Text containing URL is not properly checked about lychee HOT 9 OPEN

I guess what @lebensterben meant was that you use .md as a file extension, where parentheses have a special meaning. If you want to keep Markdown, you can put the link in angular brackets:

<https://raw.githubusercontent.com/ruzickap/ruzickap.github.io/70de7d813c51bbe21762a340c3fa76ced79dedb0/assets/img/posts/2022/2022-09-02-my-sony-a7-iv-settings/Sony_A7_IV_(ILCE-7M4)_-_by_Henry_S%C3%B6derlund_(51739988735).avif>

This works as expected.

Otherwise change the file extension to .txt and you'll find that it also works and finds the full link.

from lychee.

The same issue can be reproduced with:

$ cat > test.md << \EOF
Test: https://upload.wikimedia.org/wikipedia/commons/8/8b/Sony_A7_IV_%28ILCE-7M4%29_-_by_Henry_S%C3%B6derlund_%2851739988735%29.jpg
EOF

$ lychee -v test.md

from lychee.

https://raw.githubusercontent.com/ruzickap/ruzickap.github.io/70de7d813c51bbe21762a340c3fa76ced79dedb0/assets/img/posts/2022/2022-09-02-my-sony-a7-iv-settings/Sony_A7_IV

this is not found

from lychee.

you used a markdown file so you should use a proper link.

the parentheses have special meaning in markdown syntax and thus the parser treats the input after ( differently.

from lychee.

The full URL is (you need to copy the whole URL to test it):

https://raw.githubusercontent.com/ruzickap/ruzickap.github.io/70de7d813c51bbe21762a340c3fa76ced79dedb0/assets/img/posts/2022/2022-09-02-my-sony-a7-iv-settings/Sony_A7_IV_(ILCE-7M4)_-_by_Henry_S%C3%B6derlund_(51739988735).avif

If we consider, that I do not want to use it as a link, but as "normal" text, then it may be:

• Ignored by lychee and not scan at all
• Recognized as URL and then tested as "full URL" including all the characters which are part of the URL

-> Lychee should not recognize the patter as URL and then checked "malformed URL" + reported as non-working URL (404).

^^^ It is just my humble opinion...

What do you think about it?

from lychee.

Also see https://meta.stackexchange.com/questions/13501/links-to-urls-containing-parentheses.

from lychee.

We need something like https://github.com/rust-lang/rust/blob/a3c9eede5d50be0231f229bf28a271cd509861a3/src/librustdoc/passes/lint/bare_urls.rs#L61-L89

from lychee.

I'm not sure if I understand it properly. The link you mentioned is saying:

There are many ways to format URLs in Markdown. Parentheses work better with some methods than others.

1. Bare             - http://example.com/test(1).html
2. Bare, encoded    - http://example.com/test%281%29.html
3. Delimited        - <http://example.com/test(1).html>
4. Hyperlink        - with <a href="http://example.com/test(1).html">parens</a>
5. Markdown Link #1 - with [parens][1]
6. Markdown Link #1 - with [parens at end][2]
7. Markdown Link #2 - with [parens](http://example.com/test(1).html)
8. Markdown Link #2 - with [parens at end](http://example.com/test(1))  

In this issue I'm describing (1) - Bare : where I do not want to use the URL as link, but as a standard "non-clickable" pure text.

The question is - if the bare / plain text / clean "URL-text" (including all the characters) in Markdown file should be supported by lychee or not.

Honestly I was not able to find any "official" Markdown standard which is saying - if writing bare URLs is something bad except https://github.com/updownpress/markdown-lint/blob/master/rules/034-no-bare-urls.md.

Anyway - feel free to close this ticket if it is too difficult to support bare URLs in Markdown (containing special characters) by lychee.

Thank you...

from lychee.

@ruzickap

https://ikatyang.github.io/tree-sitter-markdown/

you can use this to test what a markdown parser will treat ( in markdown text.

this is the syntax parser for Neovim and Emacs >30.0.

from lychee.