yours-truly-phil / codey Goto Github PK

Just saw some code blocks that looked like valid java code but I didn't see the ✨ reaction added to it from the CodeFormatter.
Turns out, I didn't know that it's valid to start the content of a code block in the line of the ```. I always expected either the three ``` and a newline or the three ```, then some text (like 'java', 'cpp', etc.) which is the language attribute and then a newline.
I need to be able to recognize whether or not the first word in a codeblock is the language attribute, so I'm looking for a list with shorthands for all supported languages.
So I'm thinking I will either use a comprehensive list like https://github.com/github/linguist/blob/master/lib/linguist/languages.yml to parse the language attribute, or create my own list with all the languages the bot actually supports, which as of now would be just ["java"].

I'm both looking into https://github.com/vsch/flexmark-java and https://github.com/atlassian/commonmark-java

So far, both don't do what I expect:

It's odd that neither of them recognizes the end of the codeblock if trying to parse:

"""
Hello, what's going on,
now comes a code block:
```java
I'm inside the code block
hooray!```
I come after the code block"""

The result after parsing for both libraries is one Paragraph that contains two children:
First child: "Hello, what's going on,"
Second child: "now comes a code block:"

The a second node of type FencedCodeBlock with one child spread over three lines:
Line 1 of FencedCodeBlock: "I'm inside the code block\n"
Line 2 of FencedCodeBlock: "hooray!```\n"
Line 3 of FencedCodeBlock: "I come after the code block"

I expected to have a Paragraph, then a code block and then another paragraph after the codeblock and no ' ```' as part of the content.