section selector with level specifier?

[yshavit]

Right now, section selectors don't differentiate between the level of header:

# foo

☝️ this...

## foo

☝️ ... matches the same as this.

Do we want to add an option to specify the level?

[yshavit]

Assuming yes, what would the syntax be?

Suggestion:

{1}# foo

This would match an h1 heading: # foo but not ## foo.

Alternatively, we could do:

#{1} foo

This is clearer, but I'd need to think how it would interact with the possible quoting options being discussed in #56 (e.g. {"EOF"}text text textEOF).

Regardless, we could also add >, <, >=, <= in front of the number:

{>=2}# foo

... would match ## foo and ### foo but not # foo.

[danielwerg]

What about regex style?

#{1}: h1
#{2,3}: h2, h3
#{2,}: h2, h3, h4, h5, h6 (h7? :D)

[yshavit]

Oh, I like that!

[PhilipHayes]

Did we ever get around to this?

[yshavit]

Thanks for the nudge! I had a bunch on my plate, and this fell off my radar.

I have a PR up for it now: #419. Take a look (particularly at the description, and/or integ tests) and tell me what you think!

[PhilipHayes]

yeah that's perfect!

I was going to ask about getting only the headers but I think this would do it:

on Mac for example
find doc/* -name '*.md' -exec mdq "# " {} \; | grep '^#' <- gets ALL headers at All levels
find doc/* -name '*.md' -exec mdq "# " {} \; | grep '^# ' <- gets all h1's
but it's raw strings which is fine. The json output is incredibly helpful. and combined with jq, there's recipe potential.

[yshavit]

This is now possible as of #419, released in v0.10.0.

Syntax:

#{2} "matches exactly h2 sections"
#{2,} "matches h2 - h6 sections"
#{,2} "matches h1 - h2 sections"
#{2,4} "matches h2 - h4 sections"