NanocKt is a pedagogical and experimental toy project meant to mimic the DSL and logic behind the original Nanoc static site generator.
This project shouldn't be used by anyone, there is no goal to make it a product on par with the original Nanoc project.
This is a Frankenstein clone, meaning:
it's ugly, it's barely working, and it's not meant to live for long.
The original motivation was to determine what a Ruby DSL could look like as a Kotlin DSL. How could original syntax be reused almost identically.
A Nanoc website logic mostly resides in the Rules file describing how
to process content (compile rules) and where to put the result (route rules).
Example in Ruby for Nanoc:
preprocess do
# ignore unpublished items
@items.delete_if { |item| item[:published] == false }
end
# simply copy other assets
passthrough '/assets/fonts/**/*'
passthrough '/assets/img/**/*'
passthrough '/attachments/**/*'
compile '/notes/*' do
filter :kramdown
filter :colorize_syntax, default_colorizer: :rouge
layout '/default.*'
end
compile '**/*' do
unless item.binary?
filter :kramdown
item_layout = item[:layout] || 'default'
layout "/#{item_layout}.*" unless item_layout == 'none'
end
end
route '/notes/*' do
raise "item #{item.identifier} doesn't have a slug" if item[:slug].nil?
slug = item[:slug]
post_date = Date.parse(item[:created_at].to_s)
month_2d = format('%02d', post_date.month)
day_2d = format('%02d', post_date.mday)
"/#{post_date.year}/#{month_2d}/#{day_2d}/#{slug}/index.html"
end
route '**/*' do
if item.binary?
item.identifier.to_s
else
"#{item.identifier.without_ext}.html"
end
endEquivalent in Kotlin with NanocKt:
fun main(args: Array<String>) = nanoc(args) {
preprocess {
// ignore unpublished items
items.removeIf { it["published"] != true }
}
// simply copy other assets
passthrough("/assets/fonts/**/*")
passthrough("/assets/img/**/*")
passthrough("/attachments/**/*")
compile("/notes/*") {
filter(commonmark)
filter(colorizeSyntax, mapOf("defaultColorizer" to "rouge"))
layout("/default.*")
}
compile("**/*") {
if (!item.isBinary) {
filter(commonmark)
val itemLayout = item["layout"] ?: "default"
if (itemLayout != "none") {
layout("/${itemLayout}.*")
}
}
}
route("/notes/*") {
checkNotNull(item["slug"]) { "item ${item.identifier} doesn't have a slug" }
val slug = item["slug"] as String
val postDate = LocalDate.parse(item["created_at"] as String, DateTimeFormatter.ISO_DATE)
val month2d = postDate.month.toString().padStart(2, '0')
val day2d = postDate.dayOfMonth.toString().padStart(2, '0')
"/${postDate.year}/${month2d}/${day2d}/${slug}/index.html"
}
route("**/*") {
if (item.isBinary)
item.identifier.toString()
else
"${item.identifier.withoutExt}.html"
}
}- everything should be a rule (incl.
filter,layout, ...) - consider kscript vs
main - no item dependency management
- no auto-prune
- no cache/incremental compilation
- no management of
create/update/skip/identical/delete - no management of item representation
- no management of item snapshot
- no equivalent of
:erbfilter (could kscript be helpful here?) - a lot more!
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent