feat(markdown-docx): add codeblock transformer - #397

[K-Kumar-01]

Signed-off-by: k-kumar-01 kushalkumargupta4@gmail.com

Improves transformer by allowing codeblock transformation from OOXML to CiceroMark and vice-versa.

Changes

• Add transformation logic: OOXML <-> CiceroMark
• Add rules
• Add test

Screenshots or Video

Author Checklist

• Ensure you provide a DCO sign-off for your commits using the --signoff option of git commit.
• Vital features and changes captured in unit and/or integration tests
• Commits messages follow AP format
• Extend the documentation, if necessary
• Merging to master from fork:branchname

[algomaster99]

Tried to run your code and found that it doesn't parse the literal \\n. Here is a failing test case:

{
  "$class": "org.accordproject.commonmark.Document",
  "xmlns": "http://commonmark.org/xml/1.0",
  "nodes": [
    {
      "$class": "org.accordproject.templatemark.ContractDefinition",
      "name": "top",
      "nodes": [
        {
          "$class": "org.accordproject.commonmark.CodeBlock",
          "text": "cout<<\"\\n\";\n"
        }
      ]
    }
  ]
}

[K-Kumar-01]

Tried to run your code and found that it doesn't parse the literal \\n. Here is a failing test case:

{
  "$class": "org.accordproject.commonmark.Document",
  "xmlns": "http://commonmark.org/xml/1.0",
  "nodes": [
    {
      "$class": "org.accordproject.templatemark.ContractDefinition",
      "name": "top",
      "nodes": [
        {
          "$class": "org.accordproject.commonmark.CodeBlock",
          "text": "cout<<\"\\n\";\n"
        }
      ]
    }
  ]
}

I don't think this would be necessary as the transformer will be used by lawyers primarily and won't generally use \n in contracts.
@dselman Could you please tell if this is necessary?

[algomaster99]

@dselman Could you let us know if contracts have escape sequences, like \n, in code blocks? Currently, when we encounter \n while parsing a codeblock, we convert it to a linebreak. However, \\n is also treated as a line break even though it is literally \n, for example \n in cout<<"\n";. @K-Kumar-01 is arguing that we don't need to parse this literal as it would never be used in contracts. I think if ciceromark_parsed transformer is able to distinguish then our transformer should be capable too.

[dselman]

@dselman Could you let us know if contracts have escape sequences, like \n, in code blocks? Currently, when we encounter \n while parsing a codeblock, we convert it to a linebreak. However, \\n is also treated as a line break even though it is literally \n, for example \n in cout<<"\n";. @K-Kumar-01 is arguing that we don't need to parse this literal as it would never be used in contracts. I think if ciceromark_parsed transformer is able to distinguish then our transformer should be capable too.

In the Commonmark Dingus you can play with code blocks and see how they are rendered as HTML:

<code_block>this is \n a test.

This is more text
</code_block>

Note that neither \n nor \\n are interpreted as newlines, and any formatting (e.g. **) is ignored.

I think we can ignore \\n -- as long as newlines are mapped to/from OOXML and formatting within the code block is ignored.

[dselman]

Typo? This doesn't seem to be related to a thematic break?

[dselman]

Once the typo in the JSDoc is fixed I will merge.

[K-Kumar-01]

@dselman
Made the changes.
Let me know if there is anything else required.

[algomaster99]

Looks good to me!