I created a sh sample code with some comments inside a fenced code block, place it after an list item and then indent it.
If I indent the code block, the syntax highlighting loses its color.
The bug is in the markdown editor and preview.

Does this issue occur when all extensions are disabled?: YES

Basic Information

• VS Code Version: 1.77.3 (user setup)
• OS Version: Windows_NT x64 10.0.19045
• Color Theme: Dark+ (default dark)

Steps to Reproduce:

1. Open markdown file
2. Create list
3. Add a code block (see sample code below)
4. See the result (see screenshots below)

Sample Code

1. List One

```sh
sudo apt-get purge --autoremove -y google-chrome-stable

# with alias
remove -y google-chrome-stable
```

1. List Two

    ```sh
    sudo apt-get purge --autoremove -y google-chrome-stable

    # with alias
    remove -y google-chrome-stable
    ```

1. List Three

    ```sh
        sudo apt-get purge --autoremove -y google-chrome-stable

        # with alias
        remove -y google-chrome-stable
    ```

Screenshots

Actual behavior

Expected behavior

Originally from @adorade in microsoft/vscode#180246

runing 2.0-beta.12.26

Header item will convert to H1-6 with id.
If the contents is only CJK chars, last quotation mark for id is short.

here is a sample.

# 第一階層

top level

## 第二階層

second level

## 第三階層

third level

Show Preview shots

output from Convert Document to HTML

<h1 id=">第一階層</h1>

<p>top level</p>

<h2 id=">第二階層</h2>

<p>second level</p>

<h2 id=">第三階層</h2>

<p>third level</p>

I expect this output.

<h1 id="">第一階層</h1>

<p>top level</p>

<h2 id="">第二階層</h2>

<p>second level</p>

<h2 id="">第三階層</h2>

<p>third level</p>

The Markdown grammar has this for lists:

lists = {
    patterns = (
        {   name = 'markup.list.unnumbered.markdown';
            comment = 'Currently does not support un-indented second lines.';
            begin = '(^|\G)([ ]{0,3})([*+-])([ ]{1,3}|\t)';
            while = '\G([ ]{4}|\t|$)';
            beginCaptures = { 3 = { name = 'punctuation.definition.list.markdown'; }; };
            patterns = (
                {   include = '#list_paragraph'; },
                {   include = '#block'; },
            );
        },
        {   name = 'markup.list.numbered.markdown';
            begin = '(^|\G)([ ]{0,3})([0-9]+\.)([ ]{1,3}|\t)';
            while = '\G([ ]{4}|\t|$)';
            beginCaptures = { 3 = { name = 'punctuation.definition.list.markdown'; }; };
            patterns = (
                {   include = '#list_paragraph'; },
                {   include = '#block'; },
            );
        },
    );
};

For some reason it hits for lists starting after a newline:

But not right after a paragraph:

And I can't imaging why. The (^|\G) capture group should hit looking only at the start of a line, not what precedes it.

Why does it behave wrongly (IMHO)?

To reproduce:

• Open a blank document.
• Set document type to Markdown.
• Type #hi.
• Type command-] or mouse to Bundles > Markdown > Headings > Increase Heading Level.

Result:

Failure running “Increase Heading Level”.
Traceback (most recent call last):
  File "Increase Heading Level", line 16, in <module>
    heading = m.group(2).strip()
AttributeError: 'NoneType' object has no attribute 'group'

TextMate version 2.0.0-alpha.9346.

Steps to reproduce:

• create a markdown document

• turn on soft wrapping

• create a long header name using hashes, for example:

  ## This is my really long chapter title: don't you wish it were shorter?
  

Expected result:

• I would have expected the enlarged typeface for the header to be soft wrapped as the regular text is.

Actual result:

• header runs off the right hand side of the page, hiding the text instead of wrapping it. Here is an example.

Environment:

• OS X 10.8.3, TextMate version 2.0-alpha.9413

From @connor4312 in microsoft/vscode#91803

[link](url "title")
[link](url 'title')

Both are valid, but only the double quotes are tokenized correctly.

Hi,

When printing the Markdown preview (⌃⌥⌘P) via ⌘P to a PDF-file or to Preview.app, crashes TM completely.

fyi, printing via the Multi Markdown bundle goes OK.

--feek

I can see that the bundle is installed, but I don't have any highlighting. This is what I did:

    mkdir -p ~/Library/Application\ Support/TextMate/Bundles
    cd ~/Library/Application\ Support/TextMate/Bundles
    git clone https://github.com/textmate/markdown.tmbundle.git markdown.tmbundle

    touch /tmp/test.md
    open /tmp/test.md

I also tried .markdown as an extension.

