Extract a part from CommonMark/Markdown docs.

submark: Extract a part from CommonMark/Markdown docs

submark is a CLI program (and GitHub action) to extract some particular section from a given CommonMark/Markdown document. I use it for myself to extract the latest version section from the CHANGELOG.md file, and then reuse the text for the corresponding release note on GitHub releases, during automated release process which is run on CI/CD.

GitHub action: dahlia/submark

Although it's a standalone CLI program, you can use it through a handy GitHub action dahlia/submark:

- id: extract-changelog
  uses: dahlia/[email protected]
  with:
    input-file: CHANGELOG.md
    heading-level: 2
    heading-title-text: version ${{ github.ref_name }}
    ignore-case: true
    omit-heading: true

# The output-file refers to the path of the temporary file which contains
# the only extracted part:
- run: cat ${{ steps.extract-changelog.output.output-file }}

# The output-text contains the text of the extracted part.
- run: echo $CHANGELOG
  env:
    CHANGELOG: ${{ steps.extract-changelog.output.output-text }}

Input parameters

• input-text: The input CommonMark/Markdown text. Mutually exlclusive with the input-file parameter.
• input-file: The input CommonMark/Markdown file path. Mutually exlclusive with the input-text parameter.
• heading-level: The heading level of the section to extract.
• heading-title-text: Extract the section with the exact this heading-title-text (and the heading-level). Note that it tries to match to the heading title with no markup, which means heading-title-text: "foo bar" matches to both # foo bar and # _foo_ **bar**. Mutually exclusive with the heading-title-regex parameter.
• heading-title-regex: Similar to the heading-title-text parameter except that it takes a regular expression. Note that it tries to match to the heading title with no markup, which means heading-title-regex: "fo+ ba[rz]" matches to both # foo bar and # _foooo_ **baz**. Mutually exclusive with the heading-title-text parameter.
• ignore-case: Ignore case distinctions. (Default: false.)
• omit-heading: Omit a leading heading. (Default: false.)
• columns: Limit the maximum characters per line of the output. No limit by default.

Output parameters

• output-text: The text of the extracted part.
• output-file: The path to the temporary file which contains the only extracted part.

Download & installation

First of all, if you need to manually invoke submark on GitHub Actions, dahlia/submark/setup action is the easiest way to install it.

On the other CI/CD products, use the officially distributed executables. Prebuilt binaries for the following platforms and architectures are available on GitHub releases:

• Linux (x86_64)
• macOS (x86_64)
• Windows (win64)

For other platforms and architectures, you need to build it by yourself. It's written in Haskell, so you need to install Haskell Stack first. It can be built in the same way other Haskell programs are:

$ stack setup && stack install

Usage examples

The following examples use John Gruber's original Markdown introduction:

$ wget https://daringfireball.net/projects/markdown/index.text

Extracting an H3 heading Command-Line:

$ submark --h3 "Command-Line" index.text
### Command-Line

Use the `--html4tags` command-line switch to produce HTML output from a
Unix-style command line. E.g.:

    % perl Markdown.pl --html4tags foo.text

Type `perldoc Markdown.pl`, or read the POD documentation within the
Markdown.pl source code for more information.

Note that its textual style differ from the original text. Since submark internally builds an abstract syntax tree from the given input text and then render again the result tree, it doesn't maintain trivial styles (e.g., whitespaces, ATX vs. Setext headings) but only the semantics.

There are options from --h1 to --h6:

$ submark --h2 "Download" index.text
## Download

[Markdown 1.0.1](http://daringfireball.net/projects/downloads/Markdown_1.0.1.zip)
(18 KB) -- 17 Dec 2004

Options from --h1-regex to --h6-regex take a regular expression instead of an exact heading text:

$ submark --h2-regex "^(Down|Up)load$" index.text
## Download

[Markdown 1.0.1](http://daringfireball.net/projects/downloads/Markdown_1.0.1.zip)
(18 KB) -- 17 Dec 2004

The leading heading can be omitted:

$ submark --h2 "Download" --omit-heading index.text
[Markdown 1.0.1](http://daringfireball.net/projects/downloads/Markdown_1.0.1.zip)
(18 KB) -- 17 Dec 2004

Matching is case sensitive by default, but case can be ignored using the option -i/--ignore-case:

$ submark --h2 "DOWNload" index.text

$ submark --h2 "DOWNload" --ignore-case index.text
## Download

[Markdown 1.0.1](http://daringfireball.net/projects/downloads/Markdown_1.0.1.zip)
(18 KB) -- 17 Dec 2004

By Unix convention, - means pipe:

$ submark --h2 "Download" - < index.text
## Download

[Markdown 1.0.1](http://daringfireball.net/projects/downloads/Markdown_1.0.1.zip)
(18 KB) -- 17 Dec 2004

There's -o/--out-file option as well:

$ submark -o download.md --h2 "Download" index.text

Software license

This software is distributed under GPL 3.