>Keep your prompts open-ended: Avoid prompts that are too specific or narrow in scope. Instead, use prompts that allow the model to explore a range of ideas and perspectives.</p></li><li><p>Provide context: Help the model understand the topic by providing relevant background information and context. This will help the model generate more coherent and relevant responses.</p></li><li><p>Avoid leading questions: Avoid prompts that suggest a particular answer or perspective. Instead, use prompts that encourage the model to think creatively and generate original content.</p></li><li><p>Use natural language: Use prompts that are written in natural language, rather than technical jargon or overly formal language. This will help the model generate more natural-sounding responses.</p></li><li><p>Experiment with different prompts: Try different prompts to see what works best for your needs. Some prompts may work better than others depending on the specific task or application.</p></li></ol></div>

(The <div> has at least one class which I believe was inserted by the Dark Reader extension which I'm using.)

The extension as downloaded from the Chrome Web Store only outputs the following Markdown:

**ChatGPT:**
When using a large language model like GPT-3, it's important to choose prompts that are specific enough to elicit a coherent response, but not so restrictive that they limit the model's creativity and ability to generate diverse and interesting content.
Here are some tips for creating prompts that are suitable for use with a large language model:


 --------

I.e. the ordered list following the first paragaph is missing entirely.

ReadME has:

But no dev branch exists yet.

It'd be nice to copy to clipboard instead of just having it as a download

Something that might look like this:

It's good to push users to the chrome extension :D

Current:

Better:

I know this isn't great because you're losing some attribution here to your socials.

I recommend either having this as a tooltip pop up if possible, or maybe removing a little to fit better.

When this setting is set as disabled, the extension can not export to markdown.

Easy fix, just add "items-stretch" to the following line of code:

inputActionNode = document.querySelector(
	"div[class*='relative flex h-full flex-1 md:flex-col']"
);

becomes

inputActionNode = document.querySelector(
	"div[class*='relative flex h-full flex-1 items-stretch md:flex-col']"
);

This is a great extension and works really well. Just what I was looking for.

I did notice one issue that cause my codeblocks to be malformed. The [```python] needs to have a linebreak before it render properly. I was able to fix this with find/replace. Just add a linebreak before this tag, and it looks bueno...

Thanks,
Tony

Input:

Output markdown:

**ChatGPT:**
​


 --------
**You:**
You're going to be an expert python programmer.

 --------
**ChatGPT:**
As an AI language model, I am already proficient in Python programming and can assist with a wide range of tasks related to Python. Please let me know how I can help!


 --------
**You:**
I want you to write a hello world program in Python

 --------
**ChatGPT:**
Certainly! Here is a simple "Hello, world!" program in Python:
<pre><span class="">python</span><button class="flex ml-auto gap-2"><svg stroke="currentColor" fill="none" stroke-width="2" viewBox="0 0 24 24" stroke-linecap="round" stroke-linejoin="round" class="h-4 w-4" height="1em" width="1em" xmlns="http://www.w3.org/2000/svg"><path d="M16 4h2a2 2 0 0 1 2 2v14a2 2 0 0 1-2 2H6a2 2 0 0 1-2-2V6a2 2 0 0 1 2-2h2"></path><rect x="8" y="2" width="8" height="4" rx="1" ry="1"></rect></svg>Copy code</button><code class="!whitespace-pre hljs language-python"><span class="hljs-built_in">print</span>(<span class="hljs-string">"Hello, world!"</span>)
</code></pre>This program will output the string "Hello, world!" to the console.


 --------

VScode Preview:

Expected output would probably be better to remove the extra html/css and use the built in ``` code formatting in markdown.
Probably remove the copy button remnant too.

Great tool!

Issue: It doesn't seem to retain formatting. It looses bullet points, makes lists turn out as simple lines. The entire thing basically just comes out as a single paragraph.

Edit: Hm, it doesn't always happen. I'll investigate further to create a repro and reopen the issue (maybe I'm doing something wrong)...

import org.springframework.data.domain.Page;
import org.springframework.data.domain.Pageable;
import org.springframework.data.domain.Sort;
import org.springframework.data.jpa.domain.Specification;
import org.springframework.data.jpa.repository.support.JpaEntityInformation;
import org.springframework.data.jpa.repository.support.SimpleJpaRepository;
import org.springframework.transaction.annotation.Transactional;

import javax.persistence.EntityManager;
import javax.persistence.TypedQuery;
import java.util.concurrent.TimeUnit;

public class GenericRepositoryImpl<T> extends SimpleJpaRepository<T, Long> implements GenericRepository<T> {

    private final EntityManager entityManager;

    public GenericRepositoryImpl(JpaEntityInformation<T, ?> entityInformation, EntityManager entityManager) {
        super(entityInformation, entityManager);
        this.entityManager = entityManager;
    }

    @Override
    public Page<T> search(String searchQuery, Pageable pageable) {
        return search(searchQuery, null, pageable);
    }

As you may have noticed when we export the Markdown, the code has html escape characters. Also there is always an additional blank link at the start of the code