There's an issue with the HTML pattern that breaks if open & close tags are on separate lines in certain cases:

It would be awesome if you were able to refresh the Markdown preview window without having to close and reopen the preview.

One option could be a key command.
A better alternative would be to have the Preview window live update like the Mou app does.

Try e.g. ‘=⇥’ in (otherwise) empty document or call the commands via the menu.

This produces a ruby ‘rstrip called for nil’ error.

Ideally we should show a tool tip that explain intended usage.

I am unable to find where this happens (in the bundle), but would like to disable it. Marked thinks the “#” is a character to be displayed, and not a delimiter.

e.g.

Lorem or

and you get this:

Lorem

Marked thinks it’s supposed to display like this:

Lorem #

that is NOT what I want.

Any way to disable this? Or is this a standard?

Thanks
Mark

ADDENDUM

• I just turned off using “discount” (aka GFM) in Marked, and now headers format correctly.
• I also noticed that COMMAND+ does not do add the trailing “#"

If I make a markdown file containing something like this:

# does something
print("Hello, world")

The comment # does something is a really big font, since it's displaying as if it were a <h1>. If nobody else wants to, I might have a go at fixing this.

TM2 always autocompleted markdown lists. E.g., after typing "- list item one" and hitting return, the next line would display "- ", the hyphen and space there waiting for me with the cursor placed after the space to start writing the next bullet point.

This no longer happens and I'm wondering why. Is it just my version or do others have this problem?

Also, when I tab in the text appears in a different font and breaks the page wrap. Screenshot: http://cl.ly/ZrX5

Am I setup wrong on this or is this something in the bundle itself?

Thanks, all.

show, don’t tell

-- somebody

• please support task list items
• with checkboxes instead of bullets
• within a list,
• as does Github
• the end

Hi,

When I run the new item or new subitem commands I get a ruby error. The parse function that the new item/subitem calls from the markdown.rb file uses a method str.to_a() which doesn't seem to work. This is on OS X Mavericks, which may be using a new version of ruby than this script was expecting?

Hello!

This could be my theme doing this, but I got consistent sizing issues with several different themes, so I just wanted to check in.

I'd like my font sizes to be consistent here. How can I go about doing that?

new installed TextMate version 2.0-alpha.9517 on
same error when i try to preview a markdown file

    Preview: line 4: /lib/webpreview.sh: No such file or directory
    Preview: line 5: html_header: command not found
    Preview: line 13: Markdown.pl: command not found
    Preview: line 13: SmartyPants.pl: command not found
    Preview: line 28: html_footer: command not found

https://github.com/textmate/textmate/wiki/Reverting-To-Defaults doesn't work for me.

Textmate markdown bundle: Insert Image tag Dragaction Command

Dragging a file, that has spaces in the filename will break the inserted link.
The additional line in the command converts spaces to %20.
While Firefox and other browser don't mind mucht for spaces, the mardown preview won't show images.

change

base=${TM_DROPPED_FILE##*/}
alt=$(tr <<<${base%.*} '[_-]' ' '|perl -pe 's/(\w+)/\u$1/g')
if [ -n "$TM_DROPPED_FILE_STRIP_ROOT" ]
then
    file=$(sed -e "s|^$TM_DROPPED_FILE_STRIP_ROOT||" <<< $TM_DROPPED_FILE);
else
    file=${TM_DROPPED_FILE};
fi
echo -n "![\${1:$alt}]($file)"

to

base=${TM_DROPPED_FILE##*/}
alt=$(tr <<<${base%.*} '[_-]' ' '|perl -pe 's/(\w+)/\u$1/g')
if [ -n "$TM_DROPPED_FILE_STRIP_ROOT" ]
then
    file=$(sed -e "s|^$TM_DROPPED_FILE_STRIP_ROOT||" <<< $TM_DROPPED_FILE);
else
    file=${TM_DROPPED_FILE};
fi
file=`echo $file | sed -e 's/ /%20/g'`
echo -n "![\${1:$alt}]($file)"

Would it be possible add support for a few extended markdown features? I'm mainly thinking of support for the commonly used fenced code blocks. If you're concerned about automatically formatting non-standard markdown, then perhaps this could be something that's turned off by default, which can be enabled within the bundle settings.

I understand someone made a request for full github flavoured markdown support #9. If indeed you'd prefer to split this out into a new bundle using markdown as it's base, can I make a vote for that? I think many of these extensions, at least fenced code blocks, are common enough to warrant making it an official bundle.

Perhaps there's an opportunity here to even support the class hinting that fenced code blocks allow, and which is commonly used for code highlighting. So if someone marks a fenced code block with the ruby class hint, then Markdown can render the content of that code block using the grammer from the ruby bundle. That'd be pretty awesome.

