#2686

[trungtruong1]

I found another issue in the repo, which is #2686. It is a relatively new issue.

[trungtruong1]

I managed to fix this issue.

Before:

After:

I created another branch and pushed the changes.
This issue is about TCL and I have some knowledge about this.

[IshaanVerma1130]

i'll run tests for your commit.

[trungtruong1]

Thanks! @IshaanVerma1130

[IshaanVerma1130]

It is still marking it as TEXT when it should be TEXT>Whitespace @trungtruong1

[trungtruong1]

I think \ and $ shouldn't be marked as TEXT.Whitespace

[IshaanVerma1130]

It's '\n' in the input code. But it's reading it separately as '\' and 'n'.

I believe it should be reading them as newlines as indicated by the comments in the code.

[trungtruong1]

Ah I see, missed that part, I'll fix it

[trungtruong1]

Hello, I made a simple fix to the issue. This is the new output for a test file:

Also, I think it is better to mark \(.) as String.Escape. I use a highlighter in VSCode for Tcl and found out that they do it too.

This is the code highlighted by the extension. They counted the brackets } incorrectly but that does not apply to our case.

This is my result after the fix:

I believe this fixed a HACK in the lexer as well.
The fix is in 0285053

[IshaanVerma1130]

I believe it should still be marked as TEXT.WHITESPACE. Looking at the other Lexers and what they do. Also according to this message, the issue says that some characters should be read as escape character:

and just below that it said the sometimes it adds special meanings probably telling that it should be read as newline or TEXT.WHITESPACE according to the Pygments methodology.

Take a look at this: https://www.learnxinyminutes.com/docs/tcl/
Want to clarify this with the author?

[trungtruong1]

I think asking them would be nice, would you mind adding a comment to the issue asking for clarification?

[IshaanVerma1130]

The author still hasn't replied. T_T

[IshaanVerma1130]

@trungtruong1 The \n at the last should be recognized as TEXT.WHITESPACE but is instead marked as TEXT. Please check.
The other \\n are correctly marked.

[trungtruong1]

That last \n is lexed using the original logic of the lexer. My modification did not result in that. You can check the output example in this:

https://github.com/ZuxingGit/pygments/blob/2686/tests/snippets/tcl/test_vars.txt