Specialized Remark that only transforms (or adds) one Markdown section.
Remark works on an abstract syntax tree, which means some information is lost. Not semantics, but formatting, style, whitespace, i.e. your personal preferences. What's more, converting the AST back to markdown will add formatting and possibly result in unwanted diff noise.
Meant for sections of generated markdown, section-remark
runs on a subset of the tree (a range of nodes starting with a certain header), stringifies it and concatenates the result with the rest of your markdown. Other sections are left alone.
Returns a unified
processor, so like the regular remark
, you can use any plugin.
As an example, let's generate a Contributors section:
const remark = require('section-remark')
const contributors = require('remark-contributors')
const report = require('vfile-reporter')
const vfile = require('to-vfile')
remark('Contributors')
.use(contributors)
.process(vfile.readSync('README.md'), function (err, file) {
console.error(report(err || file))
vfile.writeSync(file)
})
This does nothing if the contributors section doesn't already exist. To either replace it or add it to the end of the document, do:
remark('Contributors', { add: true })
To replace it or add it after another section:
remark('Contributors', { after: 'API' })
We can use any test
supported by mdast-util-heading-range
. If we want to generate an API section and insert it after a usage or example section:
remark('API', { after: /^(usage|example)$/i })
Same goes for the first argument. If we want to replace or add a TOC:
remark(/^(table of )?contents$/i, { add: true })
This will throw an error though if the TOC doesn't exist, because section-remark
can't convert the regular expression to a title. We should define the text
option, so that a "Contents" header will be renamed to "Table of Contents". And let's insert the TOC before another section.
remark(/^(table of )?contents$/i, {
text: 'Table of Contents',
before: /^usage|example$/i
})
Lastly, if you use before
or after
and the section does not already exist, its header will inherit the depth of its before
or after
sibling.
Returns a unified
processor that transforms a section starting with a header which matches test
(see mdast-util-heading-range
). Options:
add
: add to end of document, if not foundbefore
: insert before section, if not foundafter
: insert after section, if not foundtext
: text content of header, defaults totest
if a stringdepth
: header depth, starting at 1, defaults to sibling depth or 2.
With npm do:
npm install section-remark
MIT © Vincent Weevers