Thoughts?

I expected the markdown bundle to work the same as previous MD bundle in TM1.

1. Type a hyphen and space to start a list: "- "
2. type some shit.
3. hit Return... expected behavior is another "- " being added to the beginning of the next line. This does not happen.
4. another shot of tequila.... try again. no dice. get terrified/angry.

I've discovered that this MD bundle expects lists to use asterisks ( * ) instead of hyphens and "Enter" instead of "Return" (fn+return on laptop). Both of these require extra keystrokes (sub-optimal) and it seems unnecessary to change the behavior from TM1.

This is my humble request for an update to this behavior.

Big fan. Thanks.

Have been using this bundle with the latest Textmate 2.0 alpha, and find the use of different (variable width) fonts, and at different sizes, distracting.

It ruins the whole typewriter vibe that you get from writing in a text editor--not to mention that there's just something really uncomfortable about combining fixed-width and variable fonts in the same view. It's exposure to this kind of overdesign that writers move to Markdown to escape...

Is it possible to have an option, or a separate bundle in the repository, that sticks to one font and size throughout, like the old bundle that came with Textmate 1.x?

I can create one new subitem in a list but when I try to create another list item I get this error message:

Failure running “New Item”.

/Users/geoffreysnyder/Library/Application Support/TextMate/Managed/Bundles/Markdown.tmbundle/Support/lib/markdown.rb:320:in parse': undefined methodlines' for nil:NilClass (NoMethodError)
from New Item:11

As long as I only create lists without subitems the create new item works as expected.

Are there plans to add formatting for YAML front matter in a Markdown document? This is getting to be a pretty standard thing in many flat-file CMSs.

If there are no plans, how would I go about adding such a thing?

Is it possible to add support for GitHub-specific syntax?

I'm especially interested in "Fenced code blocks"

Github Markdown can use triple backticks to start and end a code section, this is useful because it's the only way to specify a language for syntax highlighting.

Would it be appropriate to include this in the markdown bundle (given it's github flavored markdown)?

As of now README files look a a little sad if you're using the triple backtick syntax.

I get the following error message:

Failure running “Syntax Cheat Sheet”.
Syntax Cheat Sheet: line 4: /lib/webpreview.sh: No such file or directory
Syntax Cheat Sheet: line 5: html_header: command not found
Syntax Cheat Sheet: line 6: Markdown.pl: command not found
Syntax Cheat Sheet: line 7: html_footer: command not found

Just after updating to TextMate version 2.0-alpha.9495 it stopped working properly.

I just noticed this happening. I rarely use the SETEXT. Any help as to why I am getting this. I’m not sure where to start.

Yosemite 10.10.1
TextMate version 2.0-beta.6.5
ruby 2.0.0p481
latest version of markdown bundle

update

it appears that “/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/bin/ruby” does not exist. “/System/Library/Frameworks/“ is as far as it goes.

If in the markdown code there is a *, everything after that character will have italic style.
I saw this recently in the GitHub diffs (e.g. this one):

When specifying Markdown (MathJax) math blocks for writing equations, some equation characters will break the syntax highlighting for the file, making all of the text below the equation unhighlighted. This happens most frequently with equations that have underscores in them. Here is an example:

If the offending equation is removed, the highlighting is restored:

Everything inside of double-dollar signs ($$) should be considered part of the equation.

Snapshot is self explanatory:

Compare it with how this code looks:

///========================== main.cpp ==========================
/// iostream provides basic input/output functionalities (cout)
#include <iostream>
/// iostream is part of the "std" namespace. The keyword "using" can 
/// be employed to avoid typing std::cout throughout the code
using namespace std;

/// main is the entry point of your program. It needs to be 
/// defined and always have this signature. The argc and argv 
/// are used to pass parameters from command line.
int main(int /*argc*/, char** /*argv*/){
    /// @todo find out how to print "helloworld" to terminal
    /// @solution  
    cout << "helloworld" << endl;

    /// The main function should always return a value.
    /// This tells the application whether it exited 
    /// correctly or not (legacy error management)
    return EXIT_SUCCESS;
}

Also, disabling soft-text wrap in textmate whenever you are in a code block would be a good idea!

In the TextMate 2 markdown bundle, why do I have to manually type the numbers in my numbered lists?

After doing:

1. asdf

and pressing Enter, why do I manually have to type

1. foo
2. bar

and if I insert an element in between, why do I have to manually change the other numbers? Shouldn't it be more intelligent than this?

Just wondering.

Would it be possible for you to add YAML formatting for front matter that is often used in Markdown files?

Fletcher added MultiMarkdown specific bundle functions.

cf. https://github.com/fletcher/markdown.tmbundle