new lines in gr.Markdown

[hysts]

Describe the bug

Not sure if this is a bug or just a breaking change, but the way gr.Markdown handles line breaks seems to have been changed in gradio==3.41.x.

Consecutive lines in markdown are often rendered as one long line unless there is blank line in it. GitHub renders markdown that way, and also it was the way gr.Markdown rendered markdown with gradio==3.40.1.
But now with gradio==3.41.2, consecutive lines are rendered as separate lines. Is this an intentional change or a bug?

Have you searched existing issues? 🔎

• I have searched and found no existing issues

Reproduction

import gradio as gr

markdown = """
The quick brown fox jumps over the lazy dog. The quick brown fox jumps over
the lazy dog.

The quick brown fox jumps over the lazy dog. The quick brown fox jumps over the lazy dog.

The quick brown fox jumps over the lazy dog. The quick brown fox jumps over\
the lazy dog.
"""

with gr.Blocks() as demo:
    gr.Markdown(markdown)
demo.queue().launch()

https://huggingface.co/spaces/hysts-debug/markdown-new-line-gradio-3-40-1
https://huggingface.co/spaces/hysts-debug/markdown-new-line

Screenshot

gradio==3.40.1

gradio==3.41.2

Logs

No response

System Info

Gradio Environment Information:
------------------------------
Operating System: Linux
gradio version: 3.41.2
gradio_client version: 0.5.0

------------------------------------------------
gradio dependencies in your environment:

aiofiles: 23.2.1
altair: 5.0.1
fastapi: 0.103.0
ffmpy: 0.3.1
gradio-client==0.5.0 is not installed.
httpx: 0.24.1
huggingface-hub: 0.16.4
importlib-resources: 6.0.1
jinja2: 3.1.2
markupsafe: 2.1.3
matplotlib: 3.7.2
numpy: 1.25.2
orjson: 3.9.5
packaging: 23.1
pandas: 2.0.3
pillow: 10.0.0
pydantic: 1.10.12
pydub: 0.25.1
python-multipart: 0.0.6
pyyaml: 6.0.1
requests: 2.31.0
semantic-version: 2.10.0
typing-extensions: 4.7.1
uvicorn: 0.23.2
websockets: 11.0.3
authlib; extra == 'oauth' is not installed.
itsdangerous; extra == 'oauth' is not installed.


gradio_client dependencies in your environment:

fsspec: 2023.6.0
httpx: 0.24.1
huggingface-hub: 0.16.4
packaging: 23.1
requests: 2.31.0
typing-extensions: 4.7.1
websockets: 11.0.3

Severity

I can work around it

[pngwn]

This was intentional in order to make it consistent with our other uses of markdown in gradio.

This consistent with how markdown behaves in GitHub comments but not in rendered files.

If we change this settings we'll need to change it for everything.

[hysts]

Thanks. I see. Yeah, GitHub renders lines in comments in separate lines, but GitHub renders consecutive lines in README.md etc. in one line, and VS Code also renders markdown in the same way.

I feel like there are pros and cons to both, so maybe it would be nice to have an optional argument in gr.Markdown to switch how to render lines.

[abidlabs]

I think its better not too add too many parameters to gr.Markdown(). Since there's a very easy workaround, I'd prefer marking this as a breaking change and moving on.

[hysts]

By a very easy workaround, do you mean adding \ everywhere? I was aware of this workaround, but the reason I opened this issue was actually because I found it cumbersome and didn't want to do it. People need to add and delete \ every time they edit markdown string.

I got interested in what other libraries were doing. Looks like Python's markdown library has an extension to render new line as <br>, and markdown-it also has an option break: true to do so. Looking at these, I feel like there's a need for both options.
Also, I wonder if it is really better to render new line as <br> by default. Just out of curiosity, what is the reason for this change in behavior? Changing the other uses of markdown in gradio to not render newline as <br> was also an option, no?

[abidlabs]

Oh sorry I misunderstood. I thought the change went the other way around -- we no longer respected single line breaks, so my suggestion would have been to just add a line break. This is a bit more annoying to deal with and I agree not the typical way I've seen Markdown rendered. I think its reasonable to add a flag to control this behavior, particularly for gr.Markdown() since its a component dedicated to rendering Markdown (whereas for gr.Chatbot for example we can be more opinionated with how we deal with Markdown)

Thanks for clarifying @hysts